Alpha 1,3-fucosyltransferase-VII regulates the signaling
molecules of the insulin receptor pathway
Qiu-yan Wang
1
, Ying Zhang
1
, Hai-jiao Chen
2
, Zong-hou Shen
1
and Hui-li Chen
1
1 Key Laboratory of Glycoconjugate Research, Shanghai Medical College, Fudan University, Shanghai, China
2 Department of Urology, Zhong-shan Hospital, Fudan University, Shanghai, China
Glycosylation is important and the most common form
of post-translational modification that regulates many
aspects of protein function [1,2]. In recent years,
increased attention has been paid to the relationship
between structural changes in surface glycans and sur-
face receptor signaling. It has been reported [3] that
overexpression of N-acetylglucosaminyltransferase
(GnT)-III introducing a bisecting N-acetylglucosamine
Keywords
epidermal growth factor receptor; a1,3-
fucosyltransferase-VII; human
hepatocarcinoma cell line; insulin receptor;
signaling molecules
Correspondence
H. Chen, Key Laboratory of Glycoconjugate
Research, Ministry of Health, Department of
Biochemistry, Shanghai Medical College,

Fudan University, Shanghai 200032, China
Fax: + 86 21 6416 4489
Tel: + 86 21 5423 7223
E-mail:
(Received 19 July 2006, revised 13 November
2006, accepted 17 November 2006)
doi:10.1111/j.1742-4658.2006.05599.x
Two H7721 human hepatocarcinoma cell lines showing moderate and high
expression of a1,3-fucosyltransferase (FucT)-VII cDNA were established
and designated FucTVII-M and FucTVII-H, respectively. In a1,3-FucT-
VII-transfected cells, expression of insulin receptor (InR) a- and b subunits
and epidermal growth factor receptor (EGFR) on the cell surface and in
cells, as well as the sialyl Lewis X (SLe
x
, the product of a1,3-FucT-VII)
content of the EGFR were unchanged. However the level of SLe
x
on the
InR a subunit (InR-a) was increased dramatically. Tyrosine autophosphory-
lation of InR-b , but not EGFR, was elevated. Concomitantly, tyrosine
phosphorylation of insulin receptor substrate-1 (IRS-1), Ser ⁄ Thr phos-
phorylation of protein kinase B (PKB; Akt), p42 ⁄ 44 mitogen-activated pro-
tein kinase (MAPK), MAPK kinase (MEK), and the protein of some other
signaling molecules, such as phosphoinositide-dependent kinase-1 (PDK-1),
novel protein kinase (PKN), c-Raf-1 and b-catenin were also upregulated.
The activities of PKB and transcription factor TCF were concomitantly sti-
mulated. Upregulation of InR signaling molecules and their phosphoryla-
tion was correlated with the level of SLe
x
on InR-a and a1,3-FucT-VII

expression in cells. In addition, the phosphorylation intensity and differ-
ence in phosphorylation intensity between cells with different levels of
a1,3-FucT-VII expression were attenuated significantly by the inhibitor of
InR tyrosine kinase and by the mAb to SLe
x
. Furthermore, insulin-induced
signaling was facilitated in a1,3-FucT-VII-transfected cells, particularly
FucTVII-H. These findings provide strong evidence that a1,3-FucT-VII
may affect insulin signaling by upregulating the phosphorylation and
expression of some signaling molecules involved in the InR-signaling path-
way. These effects are likely mediated by its product, SLe
x
, on the glycans
of the InR. This is the first study to report that changes in the terminal
structure of glycans on a surface receptor can modify cell signaling.
Abbreviations
CDK, cyclin-dependent kinase; EGF, epidermal growth factor; EGFR, epidermal growth factor receptor; FucT, fucosyltransferase; GlcNAc,
N-acetylglucosamine; GnT, N-acetylglucosaminyltransferase; InR, insulin receptor; IRS-1, insulin receptor substrate-1; MAPK, mitogen-
activated protein kinase; MEK, MAPK kinase; NGF, nerve growth factor; PDK-1, phosphoinositide-dependent kinase-1; PKB, protein kinase B
(Akt); PKN, novel protein kinase; TGF, transforming growth factor.
526 FEBS Journal 274 (2007) 526–538 ª 2006 The Authors Journal compilation ª 2006 FEBS
(GlcNAc) into the N-glycans of epidermal growth fac-
tor receptor (EGFR) in U373 MG glioma cells led to
decreased epidermal growth factor (EGF) binding and
autophosphorylation of EGFR, as well as reduced cell
proliferation upon EGF stimulation. It has also been
reported [4] that overexpression of GnT-III in pheo-
chromocytoma PC12 cells inhibited transient tyrosine
phosphorylation and dimerization of the nerve growth
factor (NGF) receptor (Trk) upon stimulation with

NGF and resulted in blockage of the neurite outgrowth
during differentiation. We previously found [5] that
transfection of the sense cDNA of GnT-V, an enzyme
associated with cancer progression and metastasis, into
human H7721 hepatocarcinoma cells resulted in an
increase in the level of GlcNAcb1,6Mana1,6-branch
(GnT-V product) on the N-glycans of EGFR; this pro-
moted EGF binding and tyrosine autophosphorylation,
but had little effect on expression of the EGFR protein.
The phosphorylation (at T308, S473 and tyrosine resi-
dues) and activity of protein kinase B (PKB; Akt), as
well as the phosphorylation of p42 ⁄ 44 mitogen-activa-
ted protein kinase (MAPK; ERK-1 ⁄ 2) and MAPK
kinase (MEK) before and after EGF stimulation, were
also upregulated. Conversely, in H7721 cells expressing
antisense GnT-V (GnTV-AS), the results were the oppo-
site of those seen in GnT-V sense cDNA (GnTV-S)-
transfected cells. After GnT-V-transfected H7721 cells
were treated with 1-deoxymannojirimycin, an inhibitor
of N-glycan processing at the high mannose stage, or
antibody against the extracellular glycan domain of
EGFR, the increase in PKB activity and MAPK phos-
phorylation were significantly blocked, and the differ-
ences in PKB activity and MAPK phosphorylation
among GnTV-S, GnTV-AS and mock-transfected cells
(cells transfected with empty vector) were attenuated
significantly. These findings indicated that the altered
signaling after GnTV-S or GnTV-AS transfection was
mediated by a structural change in N-glycans on the
EGFR [5]. Furthermore, Guo et al. [6] reported that

transfection of GnT-V into human fibrosarcoma
HT1080 and mouse NIH 3T3 cells to increase the Glc-
NAcb1,6-branch on N-cadherin inhibited signaling
between N-cadherin and ERK1 ⁄ 2, and consequently
reduced calcium-dependent cell–cell adhesion mediated
by N-cadherin. These results provide evidence that the
N-cadherin signaling pathway is also influenced by the
glycan structures on N-cadherin. Wang et al. [7] repor-
ted that in embryonic fibroblast cells deprived of a1,6-
fucosyltransferase (FucT-VIII), an enzyme responsible
for the synthesis of core fucose on N-glycans, EGF-
induced phosphorylation of EGFR and EGFR-
mediated JNK or ERK activation were suppressed.
Taniguchi [8] also discovered that signal transduction
of the transforming growth factor b1 (TGF b1) recep-
tor was deficient in FucT-VIII knockout mice, leading
to emphysema-like phenotypes in the lung. These
results show that the core fucose on N-glycans is essen-
tial for EGF and TGF b1 signaling. However, all the
above-mentioned structural changes in receptor glycans
are located in the core portion of N-glycan, and whe-
ther alteration of the terminal residue on the outer
chain of either N-orO-glycan can also modify surface
receptor signaling remains unclear.
It has been documented that sialyl Lewis antigens
(SLe) expressed on the surface of cancer cells, such as
SLe
x
[SAa2,3 Galb1,4 (Fuca1,3) GlcNAc-] and SLe
a

[SAa2,3 Galb1,3 (Fuca1,4) GlcNAc-], are another kind
of glycan structure involved in metastasis, and which
can serve as the ligands for E- or P-selectin expressed
on the surface of vascular endothelial cells and mediate
the adhesion of malignant cells to vascular endothe-
lium [9–11]. The final fucosylation step in Lewis anti-
gen synthesis is catalyzed by a1,3-fucosyltransferase
(a1,3-FucT). To date, six a1,3-FucTs (III to VII and
IX) have been identified. Each enzyme has a unique
acceptor–substrate binding pattern, and each generates
a unique range of fucosylated products [12,13]. Among
these, a1,3-FucT-VII, which is expressed mainly in leu-
kocytes, catalyzes sialylated substrate and produces
SLe
x
as its only product [14]. SLe
x
is always located at
the terminus of sugar chains, and a1,3-FucT-VII may
be considered a terminal glycosyltransferase that cata-
lyzes the final step in sugar-chain processing.
In our laboratory, it has been found that the surface
SLe
x
and cellular a1,3-FucT-VII of H7721 cells is
up- and downregulated by transfection of the metasta-
sis-promoting gene c-erbB2 ⁄ neu and the metastasis-
suppressive gene nm23-H1, respectively [15–17]. In
addition, surface SLe
x

was increased when H7721 cells
were treated with proliferative inducers, and decreased
after treatment with differentiative inducers [18]. The
change in SLe
x
level was proportional to a1,3-FucT-
VII expression. Moreover, the ex vivo metastatic
potential was positively correlated with surface SLe
x
and cellular a1,3-FucT-VII levels, and could be inhib-
ited by a mAb (KM93) against surface SLe
x
[15,18].
Further studies have shown that insulin also enhanced
expression of SLe
x
and a1,3-FucT-VII and the meta-
static potential of H7721 cells [19]. In addition, our
group recently found that expression of cyclin-depend-
ent kinase (CDK) inhibitor, p27
Kip1
protein, was
decreased in H7721 cells transfected with a1,3-FucT-
VII cDNA. Uninhibited CDK2 resulted from a reduc-
tion in the p27
Kip1
-stimulated phosphorylation of
retinoblastoma protein, facilitating G
1
⁄ S transition and

increasing the growth rate in the cells. These effects
Q. Wang et al. Fucosyltransferase-VII regulates insulin signaling
FEBS Journal 274 (2007) 526–538 ª 2006 The Authors Journal compilation ª 2006 FEBS 527
were correlated with an increase in surface SLe
x
on
H7721 cells expressing different a1,3-FucT-VII intensi-
ties, and could be blocked by SLe
x
antibody in a dose-
dependent manner, indicating that p27
Kip1
expression
was influenced by a1,3-FucT-VII and its product SLe
x
[20]. Therefore, it is interesting to study whether trans-
fection of a1,3-FucT-VII can also affect the function
of some other surface receptors and subsequently
result in altered receptor signaling.
Insulin receptor (InR) was selected to study the
effects of a1,3FucT-VII on its expression, SLe
x
content
and tyrosine autophosphorylation. InR contains two
extracellular carbohydrate-containing a subunits and
two b subunits with cytoplasmic tyrosine kinase activ-
ity [21]. InR results were compared with those from
EGFR, which also contains N-glycans on its extracel-
lular domain and tyrosine autophosphorylation sites
at its intracellular domain [22]. Furthermore, insulin

receptor substrate-1 (IRS-1), PKB, phosphoinositide-
dependent kinase-1 (PDK-1), novel protein kinase
(PKN), p42 ⁄ 44 MAPK and MEK were analyzed as
the signaling molecules involved in InR signaling
[19,23]. Expression of b catenin and its downstream
transcription factor TCF in the Wnt signaling pathway
[24], which cross-talks with the InR pathway was
also studied. Mock cells transfected with the vector
pcDNA3.1 were used as controls.
Results
Characterization of two a1,3-FucT-VII-transfected
cell lines
As shown in Fig. 1A,B, a1,3-FucT-VII mRNA was
increased significantly in H7721 cells transfected with
a1,3-FucT-VII cDNA. In FucTVII-M (moderate
expression) and FucTVII-H (high expression) cells, it
was upregulated to 373.3 and 613.3% of the mock-
transfected cell level, respectively (both P < 0.01).
Consequently, expression of SLe
x
, the product of a1,3-
FucT-VII, was also elevated on the cell surface, to
171 and 284% of the mock-transfection value in
FucTVII-M and FucTVII-H cells, respectively (both
P < 0.01; Fig. 1C,D).
Expression of InR-a, EGFR and their SLe
x
in
a1,3-FucT-VII-transfected H7721 cells
Expression of cell-surface InR-a and EGFR were ana-

lyzed using specific antibodies and flow cytometry. The
0
0.4
0.8
1.2
Mock FucTVII-M FucTVII-H
Mock FucTVII-M FucTVII-H
nitca-
a
teb/IIV-
T
c
u
F
Mock
B
FucTVII-M
FucTVII-H
*
*
0
40
80
120
160
elSfonoisserpxE
x
*
*
A

D
C
Mock
FucTVII-M
FucTVII-M FucTVII-H
FucT-VII
β
-actin
FucTVII-H
497 bp
789 bp
0
10
0
10
1
10
2
10
3
10
4
FL1-H
( - ) Control Mock
10
0
10
1
10
2

10
3
10
4
FL1-H
10
0
10
1
10
2
10
3
10
4
FL1-H
10
0
10
1
10
2
10
3
10
4
FL1-H
M1
M1
M1M1

200
Counts
0 200
Counts
0 200
Counts
0 200
Counts
Fig. 1. Characterization of a1,3-FucT-VII cDNA-transfected H7721 cells. (A) RT-PCR profiles of a1,3-FucT-VII mRNA in mock- and a1,3-FucT-
VII-transfected cells. (B) Relative expressions of a1,3-FucT-VII mRNA in mock- and a1,3-FucT-VII transfected cells (n ¼ 3). (C) Fluorescence-
activated cell spectra of cell-surface SLe
x
on mock- and a1,3-FucT-VII transfected cell lines. (D) Relative expressions of surface SLe
x
on
mock- and a1,3-FucT-VII transfected cells (n ¼ 3). Mock, H7721 cells transfected with pcDNA3.1 vector; FucTVII-M, H7721 cell line with
moderate expression of transfected a1,3-FucT-VII; FucTVII-H, H7721 cell line with high expression of transfected a1,3-FucT-VII. (A) and (C)
are representative of three reproducible experiments. *P < 0.01 compared with ‘Mock’. RT-PCR and flow cytometry are described in the
Experimental procedures.
Fucosyltransferase-VII regulates insulin signaling Q. Wang et al.
528 FEBS Journal 274 (2007) 526–538 ª 2006 The Authors Journal compilation ª 2006 FEBS
results in Fig. 2A,B show that their expression was not
obviously changed in a1,3-FucT-VII-transfected cells
compared with mock-transfected cells. Results from
western immunoblots indicated that protein expression
in InR-a and EGFR was also unchanged following
transfection with a1,3-FucT-VII (Fig. 2C). However,
after immunoprecipitation and western blotting of
these receptors, and using KM93 as the probe for
SLe

x
, it was found that expression of SLe
x
on InR-a
of FucTVII-M and FucTVII-H cells was increased to
248 and 409% of the mock-transfection level, respect-
ively (both P < 0.01), whereas expression of SLe
x
on
EGFR remained unchanged (Fig. 2D,E).
Tyrosine phosphorylation of InR-b, EGFR and
IRS-1 in a1,3-FucT-VII-transfected H7721 cells
As shown in Fig. 3A,B, the amount of immuno-
precipitated InR-b was also unchanged following
a1,3-FucT-VII transfection. The relative intensity of
tyrosine autophosphorylation in immunoprecipatated
InR-b or EGFR was calculated from the intensity
ratio of the phosphorylated band to the unphos-
phorylated band. Figure 3B also shows that tyrosine
autophosphorylation of InR-b was increased to 186
and 352% of the mock-transfection value in
FucTVII-M and FucTVII-H cells, respectively
0
30
60
90
120
150
EGFRInR
Mean fluorescence intensity

Mock
FucTVII-M
FucTVII-H
AB
0
10
0
10
1
10
2
10
3
10
4
FL1-H
M1
( - ) Control-InR ( - ) Control-EGFR
FucTVII-H-InR
0
50
100
150
200
250
300
350
400
450
500

EGFR EGFR SLeX
Relative expression
Mock
FucTVII-M
FucTVII-H
InR-
α
InR-
α
SLeX
EC
InR α
EGFR
β-actin
Mock FucTVII-M FucTVII-H
FucTVII-M-InRMock-InR
FucTVII-H-EGFRFucTVII-M-EGFRMock-EGFR
200
Counts
0
10
0
10
1
10
2
10
3
10
4

FL1-H
M1
200
Counts
0
10
0
10
1
10
2
10
3
10
4
FL1-H
M1
M1
200
Counts
0
10
0
10
1
10
2
10
3
10

4
FL1-H
200
Counts
0
10
0
10
1
10
2
10
3
10
4
FL1-H
200
Counts
0
10
0
10
1
10
2
10
3
10
4
FL1-H

10
0
10
1
10
2
10
3
10
4
FL1-H
M1
200
Counts
0
200
Counts
10
0
10
1
10
2
10
3
10
4
FL1-H
M1
M1

M1
0 200
Counts
D
Mock
IP: InR-α
WB: SLeX
WB: SLeX
IP: EGFR
WB: EGFR
IP: EGFR
IP: InR-α
WB: InR-α
FucTVII-M FucTVII-H
Fig. 2. Effects of a1,3-FucT-VII transfection on expression of InR-a and EGFR and the SLe
x
content of the glycans of InR-a and EGFR.
(A) Fluorescence-activated cell spectra of InR-a and EGFR on the cell surface. (B) Relative expression of surface InR-a and EGFR (n ¼ 3).
(C) Western immunoblot profiles of InR-a and EGFR after staining with an antibody to InR-a or EGFR and horseradish peroxidase-labeled sec-
ondary antibody. (D) Western immunoblot profiles of immunoprecipited InR-a and EGFR (precipitated by CF4) after staining with antibody to
SLe
x
(KM93), InR-a or EGFR (antibody 528) and horseradish peroxidase-labeled secondary antibody to determine the SLe
x
content of InR-a
and EGFR. (E) Densitometric quantification of (D) (n ¼ 3). Mock, H7721 cells transfected with pcDNA3.1 vector; FucTVII-M, H7721 cell line
with moderate expression of transfected a1,3-FucT-VII; FucTVII-H, H7721 cell line with high expression of transfected a1,3-FucT-VII; InR,
insulin receptor a-subunit; EGFR: epidermal growth factor receptor; IP, immunoprecipitation by the antibody to the protein indicated at the
right; WB, western immunoblot with the antibody to the compound indicated at the right. (A), (C) and (D) are representative of three repro-
ducible experiments. *P < 0.01 compared with ‘Mock’. Flow cytometry, immunoprecipitation and western immunoblot are described in the

Experimental procedures.
Q. Wang et al. Fucosyltransferase-VII regulates insulin signaling
FEBS Journal 274 (2007) 526–538 ª 2006 The Authors Journal compilation ª 2006 FEBS 529
(both P < 0.01), whereas that of EGFR was
unchanged.
Tyrosine phosphorylation of IRS-1 occurs earlier in
insulin signaling. The IRS-1 protein was decreased in
a1,3-FucT-VII-transfected cells, although the change
was not statistically significant. When relative tyrosine
phosphorylation was calculated as above, it was found
that the level of phosphorylated IRS-1 was increased
to 2.8 and 8.5 times that of the mock-transfection
value in FucTVII-M and FucTVII-H cells, respectively
(both P < 0.01; Fig. 3C,D).
Phosphorylation and activity of PKB, expression
of PDK-1, PKN and phospho-PKN in
a1,3-FucT-VII-transfected H7721 cells
In insulin signaling, activation of PKB has been impli-
cated as a key step and it also has a major role in the
physiological effects of insulin [25]. As shown in
Fig. 4A,B, expression of PKB protein was not obvi-
ously altered in a1,3-FucT-VII-transfected H7721 cells,
but relative phosphorylation at T308 and S473 in PKB
(calculated from the ratio of the staining intensity of
phosphorylated protein to unphosphorylated protein
after normalization with b-actin) was apparently eleva-
ted when compared with mock-transfected cells. After
densitometric quantification, relative T308 phosphory-
lation was 149 and 205% of the mock-transfection
level in FucTVII-M and FucTVII-H cells, respectively

(both P < 0.01). Meanwhile, relative S473 phosphory-
lation was 170 and 315% of the mock-transfection
value, respectively (both P < 0.01). Increased phos-
phorylation of PKB at both T308 and S473 resulted in
an upregulation of PKB activity, measured as the
amount of phosphorylated GSK3a ⁄ b product. As indi-
cated in Fig. 4C,D, phosphorylated GSK3a ⁄ b was ele-
vated to 165 and 270% of the mock-transfected value
in FucTVII-M and FucTVII-H cells, respectively (both
P < 0.01).
PDK-1 is the enzyme responsible for PKB phosphory-
tion [26], and PKN is another substrate of PDK-1 rela-
ted to cytoskeleton and transcription factor [27].
Figure 4E,F shows that the PDK-1 protein was upregu-
lated to 162 and 198% of the control value in FucTVII-
M and FucTVII-H cells, respectively (both P < 0.01).
Protein expression of PKN and Ser ⁄ Thr phosphoryla-
tion of PKN (p-PKN) were also increased to a similar
degree. Therefore, the relative phosphorylation of
p-PKN (pPKN ⁄ PKN protein) was generally unchanged.
Expression and phosphorylation of c-Raf-1, MEK
and p42 ⁄ 44 MAPK in a1,3-FucT-VII-transfected
H7721 cells
The Ras–Raf–MEK–MAPK pathway is another
important signaling pathway in addition to the
AB
0
50
100
150

200
250
300
350
400
450
Relative intensity
kcoM
M
-
II
V
Tc
u
F
H
-IIVT
c
u
F
InR-β InR-β-Tyr-p
EGFR-Tyr-pEGFR
D
0
100
200
300
400
500
600

IRS-Tyr pIRS-1
Relative intensity
k
c
oM
M-IIVTcuF
H-IIVTcuF
Mock FucTVII-M FucTVII-H
IP: InR-β
WB: PT66
C
Mock
FucTVII-M
FucTVII-H
IP: IRS-1
WB: PT66
IP: IRS-1
WB: IRS-1
WB: PT66
IP: EGFR
WB: EGFR
IP: EGFR
IP: InR-β
WB: InR-β
Fig. 3. Effects of a1,3-FucT-VII transfection on the tyrosine autophosphorylation of two receptors and of IRS-1. (A) Western immunoblot pro-
files of immunoprecipitated InR-b and EGFR (precipitated by CF4) after staining with antibody to phosphotyrosine (PT66), InR-b or EGFR
(CF4) and horseradish peroxidase-labeled secondary antibody. (B) Densitometric quantification of (A) (n ¼ 3). (C) Western immunoblot pro-
files of immunoprecipitated IRS-1 after staining with phosphotyrosine antibody (PT66), IRS-1 antibody and horseradish peroxidase-labeled
secondary antibody. (D) Densitometric quantification of (C) (n ¼ 3). Mock, H7721 cells transfected with pcDNA3.1 vector; FucTVII-M, H7721
cell line with moderate expression of transfected a1,3-FucT-VII; FucTVII-H, H7721 cell line with high expression of transfected a1,3-FucT-VII;

WB, western immunoblot with the antibody to phosphotyrosine (PT66) or to the protein indicated on the right; IRS-1, insulin receptor sub-
strate-1; Tyr-p, tyrosine phosphorylated. (A) and (C) are representative of three reproducible experiments. *P < 0.01 compared with ‘Mock’.
Immunoprecipitation and western immunoblotting are described in the Experimental procedures.
Fucosyltransferase-VII regulates insulin signaling Q. Wang et al.
530 FEBS Journal 274 (2007) 526–538 ª 2006 The Authors Journal compilation ª 2006 FEBS
PDK-1 ⁄ PKB pathway in the insulin receptor [23]. Fig-
ure 5A,B shows that expression of MEK and p42 ⁄ 44
MAPK proteins was not apparently altered in a1,
3-FucT-VII-transfected H7721 cells. However, expres-
sion of c-Raf-1 increased significantly following a1,3-
FucT-VII transfection, being 168 and 325% of the
mock-transfection value in FucTVII-M and FucTVII-
H cells, respectively (both P < 0.01). Moreover, the
relative phosphorylation of MEK, as determined by
the ratio of p-MEK to MEK, was upregulated to 207
and 425% in FucTVII-M and FucTVII-H cells, respec-
tively (both P < 0.01), and the relative phosphoryla-
tion of p42 ⁄ 44 MAPK (the ratio of p-p42 ⁄ 44 MAPK
to p42⁄ 44 MAPK) was also increased in FucTVII-M
and FucTVII-H cells, being 2.82 and 6.01 times the
mock-transfection value (both P < 0.01).
Effect of HNMPA-(AM)
3
and KM93 on the
phosphorylation of PKB and p42/44 MAPK
In order to study whether the alteration in the phos-
phorylation of PKB and p42 ⁄ 44 MAPK was mediated
by InR kinase and surface SLe
x
, phosphorylation of

these two signaling molecules was determined before
and after cultured cells were treated with 50 lm
HNMPA-(AM)
3
(a specific inhibitor of InR tyrosine
kinase) [28] or 30 lgÆmL
)1
KM93 (SLe
x
antibody) for
24 h; corresponding untreated cells were used as the
control. It was found that the results from the untreated
cells were the same as those shown in Figs 4 and 5.
When H7721 cells were treated with HNMPA-(AM)
3
,
phosphorylation of PKB at both T308 and S473, and of
p42 ⁄ 44 MAPK was apparently decreased in mock- and
a1,3-FucT-VII- transfected cells (Fig. 6A). The decrease
in phosphorylation of PKB and p42 ⁄ 44 MAPK was
~ 40.9–76.5% (P<0.01) in a1,3-FucT-VII-transfected
cells, compared with the corresponding untreated cells
(Fig. 6B). By contrast, differences in phosphorylation
intensity for PKB and MAPK among mock, FucTVII-
M and FucTVII-H cell groups were attenuated in
HNMPA-(AM)
3
-treated cells (Fig. 6 A,B). Similarly, in
the presence of KM93, phosphorylation of both PKB
and p42 ⁄ 44 MAPK, and the differences in their phos-

phorylation intensities among the three cell lines were
also decreased significantly (Fig. 6C,D). The reduction
in phosphorylation of PKB and p42 ⁄ 44 MAPK in
a1,3-FucT-VII-transfected cells was  41.1–89.7%
(P<0.01). In the presence of HNMPA-(AM)
3
or
KM93, the rate of inhibition of phosphorylation was
correlated with expression of a1,3-FucT-VII, which was
FucTVII-H > FucTVII-H > mock-transfected cells.
However, some differences in the phosphorylation
intensities of PKB and MAPK were observed in mock-
and a1,3-FucT-VII-transfected cells in the presence of
both inhibitors, but the differences were either not sta-
tistically significant or P < 0.05.
AC
D
E
F
B
0
05
001
0
5
1
002
052
003
053

004
p374S-BKP
p
803T-BKP
Relative phosphorylation of PKB
kcoM
M-
I
IVTcuF
H-I
I
VTc
uF
0
100
200
300
400
Mock FucTVII-M FucTVII-H
ytivitc
aesanikBKPevitaleR
0
100
200
300
400
500
600
PDK-1 PKN p-PKN
y

ti
sn
e
tni
e
vi
t
aleR
Mock
FucT-VII-M
FucT-VII-H
Mock
PKB-T308
PKB-S473
PKB
β-actin
FucTVII-M FucTVII-H
Mock
Phosphorylated
GSK3 α/β
β-actin
FucTVII-M FucTVII-H
Mock
PDK-1
PKN
p-PKN
β-actin
FucTVII-M FucTVII-H
Fig. 4. Effects of a1,3-FucT-VII transfection on the phosphorylation, protein expression and activity of some signaling molecules. (A) Western
blot profiles of PKB and T308-, S473-phosphorylated PKB after staining with specific antibodies and horseradish peroxidase-labeled sec-

ondary antibody. (B) Densitometric quantification of (A) (n ¼ 3). (C) Determination of PKB activity as the amount of phosphorylated GSK3a ⁄ b
product. (D) Densitometric quantification of (C) (n ¼ 3). (E) Western immunoblot profiles of PDK-1, PKN and p-PKN after staining with speci-
fic antibodies and horseradish peroxidase-labeled secondary antibody. (F) Densitometric quantification of E (n ¼ 3). Mock, H7721 cells trans-
fected with pcDNA3.1 vector; FucTVII-M, H7721 cell line with moderate expression of transfected a1,3-FucT-VII; FucTVII-H, H7721 cell line
with high expression of transfected a1,3-FucT-VII; PKB-T308p, phosphorylated PKB at T308; PKB-S473p, phosphorylated PKB at S473. (A),
(C) and (E) are representative of three reproducible experiments. *P < 0.01 compared with ‘Mock’. Western blotting, western immunoblot-
ting and assay of PKB activity are described in the Experimental procedures.
Q. Wang et al. Fucosyltransferase-VII regulates insulin signaling
FEBS Journal 274 (2007) 526–538 ª 2006 The Authors Journal compilation ª 2006 FEBS 531
Expression of b-catenin and TCF in
a1,3-FucT-VII-transfected H7721 cells
Beta-catenin is a substrate of GSK-3 and a key mole-
cule in the Wnt and TGF-b signaling pathways [29–
31]. As shown in Fig. 7A,B, the level of b-catenin was
upregulated in a1,3-FucT-VII-transfected cells, to
242% in FucTVII-M cells and 504% in FucTVII-H
cells (both P < 0.01).
Luciferase activity was measured as an indicator of
the activity of transcription factor TCF. As shown in
Fig. 7C, TCF activity was also increased in FucTVII-
M and FucTVII-H cells, being 239 and 333% of the
mock-transfection level, respectively (both P < 0.01).
Modification of insulin signaling in
a1,3-FucT-VII-transfected cells
The effect of the above-mentioned changes in insulin-
signaling molecules on transduction of the insulin sig-
nal was further studied in a1,3-FucT-VII-transfected
cells following serum starvation. Phosphorylation at
T308 and S473 of PKB and p42 ⁄ 44 MAPK was also
selected as an indicator of signaling efficiency. It was

found that phospho-PKB-S473 was barely seen in
insulin-untreated and serum-starved cells, but was
expressed in insulin-treated cells. By contrast, phospho-
PKB-T308 and phospho-p42 ⁄ 44 MAPK were expres-
sed in both insulin-untreated and insulin-treated cells.
The intensity levels for phospho-PKB-T308 and phos-
pho-MAPK in both insulin-untreated and -treated
cells, as well as the phospho-PKB-S473 in insulin-trea-
ted cells, were FucTVII-H > FucTVII-M > mock
(Fig. 8A). In the presence of insulin, phosphorylation
of PKB and MAPK was obviously upregulated, and
was significantly higher than in the corresponding
control cells cultured in the absence of insulin. The
response to insulin stimulation was proportional to the
expression of a1,3-FucT-VII. In insulin-stimulated
FucTVII-M and FucTVII-H cells, phospho-PKB-T308
was upregulated to 215 and 398% of the mock-transfc-
tion level (both P < 0.01), and phospho-PKB-S473
was upregulated to 192 and 354% of the mock-trans-
fection level, respectively (both P < 0.01). Similarly,
phospho-MAPK was 184 and 345% of the mock-
transfection level, respectively (both P < 0.01)
(Fig. 8B).
Discussion
The results shown in Fig. 1A,B indicate that two a1,3-
FucT-VII-transfected cell lines were established with
moderate and high expression of the exogenous
cDNA. Expression of SLe
x
, the product of a1,3-FucT-

VII, was positively correlated with expression of a1,3-
FucT-VII mRNA (Fig. 1C,D).
After transfection of the a1,3-FucT-VII cDNA, pro-
tein expression of InR (including the a- and b-sub-
units) and EGFR both on the cell surface and in cells
(Fig. 2A,C), as well as the SLe
x
content of the glycans
of EGFR were unchanged, but the SLe
x
content of
the glycans of InR-a was increased significantly
(Fig. 2D,E). If the SLe
x
of InR-a glycans is compared
with that of EGFR in mock-transfected cells, it is
observed that the SLe
x
content of EGFR is far greater
than that of InR-a (Fig. 2D), suggesting that the SLe
x
on EGFR is high enough and cannot be upregulated
further by overexpression of a1,3-FucT-VII. This may
A
0
002
004
00
6
0

08
4
4
/24
p
KE
M
fa
R-c
KP
A
M
44/
24
p-
p
KEM-
p
KP
AM
Relative intensity
k
co
M
B
M-I
IV
TcuF
H-IIVTcuF
Mock

c-Raf
p-MEK
MEK
p-p42/44
MAPK
p42/44
MAPK
β-actin
FucTVII-M FucTVII-H
Fig. 5. Effects of a1,3-FucT-VII transfection on the expression of
c-Raf-1 and phosphorylation of MEK and p42 ⁄ 44 MAPK. (A) West-
ern immunoblot profiles of c-Raf-1, p-MEK, MEK, p-p42 ⁄ 44 MAPK
and p42 ⁄ 44 MAPK after staining with specific antibodies and horse-
radish peroxidase-labeled second antibody. (B) Densitometric quan-
tification of (A) (n ¼ 3). Mock, H7721 cells transfected with
pcDNA3.1 vector; FucTVII-M, H7721 cell line with moderate
expression of transfected a 1,3-FucT-VII; FucTVII-H, H7721 cell line
with high expression of transfected a1,3-FucT-VII; p-MEK, phos-
phorylated MEK; p-p42 ⁄ 44 MAPK, phosphorylated p42 ⁄ 44 MAPK.
(A) is representative of three reproducible experiments. *P < 0.01
compared with ‘Mock’. Western immunoblotting is described in the
Experimental procedures.
Fucosyltransferase-VII regulates insulin signaling Q. Wang et al.
532 FEBS Journal 274 (2007) 526–538 ª 2006 The Authors Journal compilation ª 2006 FEBS
be one reasons why a1,3-FucT-VII did not increase
the amount of SLe
x
on EGFR. Alternatively, the com-
position and structure of EGFR glycans probably dif-
fer from those of InR-a, and the EGFR glycans are

not suitable substrates for fucosylation by exogenous
a1,3-FucT-VII. In other words, SLe
x
on EGFR is
probably not synthesized by a1,3-FucT-VII, but by
other a1,3-FucTs.
Our findings showed that transfection of a1,3-FucT-
VII promoted the functional activity of InR, as verified
by increased tyrosine phosphorylation of InR-b and
IRS-1 (Fig. 3). Moreover, Ser ⁄ Thr phosphorylation of
InR signaling molecules, including PKB (Fig. 4A,B),
MEK, p42 ⁄ 44 MAPK (Fig. 5A,B) and the activity of
PKB (Fig. 4C,D) was stimulated concomitantly.
Expression of some other signaling proteins, such as
PDK-1, PKN (Fig. 4E,F), c-Raf-1 (Fig. 5A,B) and
b-catenin (Fig. 7A,B), was also upregulated by a1,3-
FucT-VII. Elevation of Ser ⁄ Thr phosphorylation in
downstream signaling molecules was presumed to be
mediated by increased tyrosine phosphorylation of InR
and IRS-1; the latter resulting from the increased SLe
x
content of InR-a. This speculation was evidenced by
the following. First, the intensity of Ser ⁄ Thr phos-
phorylation in downstream signaling molecules was
positively correlated with the intensity of tyrosine
phosphorylation in InR and IRS-1, and tyrosine phos-
phorylation was proportional to the SLe
x
content of
InR-a and also the mRNA expression of a-1,3-FucT-

VII in mock, FucTVII-M and FucTVII-H cells. Sec-
ond, inhibition of InR tyrosine autophosphorylation
by HNMPA-(AM)
3
, which inhibits EGFR tyrosine
kinase slightly [28], led to a dramatic reduction in the
Ser ⁄ Thr phosphorylation of PKB and p42 ⁄ 44 MAPK,
and obvious attenuation of the difference in phos-
phorylation intensity among three cell lines with differ-
ent a1,3-FucT-VII expression levels (Fig. 6A). Third,
blockage of cell surface SLe
x
by KM93 also resulted in
significant attenuation of the phosphorylation of PKB
and p42 ⁄ 44 MAPK, as well as the difference in phos-
phorylation intensity among three cell lines (Fig. 6B).
However, in the presence of HNMPA-(AM)
3
and
KM93, some differences in the phosphorylation inten-
sities of PKB and p42⁄ 44 MAPK were still observed
in mock- and a1,3-FucT-VII-transfected cells, indica-
ting that the SLe
x
on InR contributes a large propor-
tion, though not all, of the increased phosphorylation
caused by the overexpression of a1,3-FucT-VII. It
0
001
00

2
003
004
005
006
700
3)MA
(
-AP
M
N
H
ht
iW
3
)M
A(-
A
PMNHtuoht
i
W
Relative intensity
803T-BKP
374S-BKP
KPAM44/24p-p
*
*
*
*
#

#
#
#
#
*
*
0
00
1
002
0
0
3
00
4
005
00
6
007
008
39MK
h
tiW
3
9MKtuohtiW
Relative intensity
p8
03T
-
B

KP
p374S-BK
P
KPAM44/24p-p
*
*
*
*
*
*
#
#
#
#
Mock-1
AB
C
D
FucTVII-M
( ) HNMPA-(AM)3
PKB-T308p
PKB-S473p
p-p42/44 MAPK
β-actin
(+) HNMPA-(AM)3
FucTVII-H
Mock-1
FucTVII-M
FucTVII-H
Mock-2

FucTVII-M
FucTVII-H
Mock-1
FucTVII-M
FucTVII-H
Mock-2
FucTVII-M
FucTVII-H
Mock-2
FucTVII-M FucTVII-H
(+) KM93
( ) KM93
FucTVII-H
Mock-1
FucTVII-M
Mock-2
FucTVII-M
FucTVII-H
PKB-T308p
PKB-S473p
p-p42/44 MAPK
β-actin
Fig. 6. Effect of HNMPA-(AM)
3
and KM93 on the phosphorylation of PKB and p42 ⁄ 44 MAPK. (A) Western immunoblot profiles of phosphor-
ylated PKB and p42 ⁄ 44 MAPK in the absence and presence of HNMPA-(AM)
3
. (B) Densitometric quantification of (A) (n ¼ 3). (C) Western
immunoblot profiles of phosphorylated PKB and p42 ⁄ 44 MAPK in the absence and presence of KM93. (D) Densitometric quantification of
(C) (n ¼ 3). Mock, H7721 cells transfected with pcDNA3.1 vector; FucTVII-M, H7721 cell line with moderate expression of transfected a1,3-

FucT-VII; FucTVII-H, H7721 cell line with high expression of transfected a1,3-FucT-VII; PKB-T308p, phosphorylated PKB at T308; PKB-S473p,
phosphorylated PKB at S473; p-p42 ⁄ 44 MAPK, phosphorylated p42 ⁄ 44 MAPK; Mock-1, mock cells without HNMPA-(AM)
3
or KM93 treat-
ment; Mock-2, mock cells with HNMPA-(AM)
3
or KM93 treatment. (A) and (C) are representative of three reproducible experiments.
*P < 0.01 compared with ‘Mock-1’. #P < 0.05 compared with ‘Mock-2. Western immunoblotting is described in the Experimental proce-
dures. Samples without and with HNMPA-(AM)
3
or KM93 were examined simultaneously on the same electrophoresis gel.
Q. Wang et al. Fucosyltransferase-VII regulates insulin signaling
FEBS Journal 274 (2007) 526–538 ª 2006 The Authors Journal compilation ª 2006 FEBS 533
appears that upregulation of phosphorylation and pro-
tein expression was not mediated by EGFR, because
the SLe
x
content and tyrosine autophosphorylation of
EGFR remained constant following a1,3-FucT-VII
transfection.
In a previous insulin stimulation experiment, it was
found that H7721 cells were prone to die in serum-
free (0%) medium; therefore, 2% fetal bovine serum-
deficient medium was used. The results showed that
phospho-PKB-S473 barely appeared in cells cultured
in the serum-deficient medium (Fig. 8A). As shown
in Fig. 4A, however, there was basal expression of
both phospho-PKB-T308 and phospho-PKB-S473 in
mock- and a1,3-FucT-VII-transfected cells. These
observations suggest that phospho-PKB-T308 and

phospho-PKB-S473 are regulated by different
mechanisms. It has been documented that phospho-
PKB-T308 is regulated by phosphatidyl inositol-3-
kinase ⁄ PDK-1 [25,26], but the signal for PKB-S473
phosphorylation comes from the integrin ⁄ integrin-
linked kinase signaling pathway [32]. Sarbassov et al.
reported that PKB-S473 can be phosphorylated
directly by a kinase, named target of rapamycin
(TOR) kinase and its associated protein rictor,
0
05
0
0
1
051
002
052
003
0
5
3
00
4
0
54
KPAM44/24p-pp374S-BKP
p
803T-BKP
Relative intensity
k

coM
M-7T
F
H-7
T
F
*
*
*
*
*
*
Mock
A
B
FT7M
( ) Insulin (+) Insulin
FT7H
Mock
FT7M
FT7H
PKB-T308p
PKB-S473p
p-p42/44 MAPK
β-actin
Fig. 8. Facilitation of insulin signaling in a1,3-FucT-VII-transfected
cells. (A) Western profiles of phosphorylated PKB and p22 ⁄ 24
MAPK in insulin-untreated and -treated cells cultured in 2% fetal
bovine serum medium. (B) Quantification of phosphorylated PKB
and p22 ⁄ 24 MAPK in the presence of insulin (n ¼ 3). Mock, H7721

cells transfected with pcDNA3.1 vector; FucTVII-M, H7721 cell line
with moderate expression of transfected a1,3-FucT-VII; FucTVII-H,
H7721 cell line with high expression of transfected a1,3-FucT-VII;
PKB-T308p, phosphorylated PKB at T308; PKB-S473p, phosphoryl-
ated PKB at S473; p-p42 ⁄ 44 MAPK, phosphorylated p42 ⁄ 44 MAPK;
FT7M, H7721 cell line with moderate expression of transfected
a1,3-FucT-VII; FT7H, H7721 cell line with high expression of trans-
fected a1,3-FucT-VII. (A) is representative of three reproducible
experiments. *P < 0.01 compared with ‘Mock’. Cell culture, insulin
treatment and western immunoblotting are described in the
Experimental procedures. Samples without and with insulin treat-
ment were examined simultaneously on the same electrophoresis
gel.
0
200
400
600
A
B
C
*
*
0
5
10
15
20
25
30
Luciferase activity

*
*
Mock FucTVII-M FucTVII-H
Mock FucTVII-M FucTVII-H
Mock FucTVII-M FucTVII-H
β-actin
β-Catenin
Relative intensity
Fig. 7. Effects of a1,3-FucT-VII cDNA transfection on the expres-
sion of b-catenin and TCF activity. (A) Western blot profile of
b-catenin. (B) Densitometric quantification of (A) (n ¼ 3). (C) Trans-
activation activity of TCF measured as luciferase activity (n ¼ 3).
Mock, H7721 cells transfected with pcDNA3.1 vector; FucTVII-M,
H7721 cell line with moderate expression of transfected a1,3-FucT-
VII; FucTVII-H, H7721 cell line with high expression of transfected
a1,3-FucT-VII; TCF, T-cell factor (transcription factor). (A) is repre-
sentative of three reproducible experiments. *P < 0.01 compared
with ‘Mock’. Western blotting and luciferase assay are described in
the Experimental procedures.
Fucosyltransferase-VII regulates insulin signaling Q. Wang et al.
534 FEBS Journal 274 (2007) 526–538 ª 2006 The Authors Journal compilation ª 2006 FEBS
because a reduction in rictor or mammalian TOR
(mTOR) expression inhibited the signaling of PKB.
Rictor–mTOR complex can also facilitate the phos-
phorylation of PKB-T308 by PDK-1 [33]. Basal
expression of phospho-PKB-S473 in serum-containing
medium may result from the stimulation of insulin or
growth factors in the 10% fetal bovine serum. Our
finding that the cell response to insulin was correla-
ted with expression of a1,3-FucT-VII and SLe

x
(Fig. 8) reveals that insulin signaling was facilitated
in a1,3-FucT-VII-transfected cells.
Expression of PDK-1, PKN, c-Raf-1 and b-catenin
was upregulated after a1,3-FucT-VII transfection. The
mechanism is not well understood. However, it is rea-
sonable to speculate that upregulation may be caused
by the promotion of InR signaling, because activated
PKB can stimulate the phosphorylation of GSK-3 and
inhibit the activity of GSK-3. GSK-3 phosphorylates
b-catenin and stimulates the ubiquitination and protea-
somal proteolysis of b-catenin. During activation of
the InR pathway, GSK-3 is inactivated, which leads to
the accumulation and nuclear translocation of cyto-
plasmic unphosphorylated b-catenin. In nuclei, b-cate-
nin binds to transcription factor TCF (T-cell factor,
also called LEF, leukocyte enhancer factor) to form a
heterodimer, and transactivates the transcription of
target genes [29–31]. This model of mechanism is sup-
ported by the increased activity of TCF in this study
(Fig. 7C). Recently, we discovered that a1,3-FucT-VII
can upregulate expression of the integrin-a5 subunit at
both the mRNA and protein levels (unpublished). The
latter finding supports the suggestion that transfection
of a1,3-FucT-VII might affect the transcription of
some genes.
It would be better to use antisense a1,3-FucT-VII,
iRNA or a gene-knockout method to suppress
endogenous a1,3-FucT-VII to confirm the above
results. Unfortunately, we found that suppression of

a1,3-FucT-VII expression after transfection of anti-
sense a1,3-FucT-VII cDNA was not apparent,
because parent H7721 cells express a low level of
endogenous a-1,3-FucT-VII. Sometimes antisense
cDNA even led to cell death. When a gene of a1,3-
FucT-VII was knocked-out, almost all cells died
within 24 h. This suggests a1,3FucT-VII is essential
for the survival of H7721 cells. Therefore, construc-
tion of a plasmid containing a mutant at the cata-
lytic domain of a1,3-FucT-VII with deletion of
enzyme activity of its coding protein is very critical
if we are to determine whether the changed phospho-
rylation of signaling molecules was mediated by the
altered amount of SLe
x
on InR. This is being inves-
tigated in our laboratory.
It would be of interest to study whether the SLe
x
level of InR on insulin-responsive cells in diabetic
patients was changed. This may reveal the role of the
sugar chains on InR in the pathogenesis of diabetes.
In summary, the cDNA of a1,3-FucT-VII is able to
regulate the phosphorylation and expression of some
signaling molecules in the InR pathway, and these
effects of a1,3-FucT-VII are probably mediated by its
product, SLe
x
, on the glycans of cell-surface receptors.
Increased expression and phosphorylation of insulin-

signaling molecules leads to the facilitation of insulin
signaling. These findings provide evidence that modifi-
cation of the terminal structure of glycans on surface
receptors may also affect cell signaling. The detailed
mechanism requires further study.
Experimental procedures
Materials
H7721 human hepatocarcinoma cell line was obtained from
the Institute of Cell Biology. RPMI-1640 and liposome
Lipofectamine
TM
were purchased from Gibco ⁄ BRL (Rock-
ville, MD). Rabbit polyclonal antibodies against human
insulin receptor a- and b-subunit, IRS-1, Raf-1, PDK-1,
PKN, b-catenin and mouse mAb 528 (against human
EGFR extracellular domain), b-actin and Protein G plus-
agarose were from Santa Cruz Technology (Santa Cruz,
CA). Rabbit polyclonal antibodies against human PKB,
phospho-PKB, phospho-PKN, MEK1 ⁄ 2, phospho-
MEK1 ⁄ 2, p42 ⁄ 44 MAPK, and mAb against phospho-
p42 ⁄ 44 MAPK were from Cell Signaling Technology
(Beverly, MA). The PKB assay kit was from New England
Biolabs (NEB Ltd., Beijing, China). KM93 was from Sei-
kagaku Co. (Tokyo, Japan). The InR kinase inhibitor,
HNMPA-(AM)
3
, [28] was from Calbiochem (San Diego,
CA). Polyvinylidene difluoride membrane was from Bio-
Rad Laboratories (Hercules, CA). Phosphotyrosine anti-
body (PT66), Mes, Hepes, leupeptin, pepstatin, human

EGFR mAb CF4 (against intracellular domain), fluorescein
isothiocyanate-conjugated and horseradish peroxidase-labe-
led secondary antibodies (goat anti-mouse and anti-rabbit
IgG) were from Sigma (St. Luois, MO). Trizol, AMV
reverse transcriptase, transcription factor TCF analysis kit
(Dual-luciferase
R
reporter assay system) and Renilla lucif-
erase reporter plasmid (pRL-TK) were from Promega
(Madison, WI). The TCF reporter plasmid (TK-luciferase
reporter) was the product of Upstate Biotechnology (Lake
Placid, NY). The RT-PCR primer of a1,3-FucT-VII was
provided by TaKaRa Co. (Tokyo, Japan). Other reagents
were commercially available in China.
a1,3-FucT-VII transfected human hepatocarcinoma
H7721 cell lines were established as previously reported
[15].
Q. Wang et al. Fucosyltransferase-VII regulates insulin signaling
FEBS Journal 274 (2007) 526–538 ª 2006 The Authors Journal compilation ª 2006 FEBS 535
Cell culture and treatment
Cells were cultured at 37 °C, 5%CO
2
in RPMI-1640 con-
taining 10% fetal bovine serum, penicillin and streptomy-
cin as described previously [5,20]. In the treatment of InR
kinase inhibitor and SLe
x
mAb, the final concentration of
HNMPA-(AM)
3

, and KM93 were 50 lm and 30 lgÆmL
)1
,
respectively. The duration of treatment was 24 h. In the
experiments using insulin, the cells were precultured in 2%
fetal bovine serum (serum deficiency) medium for 24 h,
and then 10 nm of insulin was added for 10 min incuba-
tion.
Determination of a1,3 FucT-VII mRNAs with
RT-PCR
Total-cell RNA was extracted with Trizol and the cDNA
was synthesized with oligo(dT)
18
primer and AMV reverse
transcriptase from 3 lg total RNA. The RT-PCR was per-
formed in 50 lL of reaction mixture containing 5 lL
cDNA, 0.2 lm of the primer pair of a1,3FucT-VII or
b-actin (internal standard), 0.2 m d-NTP and 1unit Taq
DNA polymeraseas described previously [34]. The primer
for a1,3FucT-VII was F: 5¢-CACCTCCGAGGCATCTTC
AACTG-3¢,R:5¢-CGTTGGTATCGGCTC TCATTCA
TG-3¢. The primer for b-actin was F: 5¢-GATATCGCC
GCGCTCGTCGTCGAC-3¢,R:5¢-CAGGAAGGAAGG
CTGGAAGAGTGC-3¢ [35]. The cDNA was subjected to
denaturation at 94 °C for 5 min, followed by 28 cycles
(94 °C, 61.5 °C and 72 °C, 1 min for each) of PCR, and
incubated at 72 °C for 10 min and 4 °C for 5 min. Then
10 lL products were applied to agarose gel electrophor-
esis. The amplified DNA bands were scanned and ana-
lyzed with nih image software. The quantitative data

were obtained by the intensity ratios of a1,3FucT-VII ⁄
b-actin band.
Detection of the expression of SLe
x
, InR-a
subunit and EGFR on the cell surface using flow
cytometry
After being washed with NaCl ⁄ P
i
and blocked with 1%
BSA, the EDTA-detected cells (1 · 10
6)
were incubated
with 1 : 50 SLe
x
antibody KM93, 2.5 lgÆmL
)1
polyclonal
antibody against InR-a subunit or 528 mAb against the
EGFR extracellular domain for 45 min at 4 °C. In the
‘(–) Control’ sample the primary antibody was omitted.
Washed cells were incubated with 1 : 128 fluorescein
isothiocyanate-conjugated secondary antibody for 30 min
at 4 °C. cells were then suspended in NaCl ⁄ P
i
and sub-
jected to flow cytometry. Fluorescence-activated cell spec-
tra were drawn automatically, and the relative amount of
surface SLe
x

was expressed as mean fluorescence inten-
sity.
Immunoprecipitation of InR-a,-b, EGFR and IRS-1
Washed monolayer cells were lyzed with 200 lL lysis buffer
(50 mm pH 7.4 Hepes, 150 mm NaCl, 100 mm NaF, 1 mm
MgCl
2
, 1.5 mm EGTA, 1% Nonidet P-40, 1 mm phen-
ykmethylsulfonyl fluoride and 1 mg % of leupeptin and
pepstatin). After protein determination, cell lysate contain-
ing 500 lg protein was incubated with 5 lg of one of the
following antibodies (antibody to InR-a or InR-b, IRS-1,
CF4 antibody against EGFR intracellular domain), and
incubated at 4 °C for 1 h. Protein G plus-agarose was
added and the samples were incubated at 4 °C for 3 h for
immunoprecipitation.
Analysis of SLe
x
expression on InR-a and EGFR
using western immunoblotting
In brief, immunoprecipitated InR-a and EGFR were sub-
jected to SDS ⁄ PAGE, then transferred to a poly(vinylidene
difluoride) membrane and treated with 1 : 500 diluted anti-
SLe
x
(KM93) and anti-InR-a or anti-EGFR (528) sera in
Tris-buffered saline with 5% fat-free dry milk, followed by
1 : 500 HRP-labeled secondary antibody. Finally, the color
was developed with enhanced chemiluminescence reagents,
and followed by densitometric scanning.

Determination of tyrosine phosphorylation
of InR-b, EGFR or IRS-1 using western
immunoblotting
Monolayer cells were lysed with 200 lL lysis buffer as
described above. Immunoprecipitated InR-b or EGFR was
divided into two (for different probes) and subjected to
8% SDS ⁄ PAGE and western blotting, the membranes
were probed with 1 : 1000 phosphotyrosine mAb (PT66)
and InR-b antibody or 1 : 500 EGFR antibody (CF4) in
Tris-buffered saline with 5% fat-free dry milk, followed
by incubation with 1 : 500 diluted horseradish peroxidase-
labeled secondary antibody. The color was also developed
with an enhanced chemiluminescence reagent. The meas-
urement of tyrosine phosphorylation of IRS-1 was similar
to that of InR-b and EGFR, but the primary antibodies
used in western immunoblotting were anti-PT66 and anti-
IRS-1 sera.
Analysis of the proteins or phosphorylated
proteins of PKB, PDK-1, PKN, Raf-1, MEK, p42/44
MAPK and b-catenin using western
immunoblotting
Briefly, cells were homogenized in Mes buffer (0.1 m
pH 6.5, 150 mm NaCl, 2% Triton X-100, 25% glycerol,
1mm phenylmethylsulfonyl fluoride, 1 mg % leupeptin
Fucosyltransferase-VII regulates insulin signaling Q. Wang et al.
536 FEBS Journal 274 (2007) 526–538 ª 2006 The Authors Journal compilation ª 2006 FEBS
and pepstatin), and 50 lg supernatant protein after cen-
trifugation were subjected to 10% SDS ⁄ PAGE. The mem-
branes were treated with one of the 1 : 500-diluted
primary antibodies of the determined proteins or phos-

phorylated proteins in Tris-buffered saline with 5% fat-
free dry milk, followed by incubation with 1 : 500-diluted
horseradish peroxidase-labeled secondary antibody and
stained with enhanced chemiluminescence reagent. b-Actin
was used as loading control and stained with 1 : 800 dilu-
ted mAb and 1 : 500 horseradish peroxidase-labeled sec-
ondary antibody. The protein bands were also quantified
with densitometric analysis.
Assay of PKB activity
PKB activity was performed with assay kit according to
the instruction manual. In short, the washed cells were
lyzed with the buffer provided by the kit. After protein
determination, 500 lg of cell lysate was mixed with 20 lL
suspension of immobilized PKB antibody and incubated
at 4 °C for 3 h with shaking. After centrifugation, the
washed pellet was suspended in 40 lL kinase buffer
(20 mm Tris ⁄ HCl, pH 7.5, 5 mm b-glycerolphosphate,
2mm dithiothreitol, 0.1 mm Na
3
VO
4
,10mm MgCl
2
) and
supplemented with substrates, including 0.8 lLof10mm
ATP and 1 lg GSK-3a ⁄ b fusion protein. After incubated
at 30 °C for 60 min, the phosphorylated GSK-3a ⁄ b fusion
protein was detected with western blotting using the anti-
body to phospho-GSK-3a ⁄ b, followed by the addition of
secondary antibody and enhanced chemiluminescence rea-

gents [5].
Determination of TCF transcription factor activity
with dual luciferase reporter system
The method was performed according to the protocol of
‘dual-luciferase
Ò
reporter assay’ in the manual. In brief,
0.5 lg TCF reporter plasmid and 0.4 lg pRL-TK plasmid
was mixed in 25 lL antibiotic- and serum-free RPMI-1640
(ASF-RPMI), and 2 lL Lipofectamine
TM
was added to
23 lL of the above medium. Two separate preparations
were mixed within 5 min. After standing at room tempera-
ture for 30 min and adding 100 lL ASF-RPMI, the
150 lL plasmid ⁄ Lipofectamine mixture was added to the
ASF-RPMI washed cells already cultured in 12-well plate
for 24 h according the Lipofectamine
TM
manual. Cells
were further cultured at 37 °C, 5% CO
2
for 5 h, and
transferred to 300 lL antibiotic-free RPMI-1640 contain-
ing 20% serum for 24 h incubation. Finally, the cells were
incubated in normal RPMI-1640 for 48 h. After washing
with NaCl ⁄ P
i
, cells were lyzed and luciferase activity was
assayed according to the manual provided with the kit.

The firefly luminescence intensity of TK-luciferase was
normalized against the Renilla luminescence intensity of
pRL-TK [36].
Acknowledgements
This work was supported by the grant from National
Natural Science Foundation of China no. 30670467.
References
1 Rudd PM & Dwek RA (1997) Glycosylation: heteroge-
neity and the 3D structure of proteins. Crit Rev Bio-
chem Mol Biol 32, 1–100.
2 Mann M & Jensen ON (2003) Proteomic analysis of
post-translational modifications. Nat Biotechnol 21,
255–261.
3 Rebbaa A, Yamamoto H, Saito T, Meuillet E, Kim
P, Kersey DS, Bremer EG, Taniguchi N & Moskal
JR (1997) Gene transfection-mediated overexpression
of b1,4-N-acetylglucosamine-bisecting oligosaccharides
in glioma cell line U373 MG inhibits epidermal
growth factor receptor function. J Biol Chem 272,
9275–9279.
4 Ihara Y, Sakamoto Y, Mihara M, Shimizu M & Tani-
guchi N (1997) Overexpression of N-acetylglucosaminyl-
transferase III disrupts the tyrosine phosphorylation of
Trk with resultant signaling dysfunction in PC12 cells
treated with nerve growth factor. J Biol Chem 272,
9629–9634.
5 Guo P, Wang QY, Guo HB, Shen ZH & Chen HL
(2004) N-Acetylglucosaminyl-transferase V modifies the
signaling pathway of epidermal growth factor receptor.
Cell Mol Life Sci 61, 1975–1804.

6 Guo HB, Lee I, Kamar M & Pierce M (2003) N-Acetyl-
glucosaminyltransferase V expression levels regulate
cadherin-associated homotypic cell–cell adhesion and
intracellular signaling pathways. J Biol Chem 278,
52412–52424.
7 Wang XC, Gu JG, Ihara H, Miyoshi E, Honke K &
Taniguchi N (2006) Core fucosylation regulates epider-
mal growth factor receptor-mediated intracellular signal-
ing. J Biol Chem 281, 2572–2572.
8 Taniguchi N (2005) Biological function of core fucosyla-
tion. Glycoconj J 22, 224 (Abstract of GLYCO–XVIII).
9 Kannagi R, Izawa M, Koike T, Miyazaki K &
Kimura N (2004) Carbohydrate-mediated cell adhesion
in cancer metastasis and angiogenesis. Cancer Sci 95,
377–384.
10 Varki A (1994) Selectin ligands. Proc Natl Acad Sci
USA 91, 7390–7397.
11 Kannagi R (2004) Molecular mechanism for cancer-
associated induction of sialyl Lewis X and sialyl Lewis
A expression: the Warburg effect revisited. Glycoconj J
20, 353–364.
12 Narimatsu H (1998) Human fucosyltransferases. Tissue
distribution of blood group antigens, cancer associated
antigens and fucosyltransferase. Protein Nucleic Acid
Enzyme 48, 2394–2403 [in Japanese].
Q. Wang et al. Fucosyltransferase-VII regulates insulin signaling
FEBS Journal 274 (2007) 526–538 ª 2006 The Authors Journal compilation ª 2006 FEBS 537
13 de Vries T, Knegtel RMA, Holmes EH & Macher BA
(2001) Fucosyltransferases: structure ⁄ function studies.
Glycobiology 11, 119–128.

14 Sasaki K, Kurata K, Funayama K, Nagata M, Watana-
be E, Ohta S, Hanat N & Nishi T (1994) Expression
cloning of a novel a1,3-fucosyltransferase that is
involved in biosynthesis of the sialyl Lewis X carbohy-
drate determinants in leukocytes. J Biol Chem 269,
14730–14737.
15 Liu F, Qi HL, Zhang Y, Zhang XY & Chen HL (2001)
Transfection of c-erbB2 ⁄ neu gene up regulates the
expression of sialyl Lewis X, a1,3-fucosyltransferase VII
and metastatic potential in human hepatocarcinoma cell
line. Eur J Biochem 268 , 3501–3512.
16 Liu F, Zhang Y, Zhang XY & Chen HL (2002) Trans-
fection of nm23-H1 gene into human hepatocarcinoma
cell line inhibits the expression of sialyl Lewis X,
a1,3-fucosyltransferase VII and metastatic potential.
J Cancer Res Clin Oncol 128, 189–196.
17 Duan LL, Guo P, Zhang Y & Chen HL (2005) Regula-
tion of metastasis-suppressive gene nm23-H1 on glyco-
syltransferases involved in the synthesis of sialyl Lewis
antigens. J Cell Biochem 94, 1248–1257.
18 Liu F, Qi HL & Chen HL (2001) Regulation of differ-
entiation- and proliferation-inducers on Lewis antigens,
a-fucosyltransferase and metastatic potential in hepato-
carcinoma cells. Br J Cancer 268, 3501–3512.
19 Qi HL, Zhang Y, Ma J, Guo P, Zhang XY & Chen HL
(2003) Insulin ⁄ protein kinase B signaling pathway
upregulates metastasis-related phenotypes and molecules
in H7721 human hepatocarcinoma cell line. Eur J Bio-
chem 270, 3795–3805.
20 Wang QY, Guo P, Duan LL, Shen ZH & Chen HL

(2005) a-1,3-Fucosyltransferase VII stimulates the
growth of hepatocarcinoma cells via cyclin-dependent
kinase inhibitor p27
Kip1
. Cell Mol Life Sci 62, 171–178.
21 Brandt J, Andersen A, S & Kristensen C (2001) Dimeric
fragment of the insulin receptor a-subunit binds insulin
with full holoreceptor affinity. J Biol Chem 276, 12378–
11238.
22 Bishayee S (2000) Role of conformational alteration in
the epidermal growth factor receptor (EGFR) function.
Biochem Pharmacol 60, 1217–1223.
23 Pessin JE & Saltiel AR (2000) Signaling pathway in
insulin action. Molecular targets of insulin resistance.
J Clin Invest 106, 165–169.
24 Polakis P (2000) Wnt signaling and cancer. Gene Dev
14, 1837–1851.
25 Datta SR, Brunet A & Greenberg ME (1999) Cellular
survival: a play in three Akts. Gene Dev 13, 2905–2927.
26 Balendran A, Casamayor A, Deak M, Paterson A,
Gaffney PA, Downes CP & Alessi DR (1999) PDK-1
acquires PDK-2 activity in the presence of a synthetic
peptide derived from the carboxy terminus of PRK-2.
Curr Biol 9, 393–404.
27 Dong LQ, Landa LR, Wick MJ, Zhou L, Mukai H,
Ono Y & Liu F (2000) Phosphorylation of protein
kinase N by phosphoinositide-dependent protein kinase-
1 mediates insulin signals to the actin cytoskeleton. Proc
Natl Acad Sci USA 97, 5089–5094.
28 Saperstein R, Vicario PP, Strout HV, Brady E, Slater

EE, Greenlee WJ, Ondeyka DL, Patchett AA & Hang-
auer DG (1989) Design of a selective insulin receptor
tyrosine kinase inhibitor and its effect on glucose uptake
and metabolism in intact cells. Biochemistry 28, 5694–
5701.
29 Willert K & Nusse R (1998) b-Catenin: a key mediator
of Wnt signaling. Curr Opin Genet Dev 8, 95–102.
30 Nishita M, Hashimoto MK, Ogata S, Laurent MN,
Ueno N, Shibuya H & Cho KW (2000) Interaction
between Wnt and TGF-beta signaling pathways during
formation of Spemann’s organizer. Nature 403, 781–
785.
31 Doble BW & Woodgett JM (2003) GSK-3: tricks of the
trade for a multi-tasking kinase. J Cell Sci 116, 1175–
1186.
32 Shoukat D, Benjamin W & Gregory H (1999) Integrin-
linked kinase (ILK): a regulator of integrin and growth
factor signaling. Trend Cell Biol 9, 319–323.
33 Sarbassov DD, Guertin DA, Ali SM & Sabatin DM
(2005) Phosphorylation and regulation of Akt ⁄ PKB by
the rictor–mTOR complex. Science 307, 1098–1101.
34 Guo P, Zhang Y, Zhao JH, Wang LY, Guo HB, Zhang
XY & Chen HL (2003) Regulation on the expression
and N-glycosylation of integrin by N-acetylglucosami-
nyl- transferase V. Biochem Biophys Res Commun 310,
619–626.
35 Guo P, Zhang Y, Shen ZH, Zhang XY & Chen HL
(2004) Effect of N-acetylglucosaminyltransferase V on
the expressions of other glycosyltransferases. FEBS Lett
562, 93–96.

36 Li L, Yuan HD, Weaver CD, Mao JH, Farr GH, III
Sussman DJ, Jonkers J, Kimelman D & Wu DQ (1999)
Axin and Frat1 interact with Dvl and GSK, bridging
Dvl to GSK in Wnt-mediated regulation of LEF-1.
EMBO J 18
, 4233–4240.
Fucosyltransferase-VII regulates insulin signaling Q. Wang et al.
538 FEBS Journal 274 (2007) 526–538 ª 2006 The Authors Journal compilation ª 2006 FEBS