| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Institut de Pharmacologie et de Biologie Structurale du CNRS, 205 Route de Narbonne, 31077 Toulouse Cedex 4, France; and
* Commissariat à l’Energie atomique, Biochimie des Régulations Cellulaires Endocrines, INSERM U244 CEN/Grenoble, F-38054 Grenoble Cedex 9, France
1Correspondence: Institut de Pharmacologie et de Biologie Structurale du CNRS, 205 Route de Narbonne, 31077 Toulouse Cedex 4, France. E-mail: bouche{at}ipbs.fr
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Key Words: FGF-2 • mitogenicity • signaling pathway • activation • CK2
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, 2)
.
FGFs play an important role in cell growth and differentiation, neurite
outgrowth, embryogenesis, angiogenesis, and wound healing
(3)
. FGF-2 exists as four molecular forms including three
of high molecular mass (21.5, 22, and 24 kDa) initiated by alternative
CUG-codon, and one 18 kDa form initiated at canonical AUG codon
(4
, 5)
. In cells that produce FGF-2, the higher molecular
mass isoforms are found exclusively in the nucleus, whereas the 18 kDa
FGF-2 is cytoplasmic, excreted and stored in the extracellular matrix
(6
, 7)
. The pleiotropic effects in cell growth and
differentiation exhibited by 18 kDa FGF-2 reflect an intricate
combinatorial process involving potential interactions between the
growth factor, four different high-affinity transmembrane tyrosine
kinase receptors (FGF-R1 to R4), and low-affinity binding sites
corresponding to the naturally heterogeneous glycosaminoglycan chains
of heparan sulfate proteoglycans (HSPG) (1
, 8)
. The
functional domain of FGF-2 (18 kDa, 155 amino acid form) that
participates in heparin binding is composed of extended array of
positively charged lysine and arginine residues (9
10
11
12)
.
In addition, Springer et al. identified high- and low-affinity receptor
binding surfaces on FGF-2 on the basis of site directed mutagenesis and
molecular modeling (13)
. Using a molecular modeling
method, Lam et al. (14)
recently proposed a modeled
complex between heparin, FGF-2, and FGF-R1 with a 1: 2: 2 stochiometry
consistent with the reported site-specific mutagenesis and biochemical
cross-linking data.
FGF-2 binding to its receptor causes its autophosphorylation and
activation of at least two signaling pathways, mitogen-activated
protein kinase (MAPK) and phospholipase C-
(PLC), leading to
mitogenic response (15)
. The MAPK pathway is required for
proliferation of fibroblasts in culture (16)
, for
neurotrophic growth factor (NGF)- and epidermal growth factor (EGF)
-induced differentiation of PC12, and for transformation of NIH3T3
(17)
. In contrast, PLC is not involved in FGF-mediated
mitogenesis in L6 myoblast and BaF3 cells or in differentiation of PC12
cells (18
19
20)
.
In addition to these classical signal transduction pathways, some of
the mitogenic signals may be provided by FGFs themselves. Indeed,
exogenous FGF-1 and FGF-2 (18 kDa) are translocated from outside the
cell to the nucleus (21
, 22)
. For FGF-1, this nuclear
localization is necessary for mitogenic activity (23
24
25)
whereas the nuclear translocation and accumulation of FGF-2 in the
nucleolus are associated with cellular proliferation (22
, 26)
. Addition of FGF-2 to nuclei isolated from G0-arrested ABAE
cells induced an increase of phosphorylation of nuclear proteins
(27)
. One of them—nucleolin—an important component of
the nucleolus, is also a major substrate for protein kinase CK2 in
rapidly proliferating tissues (28)
. The direct interaction
of FGF-2 in vitro with CK2 stimulates CK2 activity
(29)
and suggests that CK2 could be a downstream effector
of the mitogenic signaling pathway induced by exogenously added FGF-2.
CK2, which is present in the cytoplasm and nucleus of all
eukaryotic cells, is composed of two catalytic 
and/or ’
subunits and two regulatory ß subunits, and phosphorylates a large
number of proteins (30)
. In yeast, CK2 is essential for
viability (31)
. In mammalian cells, CK2 is required for
cell cycle progression (32
, 33)
and associates with
several growth-related proteins including p53 (34)
,
I
B- (35)
, and c-Mos (36
, 37)
.
To gain new insight into the FGF-2-induced mitogenic signaling pathway, we analyzed the biological properties of FGF-2 mutants. We show that the mitogenic and differentiation activities of FGF-2 can be dissociated by the mutation of serine 117; FGF-2(S117A) is a poor inducer of proliferation, but possess an intact differentiation activity. Signal transduction through high-affinity receptors and nuclear translocation are not affected by this mutation. However, whereas FGF-2(WT)/CK2 complexes are detected in the nucleus and cytoplasm during the G1/S transition, FGF-2(S117A) is not found to be associated with CK2. In addition, we show that in early G1 phase, FGF-2(WT) seems to specifically stimulate CK2 in the nucleus whereas FGF-2(S117A) is inactive. These results suggest that FGF-2 mitogenic signal transduction pathway requires, in addition to signaling through cell surface receptors, activation of the protein kinase CK2 within the nucleus at the early G1 phase.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
. FGF-2(R116A), FGF-2(S117A), FGF-2(R118A),
FGF-2(K119A), FGF-2(K119E), FGF-2(Y120F), and FGF-2(T121A) cDNAs were
prepared by polymerase chain reaction using a primer containing the
mutated base. FGF-2(WT), FGF-2(R116A), FGF-2(S117A),
FGF-2(R118A), FGF-2(K119A), FGF-2(K119E), FGF-2(Y120F), and
FGF-2(T121A) proteins were expressed and purified as described by the
manufacturer’s instructions (Pharmacia Biotech, Brussels, Belgium)
except that they were further purified using a heparin Sepharose CL6-B
column (Pharmacia Biotech). The concentration of NaCl that eluted the
FGF-2(S117A) mutant from heparin (1.3 M) was the same as that which
eluted FGF-2(WT). The GST carrier was removed from fusion proteins by
cleavage with 25 units of thrombin (Pharmacia Biotech) in 500 µl
phosphate-buffered saline (PBS). The nucleolin purification and protein
biotinylation were performed as described previously (29)
.
Biotinylation on cystein residues of FGF-2 does not affect its
biological activity (39)
.
Competition of the interaction between CK2 and
[3H]spermine and biotinylated-FGF-2 binding to CK2
in vitro
Competition assays were based on a previous study
(40)
. Seven to fifteen pmoles of CK2, 20 pmol of tritiated
spermine (1 µCi), and 1 µM of FGF-2 peptides (generously provided
by A. Baird) were incubated at 4°C for 5 min in a final volume of 80
µl, with a final concentration of 10 mM Tris, pH 7.4. Then samples
were loaded onto a Sephadex G-50 column (Pharmacia). After a
centrifugation for 3 min at 2000 x g at 4°C, the
eluent, containing CK2-[3H]spermine and
CK2-FGF-2 or CK2-FGF-2 peptides, was collected.
[3H]spermine bound to CK2 was determined by
scintillation counter.
Biotinylated-FGF-2/CK2 ß subunit reactions were achieved as described
previously (29)
with nuclear extracts prepared from insect
cells overexpressing Drosophila melanogaster ß subunit of
CK2.
Cell culture
NIH3T3 and ABAE (adult bovine aortic endothelial) cells were
propagated in DMEM (Life Technologies, Inc., Grand Island, N.Y.)
containing antibiotics and supplemented with 10% FCS (fetal calf
serum) and 10% CS (calf serum), respectively.
[3H]Thymidine incorporation was measured as
described (24
, 26)
. BHK-21 (baby hamster kidney clone 21)
cells were grown in DMEM-F12 medium supplemented with 5% FCS and
antibiotics. The rat bladder carcinoma NBT-II and transfected FGF-R1
NBT-II (NBT-II.R1) cells obtained from Dr. J. Jouanneau (Laboratoire de
Physiopathologie du Développement, C.N.R.S.-Ecole Normale
Supérieure, Paris, France) and PC12 cells were cultured in DMEM
supplemented with antibiotics and 10% FCS. Emfi
fgf-2 -/- or Emfi fgf-2
+/+ cells, a generous gift of Dr. T. Doetschman
(University of Cincinnati College of Medicine, Cincinnati, Ohio)
(41)
, were propagated as described above for NIH3T3.
A dose response experiment allowed us to determine the optimal concentration of FGF-2(WT) and FGF-2(S117A) for PC12 and NBTII or NBT-II.R1 cell differentiation. PC12 cells grown to low confluency in 24-well plates in standard medium were stimulated with 10 ng/ml FGF-2(WT) or FGF-2(S117A) for 3 days. NBT-II or NBT-II.R1 cells grown to low confluency in 24-well plates in standard medium were activated for 3 days with 5 ng/ml FGF-2(WT) or FGF-2(S117A).
Cross-linking and binding of 125I-labeled FGF-2(WT) or
FGF-2(S117A) to BHK-21 cells
Purified proteins were radioiodinated using chloramine
T-immobilized on nonporous polystyrene beads (IODO-BEADS from Pierce
Chemical Co., Rockford, Ill.) according to the manufacturer’s
instructions. Protein-bound and free iodide were separated by
centrifugation with microconcentrators 10, with a cutoff of 10 kDa
(Amicon, Beverly, Mass.). The specific activity of the iodinated
proteins was routinely ~5 x 105 cpm/ng of
protein. The cross-linking and binding were performed as described
previously (42)
. Cross-linking products were analyzed by
Western blotting using an anti-streptavidin peroxidase antibody
(Amersham, Amersham, U.K.).
Subcellular localization of FGF-2 by immunofluorescence microscopy
The immunofluorescence microscopy was achieved as described
previously (26)
with some modifications. NIH3T3 cells were
grown on glass coverslips, fixed with 3% paraformaldehyde in PBS for
15 min at 4°C, and washed with PBS-0.5% bovine serum albumin (BSA)
(2x5 min) and then with 50 mM NH4Cl in PBS at
4°C for 20 min. The fixed cell monolayers were permeabilized with
0.1% Triton X-100 in PBS for 5 min at 20°C, washed twice with PBS,
and twice with PBS-0.5% BSA at 4°C. Treated coverslips were
incubated at 37°C for 1 h with the anti-human FGF-2 monoclonal
antibody (Transduction Laboratories, Lexington, Ky.) diluted 1/30 in
PBS-0.5% BSA. After three washes with PBS-0.5% BSA, cells were
further incubated for 1 h at 37°C with fluorescein-conjugated
goat anti-mouse IgG (Immunotech) diluted 1/40 in PBS-0.5% BSA.
Finally, the coverslips were washed with PBS-0.5% BSA (2x5 min) and
with PBS alone (2x5 min), then air dried and mounted in Mowiol.
Fluorescence of fixed immunostained cells was viewed with a Zeiss
confocal laser scanning using a 63x oil immersion objective.
Purification and subcellular localization of biotinylated
FGF-2/CK2 complexes
NIH3T3 or PC12 cells were seeded in 100 mm dishes at a density
of 105 cells/dish. After 72 h, NIH3T3 cells
were G0-arrested by serum starvation for 1 day. Cells were then either
untreated or stimulated with biotinylated FGF-2(WT) or biotinylated
FGF-2(S117A) for different times. Conditioned culture medium was
recovered and the cells were washed twice in PBS containing mixture of
protease inhibitors (Boehringer Mannheim, Mannheim, Germany). After
trypsinization, the cells were harvested in 5 ml of DMEM-10% FCS,
centrifuged for 5 min at 600 x g at 4°C, and washed
twice in PBS. The cells were disrupted for 1–2 min by gentle
homogenization in 1 ml of lysis buffer containing 15 mM Tris, pH 7.5,
150 mM NaCl, 5 mM MgCl2, 0.1% Tween 20, 0.3%
NP10, and a mixture of protease inhibitors. Nuclear and postnuclear
fractions were prepared as described previously (43)
.
Conditioned culture medium, total extract, or nuclear and postnuclear
extracts were incubated in the presence of streptavidin beads for 60
min at 4°C under constant rotation. After extensive washing, bound
proteins were directly resuspended in 1x Laemmli sample buffer and
analyzed by Western blotting using anti-FGF-2 (Sigma, St. Louis, Mo.),
anti-CK2 (UBI), and anti-CK2 ß (Calbiochem, Nottingham, U.K.)
antibodies.
Nucleolin and tyrosine phosphorylation assays; MAPK activation
Phosphorylation assays were achieved as described previously
(29)
; nuclear or cytoplasmic extracts were prepared from
NIH3T3 cells stimulated with FCS, FGF-2(WT), or FGF-2(S117A). All
reactions were stopped by adding 6 µl TCA. The resulting
TCA-insoluble material was mixed with 20 µl of 1x Laemmli sample
buffer. Phosphorylation levels were normalized relative to the quantity
of catalytic subunit of CK2 in nuclear and cytoplasmic extracts by
performing immunoblotting experiments with an anti-CK2 antibody
(data not shown).
For tyrosine phosphorylation and MAPK activation detection, NIH3T3, Emfi fgf-2 -/-, or Emfi fgf-2+/+ cells growing in 100 mm dishes were kept for 1 day in medium containing 0% FCS. Subsequently, cells were untreated or stimulated with 20 ng/ml FGF-2(WT) or FGF-2(S117A) for the times indicated. The cells were washed twice in PBS containing 100 µM sodium orthovanadate, 30 mM sodium pyrophosphate, 50 mM NaF as phosphatase inhibitors, and a mixture of protease inhibitors and lysed in 15 mM Tris, pH 8, 150 mM NaCl, 2 mM MgCl2, 0.5% Tween 20, 1% Triton X-100 supplemented with protease and phosphatase inhibitors, as described above. After centrifugation, supernatants were trichloroacetic acid precipitated and resuspended in 1x Laemmli sample buffer. Tyrosine phosphorylation and activation of endogenous extracellular signal-regulated kinase (ERK) were measured by Western blotting using an antiphosphotyrosine monoclonal antibody (UBI) and an anti-active ERK antibody (Promega, Southampton, U.K.), respectively. Total endogenous ERK was detected with an anti-ERK antibody (New England Biolabs, Beverly, Mass.).
p70 S6 kinase activity assay
The p70S6K activity assay was performed as described previously
(44)
with some modifications. G0-arrested NIH3T3 cells
cultured on 100 mm dishes were either unstimulated or stimulated with
20 ng/ml FGF-2(WT) or FGF-2(S117A) for 60 min. Cells were washed with
extraction buffer (EB) containing in 50 mM Tris, pH 8, 120 mM NaCl, 20
mM NaF, 30 mM 4-nitrophenyl phosphate, 0.1 mM PMSF, 1 mM EDTA, 6 mM
EGTA, 15 mM
Na4P2O7.
After washing, cells were lysed in EB containing 1% Nonidet P-40 on
ice. Collected cell lysate was frozen in liquid nitrogen and kept at
-80°C until assay. After thawing, the lysate was centrifuged at
10,000 x g for 30 min and then 100 µg of lysed
protein was incubated with 7.5 µg of an anti-p70S6K
monoclonal antibody (Calbiochem) at 4°C for 3 h, followed by
precipitation with 20 µl of 50% (wet v/v) protein A-Sepharose beads
for 1 h. The protein A immune complex was collected by
centrifugation, washed three times with EB containing Nonidet P-40,
once with dilution buffer (DB: 50 mM MOPS, pH 7.2, 5 mM
MgCl2, 1 mM DTT), and resuspended in 10 µl of
DB. The p70S6K activity was measured with a S6 Kinase assay kit (UBI)
according to the manufacturer’s instructions. Quantities of
immunoprecipitated p70S6K were verified by performing immunoblotting
experiments with an anti-p70S6K antibody.
In vivo FGF phosphorylation
The in vivo phosphorylation assay was performed as
described previously (45)
with some modifications.
G0-arrested NIH3T3 were preincubated at 37°C for 3 h in
phosphate-free DMEM containing 100 µCi/ml
32PO43-.
Biotinylated FGF-2(WT) or FGF-2(S117A) (20 ng/ml) was added and the
incubation was continued for 12 h. The cells were washed twice in
PBS and lysed in 15 mM Tris, pH 7.5, 150 mM NaCl, 5 mM
MgCl2, 0.1% Tween 20, 0.3% NP10 supplemented
with phosphatase and protease inhibitors. Nuclear and postnuclear
fractions were prepared as described previously (43)
and
incubated in the presence of streptavidin beads for 60 min at 4°C
under constant rotation. After extensive washing, bound proteins were
directly resuspended in 1x Laemmli sample buffer and analyzed by
sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE)
and autoradiography.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
, a natural substrate of this enzyme
involved in ribosome biogenesis (46
, 47)
. CK2 is
stimulated in vitro by polycation structures, including
polyamines such as spermine (30
, 48
, 49)
. CK2 interacts
with spermine through its regulatory ß subunit (49)
.
Using FGF-2 and synthetic peptides spanning the entire sequence of
FGF-2, we have performed competition of the interaction between CK2 and
[3H]spermine. We showed that FGF-2 displaced
[3H]spermine and that the 115–129 amino acid
sequence of FGF-2 was critical for the interaction with CK2 (Fig. 1A
). To identify the residues on FGF-2 important for both CK2
binding and FGF-2 biological properties, several mutations within this
region 115–129 were performed (Fig. 1B
). All these mutants
were first tested for the binding to CK2 in vitro and then
for mitogenic and differentiation activities. Only one showed
interesting characteristics and is described in this paper. Added to
baculovirus-infected cell extracts overexpressing ß subunits of CK2,
FGF-2(S117A) (in which serine 117 was substituted by alanine) had an
impaired ability to interact with the ß regulatory subunit of CK2
(Fig. 1C
) and to stimulate CK2 activity using nucleolin as
substrate (data not shown) compared to wild-type FGF-2.
|
Moreover, this mutant had a greatly reduced mitogenic activity (defined
as the ability of a cell to undergo S phase) compared to FGF-2(WT) in
NIH3T3 and ABAE cells. The maximal mitogenic activity of wild-type
FGF-2 is observed at 10 ng/ml and 1 ng/ml for NIH3T3 and ABAE cells,
respectively (Fig. 2A
). FGF-2(S117A) was unable to promote mitogenicity at these
concentrations and presented a mitogenic activity reduced by 80%
compared to that of wild-type growth factor at 20 ng/ml for NIH3T3
cells. We then tested the differentiation properties of FGF-2(S117A)
(Fig. 2B
). First, neuronal differentiation in the PC12
phaeochromocytoma cell line by FGF-2(WT) and FGF-2(S117A) was analyzed
(50)
. Similar to FGF-2(WT), FGF-2(S117A) promoted neurite
outgrowth in PC12 cells at the same concentration (Fig. 2B
,
left panel). Then the induction of the epithelial-mesenchymal
transition (EMT) in untransfected or FGF-R1 transfected bladder
carcinoma cells (NBTII and NBTII.R1, respectively) was examined (Fig. 2B
, medium and right panels). NBT-II and NBT-II.R1 formed
epithelial cell clusters in standard culture medium (CT). As expected,
FGF-2(WT) was able to induce the EMT in NBTII.R1 but not in
untransfected NBTII cells (51)
. Similar to wild-type
FGF-2, the FGF-2(S117A) mutant caused NBTII.R1 to lose their epithelial
character and to acquire the mesenchymal phenotype leading to
cell dissociation. Together, these results demonstrate that the
mitogenic and differentiation activities of FGF-2 can be uncoupled by a
single point mutation (S117A) that affects the binding to CK2 in
vitro.
|
FGF-2(S117A) binds to and activates FGF receptors
FGF-mediated activities require binding of the growth factor to
its high-affinity tyrosine kinase receptors. Binding properties of
wild-type and mutated FGF-2 with their receptors were compared through
cross-linking experiments (Fig. 3A
, Insert). We revealed a complex of approximately
Mr 140,000 corresponding to those of
FGF-2 (18 kDa) and FGF-R (125 kDa). Therefore, both FGF-2(WT) and
FGF-2(S117A) bound to FGF receptors. To test whether FGF-2(S117A)
competed with FGF-2(WT) for similar binding sites, the ability of
increasing concentrations of FGF-2(WT) and FGF-2(S117A) to displace
125I-FGF-2(WT) or
125I-FGF-2(S117A) bound to cell surface receptors
was analyzed. As shown Fig. 3A
, both FGF-2(WT) and
FGF-2(S117A) displaced wild-type or mutated FGF-2 with the same
efficiency. This shows that both FGF-2(WT) and FGF-2(S117A) can compete
for similar binding sites with equivalent affinities. To study whether
the mutant retained the ability to activate FGF receptors, we tested
its effect on tyrosine phosphorylation of cellular proteins in NIH3T3
cells (Fig. 3B
). As shown in Fig. 3B
, the pattern
of total protein tyrosine phosphorylation after treatment with mutant
FGF-2 for various time periods was similar to that obtained after
wild-type growth factor stimulation. Both wild-type and mutated FGF-2
induced transient (up to 30 min) tyrosine phosphorylation of cellular
proteins. This demonstrates that FGF-2(S117A) retains the ability to
interact with and stimulate the FGF receptors in the cells.
|
FGF-2-binding to cell surface receptors has been shown to initiate at
least two signaling pathways leading to mitogenic response through
activation of MAPK and PLC (15)
. By immunoblotting
experiments with antibodies raised against phosphorylated forms of
MAPK, we observed, after a short (up to 1 h) or a long (from
1 h to 12 h) stimulation, that FGF-2(S117A) mutant was able
to activate MAPK as FGF-2(WT) in NIH3T3 cells (Fig. 4A
). No activation of PLC was detected in FGF-2(WT) and
FGF-2(S117A)-stimulated NIH3T3 cells, suggesting that this enzyme is
not involved in FGF-2-induced mitogenic signal transduction pathway in
this cell line (data not shown). In addition, it was shown that p70 S6
kinase, a serine/threonine kinase that phosphorylates 40 S ribosomal
protein S6 in response to a number of extracellular stimuli
(44)
, is required for FGF-stimulated cell proliferation
but not for differentiation of endothelial cells (44)
.
p70S6K activation was examined by immunoprecipitation of p70S6K from
unstimulated and wild-type or mutated FGF-2-stimulated NIH3T3 cells,
followed by incubation in kinase buffer and
[-32P]ATP in the presence of synthetic
peptides, serving as a substrate for immunoprecipitated p70S6K. Figure 4B
shows that p70S6K is activated in a similar manner in
FGF-2(WT) and FGF-2(S117A)-stimulated cells. These results indicate
that FGF-2(S117A) triggers the activation of the main signaling pathway
involved in mitogenic response, MAPK and p70S6K.
|
The low levels of FGF-2 synthesized in NIH3T3 cells are excreted
(52
, 53)
and accumulated in the extracellular matrix
(6
, 7)
. It is therefore possible that FGF-2(S117A) could
indirectly induce the activation of MAPK through the release of FGF-2
stored in extracellular matrix. To exclude this possibility, we
evaluated the effects of FGF-2(S117A) in fgf-2 gene-ablated
embryonic fibroblasts (41)
(Emfi fgf-2
-/- cells). In these cells, the patterns of
FGF-2(WT) and FGF-2(S117A)-induced MAPK activation kinetics were the
same (Fig. 4C
). In Emfi fgf-2
-/- cells, the deactivation of MAPK occurred
later than in NIH3T3 due to a longer cell cycle (K. Baily, unpublished
observations). These data demonstrate that although FGF-2(S117A) binds
to and activates FGF receptors and stimulates MAPK and p70S6K pathways
similar to wild-type FGF-2, it is a poor inducer of proliferation.
FGF-2(S117A) is translocated to the nucleus of cells
undergoing G1-S transition
Exogenously added FGF-2 translocates to the nucleus of
proliferating cells (22
, 26
, 54
, 55)
. We first verified
that the stability of FGF-2(S117A) in conditioned culture medium was
similar to that of FGF-2(WT) (Fig. 5A
) and, by an in vitro assay, that the FGF-2
mutant dimerized as strongly as wild-type growth factor (data not
shown). Localization of FGF-2(S117A) was examined by indirect
immunofluorescence analysis using an anti-FGF-2 monoclonal antibody in
NIH3T3 cells (Fig. 5B
). Bright staining was observed in the
nucleus after stimulation of G0-arrested cells by both wild-type and
mutated growth factor, with a diffuse staining of the cytoplasm (Fig. 5B, B and D
). In control experiments, no staining was
observed when unstimulated cells were exposed to the anti-FGF-2
antibody (Fig. 5B, A
). Confocal sectioning revealed that
both wild-type and mutated FGF-2 exhibited an intranuclear localization
(Fig. 5B, C and E
). These results indicate that exogenously
added FGF-2(S117A) retains its ability of being internalized from
outside the cell to the cytoplasm and the nucleus. Moreover, these data
show that the inability of FGF-2(S117A) to induce S phase
transition is not due to a defect in internalization or transport to
the nucleus, but rather probably reflects the inability of the mutated
growth factor to interact with the appropriate intracellular targets.
|
FGF-2(S117A) is phosphorylated in living cells
Vilgrain et al. (56)
have reported the
phosphorylation of FGF-2 by kinase distinct from protein kinas A and
protein kinase C in the nuclei of human hepatoma cells, suggesting a
potential role for this phosphorylation in the regulation of cell
growth. The S117A mutation disrupts a consensus phosphorylation site
for protein kinase C (, ß, ) [(S/T)X(R/K)] in human FGF-2
(57)
. Therefore, we compared phosphorylation of
FGF-2(S117A) and wild-type FGF-2 in the cytoplasm and nucleus of living
cells. NIH3T3 cells were incubated with unlabeled biotinylated
FGF-2(WT) and FGF-2(S117A) in the presence of
32PO43-. The
mutated FGF-2 was labeled in both the nucleus and cytoplasm similar to
wild-type growth factor (Fig. 6
). These results rule out the possibility that a defect in
phosphorylation of the growth factor in the cells could be the reason
for reduced mitogenicity of FGF-2(S117A). Rather, the data suggest that
an intracellular function of FGF-2 is affected by the mutation,
probably the activation of a new downstream effector.
|
FGF-2(S117A) neither binds to nor activates nuclear CK2 in
vivo
To test our hypothesis that CK2 could be one of the
intracellular targets of exogenous FGF-2, we first isolated CK2/FGF-2
complexes in biotinylated FGF-2(WT) [biot-FGF-2(WT)] or
biotinylated FGF-2(S117A) [biot-FGF-2(S117A)] stimulated
NIH3T3 cells. As expected, biot-FGF-2(S117A) had a severely
reduced capacity to promote G1-S transition in NIH3T3 cells (Fig. 7A
). Four hours after the addition of
biot-FGF-2(WT), FGF-2(WT)/CK2 complexes were detected in
cytoplasmic and nuclear fractions (Fig. 7B
, lanes 2 and 7).
The number of these complexes increased after 12 h of stimulation,
which corresponded to the G1/S phase transition of NIH3T3 cell cycle
(Fig. 7B
, lanes 4 and 9). In contrast, FGF-2(S117A)/CK2
complexes were poorly detected in cytoplasmic and nuclear fractions
(Fig. 7B
, lanes 3, 5, 8, and 10). FGF-2/CK2 complexes were
not detected during the S phase (data not shown). FGF-2(S117A) is
poorly detected in the nuclear fraction at the beginning of the S phase
of the cell cycle (Fig. 7B
, upper panel, lane 10). Since
there was no concomitant increase in FGF-2(S117A) content in the
cytoplasm (Fig. 7B
, upper panel, lane 5), the loss of this
mutant in the nucleus was likely due to its nuclear degradation. To
support the idea that FGF-2 binding to CK2 is required for cell
proliferation, we have looked for the presence of FGF-2/CK2 complexes
in PC12 cells. Since these cells differentiate rather than proliferate
in response to FGF-2 treatment, we would expect to see a lack of
interaction between FGF-2 and CK2 in these cells. As shown Fig. 7C
, we did not observe any complexes between FGF-2 and CK2
in total cellular extracts from FGF-2-stimulated PC12 cells. Taken
together, these results suggest that this complex formation is required
for FGF-2-stimulated cell proliferation.
|
The activity of CK2 in the cytoplasmic and nuclear fractions from
FGF-2-treated cells was then determined. The appearance of
FGF-2(WT)/CK2 nuclear complexes was correlated with the stimulation of
nuclear CK2 activity (Fig. 8A
). However, the peak of CK2 activity did not correspond to
the maximal amount of detected FGF-2/CK2 complexes. Surprisingly, the
stimulation of CK2 activity was never observed with cytoplasmic
fraction from FGF-2(WT)-stimulated cells. No activation of CK2 in the
cytoplasmic and nuclear fractions from FGF-2(S117A)-treated cells was
detected. To further study the specificity of CK2 activation by FGF-2,
we examined the effect of serum (FCS) on the modulation of CK2 activity
(Fig. 8B
). Serum is known to mimic the effects of diverse
growth factors. We found that CK2 activity did not increase in nuclear
and cytoplasmic extracts from serum-stimulated cells. Thus, in our
experimental conditions, FGF-2 specifically activates nuclear CK2 in
early G1 phase.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). However,
like wild-type FGF-2, this mutant promotes neurite outgrowth in rat
PC12 cells and induces epithelial-mesenchymal transition in
FGF-R1-transfected bladder carcinoma cells (NBTII.R1 cells) (Fig. 2B
). Neither FGF-2(WT) nor FGF-2(S117A) is able to induce
this transition in untransfected cells (NBTII cells), demonstrating
that FGF-2(S117A) recognizes the FGF-R1 and induces signaling pathways
implicated in the differentiation of NBTII.R1 cells similar to
wild-type FGF-2.
FGF-2 exerts its effects in cell growth and differentiation through a
dual receptor system in which it is suggested that the low-affinity
HSPG molecules may act to deliver FGF to high-affinity tyrosine kinase
receptors (8
, 58)
. The identification of FGF-2 functional
domains that participate in binding to heparin and tyrosine kinase
receptors (14)
implies that the S117 residue is not
involved in these interactions. Indeed, our results demonstrate that
the mutation S117A does not affect the binding properties of FGF-2 to
high-affinity receptors (Fig. 3A
) and heparin (data not
shown). The interaction of FGF-2 with tyrosine kinase receptors leads
to tyrosine phosphorylation of several molecules and to the activation
of at least three signaling pathways described to be involved in the
mitogenic signal transmission of FGF-2: PLC, p70S6K, and MAPK
(15
, 44)
. In agreement with previous studies using
different cell types (18
, 19)
, we show that PLC
signaling pathway activation is not involved in mitogenic signal
transduction induced by FGF-2 in NIH3T3 cells (data not shown).
Although FGF-2(S117A) induced protein tyrosine phosphorylation (Fig. 3B
) and activated the MAPK and p70S6K pathway in NIH3T3
cells as did wild-type FGF-2 (Fig. 4A, B
), this mutant was a
poor inducer of proliferation. Thus, similar to FGF-1 (23
, 24
, 59)
, our results demonstrate that activation of tyrosine kinase
of FGF receptors, particularly MAPK stimulation, is insufficient to
program the cells to enter S phase of the cell cycle. In addition, we
show that FGF-2-induced p70S6K activation is insufficient to stimulate
cell proliferation too.
It has been shown that FGF-1 and 2 are translocated to the
nucleus of cells undergoing G1-S transition (23
, 24
, 26)
.
Several authors suggest that signaling events in addition to the
activation of FGF-R are necessary to mediate cellular responses by FGFs
(24
, 59
, 60)
. Wiedlocha et al. (54
, 55
, 61
62
63)
propose a dual mode of signal transduction by externally
added FGF-1 involving the stimulation of tyrosine kinase activity of
the receptors and nuclear translocation of FGF-1, perhaps via the
high-affinity receptors. Transport of FGF-1 to the nucleus is essential
for DNA synthesis (23
24
25)
. It is therefore possible that
the translocation of FGF-2 to the nucleus could be relevant to its
mitogenic activity. The defect in FGF-2(S117A) mitogenic activity is
not due to its inability to be translocated to the nucleus (Fig. 5B
). However, we observed that the mutated growth factor is
not seen in the nucleus at 12 h (Fig. 7B
, lane 10).
This suggests that it is degraded before the G1-S transition. We can
speculate that this degradation is a defect in an interaction with
appropriate nuclear targets or in its phosphorylation in the nucleus.
It has been proposed that phosphorylation of FGF-1 might be important
for its mitogenic activity (45)
. Since the S117A mutation
does not affect the phosphorylation of FGF-2 (Fig. 6)
, the lack of
interaction with appropriate nuclear target is likely to be responsible
for the degradation of the mutated growth factor.
We previously reported that the addition of FGF-2 to isolated nuclei
from G0-arrested cells stimulates RNA polymerase I activity and induces
the phosphorylation of nucleolin, which is a substrate of CK2 and plays
a pivotal role in preribosomal RNA metabolism (22
, 47)
. In
line with these data, several authors described that CK2 or, more
specifically, its ß subunit, was located mainly in the nucleus of
growing cells but spread throughout the cells that are G0-arrested or
confluent (64
65
66)
. Therefore, a coordinated
spatio-temporal distribution of FGF-2 and CK2 in the nucleus could
allow the regulation of specific nuclear protein phosphorylations, such
as nucleolin phosphorylation, during the G1-S transition.
Recently, we have demonstrated that, in vitro, FGF-2
directly stimulates CK2 activity by interacting with its regulatory ß
subunit. FGF-2 seems to specifically activate this enzyme, since FGF-1,
insulin, NGF, and EGF do not directly modulate CK2 activity
(29)
. Here we show that complexes between FGF-2(WT) and
CK2 can be recovered in vivo from the cytosol and the
nucleus of cells undergoing G1-S transition (Fig. 7B
).
FGF-2/CK2 complexes were not detected during the S phase (data not
shown). We observed the stimulation of CK2 by FGF-2 only in the nuclear
fraction, and this stimulation reached a maximum in early
G1 phase (4 h after FGF-2 stimulation) (Fig. 8A
). This maximal stimulation of CK2 activity did not
correlate with the maximal quantity of complexes, suggesting that a
process of CK2 desensitization exists for quenching signal
transduction. Altogether, these data suggest that although
FGF-2(WT)/CK2 complexes can be found both in the cytoplasm and in the
nucleus, only the nuclear ones could be functionally active. Moreover,
the stimulation of nuclear CK2 activity in early
G1 phase was not observed with serum (Fig. 8B
), suggesting that the activation of this kinase was
specific to FGF-2. Our FGF-2(S117A) mutant impaired in its mitogenic
activity did not interact in vivo with CK2. In addition,
FGF-2 does not interact with CK2 in PC12 cells in which the growth
factor induces a differentiation response rather than a mitogenic one.
Taken together, our results raise the possibility that CK2 may be an
essential element in nuclear signaling pathway involved in
FGF-2-induced mitogenic signal in early G1
phase.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
) or FGF-2(S117A) (
).
[3H]Thymidine was added in NIH3T3 and ABAE cultures for
the last 8 h and 4 h, respectively, then incorporated
radioactivity was measured. B) PC12 cells were
unstimulated (CT) or stimulated with 10 ng/ml FGF-2(WT) and
FGF-2(S117A). Neurite outgrowth was observed 3 days after growth factor
treatment. Low confluency NBT-II or FGF-R1-transfected NBT-II
(NBT-II.R1) was unstimulated (CT) or stimulated with 5 ng/ml FGF-2(WT)
and FGF-2(S117A). Cell dissociation reflecting the
epithelial-mesenchymal transition was observed 3 days after growth
factor treatment.
), unlabeled FGF-2(S117A) (
) or 20 ng/ml
125I-FGF-2(S117A) and increasing concentrations of
unlabeled FGF-2(WT) (•), unlabeled FGF-2(S117A) (
). Insert)
Confluent BHK-21 cells were incubated for 3 h at 4°C without
(lane 1) or with biotinylated FGF-2(WT) (biot-FGF-2(WT),
lane 2) and biotinylated FGF-2(S117A)
(biot-FGF-2(S117A), lane 3). At the end of the
incubation, the cells were washed and treated with DSS cross-linking
agent. Cross-linking products were detected by Western blotting using
an anti-streptavidin peroxidase antibody. Migration of molecular mass
standards (kDa) is indicated to the right. B)
G0-arrested NIH3T3 cells were untreated or stimulated for various time
periods with wild-type or mutated FGF-2. Subsequently, the cells were
lysed and the proteins from the postnuclear supernatant were resolved
by SDS-PAGE. Tyrosine-phosphorylated proteins were detected by Western
blotting using antiphosphotyrosine monoclonal antibody. The arrows at
left indicate the major proteins phosphorylated in response to FGF-2.
) represent cytoplasmic CK2 activity in FGF-2(WT)- and
FGF-2(S117A)-stimulated cells, respectively. B) Serum
does not activate nucleolin phosphorylation by CK2. NIH3T3 cells were
treated for different times without or with 20 ng/ml growth factor or
20% FCS (fetal calf serum). Experiments were carried out and analyzed
as described in panel A. (