| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Original Articles: Hormones and Reproductive Health |
5ß1 Integrin via Mitogen-Activated Protein Kinase Pathway1
Departments of Anatomy and Cell Biology (L.M.G., T.M., P.K.L.) and Pathology (C.C.), The University of Western Ontario, London, Ontario, Canada N6C 5C1
Address all correspondence and requests for reprints to: Peeyush K. Lala, M.D., Department of Anatomy and Cell Biology, Medical Science Building, The University of Western Ontario, London, Ontario, Canada N6A 5C1. E-mail: pklala{at}julian.uwo.ca
Abstract
A highly migratory subpopulation of the human placental trophoblast,
known as the extravillous trophoblast (EVT), invades the uterus and its
vasculature, to establish adequate exchange of key molecules between
the maternal and fetal circulations. During their formation, EVT cells
selectively acquire 
5ß1 integrin. We had shown that 5ß1 is
required for their migratory function, and that EVT cell migration is
stimulated by insulin-like growth factor-binding protein (IGFBP)-1
produced by the uterine decidua. The present study examined whether
this stimulation is dependent on binding of the Arg-Gly-Asp (RGD)
domain of IGFBP-1 to an RGD binding site on the 5ß1 integrin,
followed by activation of focal adhesion kinase (FAK) and stimulation
of the mitogen-activated protein kinase (MAPK) pathway. IGFBP-1
treatment increased migration of EVT cells, whereas an anti-5ß1
integrin antibody blocked migration regardless of IGFBP-1 treatment.
Migration stimulation by IGFBP-1 was abrogated by pretreatment with a
Gly-Arg-Gly-Asp-Ser-Pro (GRGDSP), but not a
Gly-Arg-Gly-Glu-Ser-Pro (GRGESP) hexapeptide, and by mutation of the
RGD domain of IGFBP-1 to Trp-Gly-Asp (WGD). IGFBP-1 treatment caused a
rapid localization of immunoreactive FAK to cellular lamellipodia, a
rapid increase in phosphorylation of FAK and extracellular-signal
regulated kinases 1 and 2. Preincubation of EVT cells with Herbimycin
A, a tyrosine kinase inhibitor, abrogated IGFBP-1 effects; whereas an
MAPK kinase inhibitor, PD 98059, reduced migration regardless of
IGFBP-1 treatment. These results indicate that IGFBP-1 stimulation of
EVT cell migration occurs by binding of its RGD domain to the 5ß1
integrin, leading to activation of FAK and stimulation of MAPK pathway.
THE HUMAN PLACENTA is a highly
invasive structure, in which a migratory subpopulation of the placental
trophoblast cells, known as the extravillous trophoblast (EVT),
invades the uterus and remodels its vasculature to establish an
adequate exchange of key molecules between the maternal and the fetal
circulations. EVT cells arise by proliferation and differentiation of
cytotrophoblast stem cells at certain sites within the chorionic villi,
resulting in migratory cell columns, which eventually anchor the
placenta to the uterine wall (1, 2). EVT cell migration
and invasion into the uterus remains largely confined to the
endometrium-myometrium junction and continues until
midgestation. Studies directed at elucidating mechanisms that
regulate EVT cell proliferation, migration, and invasion have
identified that this regulation is provided by a variety of factors in
the EVT cell microenvironment, including growth factors, growth factor
binding proteins, proteoglycans, and extracellular matrix (ECM)
components (3, 4). During the differentiation of
cytotrophoblast stem cells into the EVT cell pathway, they acquire a
selective integrin (ECM receptor) profile; the cells lose 6ß4 and
gain 5ß1 integrins (5, 6), implicating 5ß1 in
EVT cell function. Indeed, our laboratory has shown that access to cell
surface 5ß1 is essential for EVT cell migration (7).
Recent studies have revealed that EVT cell migration is stimulated in
an autocrine manner by EVT cell-derived insulin-like growth factor
(IGF)-II and in a paracrine manner by decidua-derived IGF-binding
protein (IGFBP)-1 (4, 7). Migration stimulation by either
molecule can occur independent of their binding to each other, and
invasion-stimulating effects of either molecule are explained primarily
by this migration-promoting action (4). It was suggested
that the IGFBP-1 effects were mediated via 5ß1 integrin (4, 7).
A family of six IGFBPs that can bind both IGF-I and IGF-II with high
affinity have been identified and sequenced (8). Though
these proteins can modulate IGF actions on target cells in an
inhibitory (as well as potentiating) manner (9), there is
now a growing body of evidence to suggest that some of the IGFBPs are
able to exert IGF-independent actions (10, 11). IGFBP-1
contains an Arg-Gly-Asp (RGD) domain (12) similar to
certain ECM proteins such as fibronectin, capable of binding to the RGD
recognition sites of certain integrins. In vitro studies
have demonstrated that IGFBP-1, via its RGD domain, can bind to the
5ß1 integrin, leading to IGF-independent stimulation of migration
in Chinese hamster ovary (CHO) cells (13), as well as
porcine vascular smooth muscle cells (14). However, the
precise signaling mechanisms underlying this stimulation remained
unclear. IGF-independent inhibition of protein synthesis in human
skeletal muscle cells, in response to IGFBP-1, was reported to occur
via ß1 integrin binding and stimulation of a rapamycin-sensitive
signaling pathway (15).
Interaction of the cell surface with the ECM components, via a large family of integrin receptors, seems to play an important role during morphogenesis (16), as well as in tumorigenesis (17). In vitro, occupation of integrins with their ligands leads to formation of focal adhesions resulting from an interaction of the cytoplasmic tails of integrins with the actin-containing cytoskeleton. Focal adhesions occur where integrins are clustered. This clustering was shown to increase phosphorylation of a focal adhesion-associated kinase tyrosine kinase (18), now known as focal adhesion kinase (FAK).
Integrins have short cytoplasmic domains without endogenous catalytic activity. To function in signal transduction, these domains must interact with other catalytic molecules for signaling to the cell interior (19, 20). Occupation of integrins with ligands can lead to activation of members of the mitogen-activated protein kinase (MAPK) cascade, especially extracellular-regulated protein kinases (ERK) (21), as well as FAK (22). Furthermore, growth factors capable of binding to integrins can also lead to FAK phosphorylation (23). There is currently a conflicting body of evidence as to whether FAK is involved (24) or not involved (25) in integrin-triggered ERK activation. Similarly, data exist to show that Ras is either required (26) or not required (27) for integrin-mediated MAPK activation.
Currently, the precise signaling pathway used by IGFBP-1 for
IGF-independent stimulation of EVT cell migration remains unknown. The
present study used a human EVT cell line HTR-8/svneo, produced in our
laboratory from a first-trimester placenta (28), to
identify molecular mechanisms responsible for IGFBP-1-mediated
stimulation of EVT cell migration. We tested whether this stimulation
was dependent on binding of the RGD domain of IGFBP-1 to an RGD binding
site on the 5ß1 integrin, followed by activation of FAK and
stimulation of MAPK pathway.
Materials and Methods
Reagents and antibodies
Cells were cultured in RPMI 1640 (Life Technologies, Inc., Burlington, ON) containing 10% FBS (Life Technologies, Inc.) with 2% penicillin/streptomycin (Life Technologies, Inc.). In migration assays, cells were suspended
in RPMI 1640 with 1% FBS and supplemented with 0.01% BSA
(Sigma, Oakville, ON). Recombinant IGFBP-1, produced by
CHO cells, was purchased from Upstate Biotechnology, Inc.
(Lake Placid, NY). Mutated IGFBP-1, derived from CHO cells, was
kindly provided by David Clemmons, University of North Carolina, Chapel
Hill, NC. MTT [3-(4,
5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide] was
purchased from Sigma. A polyclonal goat antihuman 5ß1
antibody was purchased from Chemicon International Inc. (Mississauga,
ON). This antibody has been shown to block binding of fibronectin to
human endothelial cells via the RGD binding domain of the 5ß1
integrin (29). Gly-Arg-Gly-Asp-Ser-Pro (GRGDSP) and
Gly-Arg-Gly-Glu-Ser-Pro (GRGESP) hexapeptides were purchased from
Life Technologies, Inc. Herbimycin A, a tyrosine kinase
inhibitor, was purchased from Sigma. PD 98059
(2-[2'-amino-3'methoxyphenyl]-oxanaphthalen-4-one) is a compound that
specifically inhibits MAPK kinase (MEK) and was purchased from
Calbiochem (San Diego, CA). Mouse monoclonal antibody
against phosphotyrosine, mouse monoclonal antihuman phosphorylated-ERK
antibody, polyclonal goat antihuman ERK-1 (cross-reactive with ERK-2),
and rabbit antihuman FAK antibody were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Horseradish
peroxidase-conjugated goat antimouse IgG, goat antirabbit IgG, and
swine antigoat IgG, as well as fluorescein
isothiocyanate-conjugated goat antirabbit IgG, were purchased
from Cedarlane (Hornby, ON).
EVT cell line and culture
HTR-8 is an EVT cell line produced in this laboratory from
primary cultures of cytotrophoblast cells, growing out of
first-trimester chorionic villus explants and propagated further, as
detailed earlier (30). These cells are short-lived (living
up to 12–15 passages) and express all the markers of the EVT cells
in situ, such as cytokeratin 7, 8, and 18; placental type
alkaline phosphatase; high-affinity urokinase type plasminogen
activator receptor; HLA framework antigen W6/32, IGF-II messenger RNA
and protein; and a selective repertoire of integrins 1, 3, 5,
v, ß1, and vitronectin receptor vß3/ß5
(30). These cells also express HLA-G messenger RNA
and protein when grown on Matrigel or laminin (31).
HTR-8/svneo cell line, employed in the present study, was produced by
immortalization of HTR-8 cells with sv40 Tag transfection
(28). It shares fully all the phenotypic and functional
(proliferative, migratory, and invasive) characteristics of HTR-8
cells, including responsiveness to migration-stimulating signals of
IGF-II and IGFBP-1 and migration-inhibitory effects of transforming
growth factor (TGF)-ß. Like the parental HTR-8 cells, this cell line
is incapable of anchorage-independent growth or tumorigenicity in nude
mice (indicating the absence of a transformed phenotype)
(28), thus providing an exquisite in vitro
model for studies of EVT cell biology. Other laboratories have
confirmed that phenotypic and functional behaviors of these cells
during Matrigel invasion simulate those of EVT cells in primary
cultures (32). In the present study, cells between 70–90
passages were used and grown in RPMI 1640 supplemented with 10%
FBS.
MTT assay
Cell proliferation levels were evaluated using MTT as a marker for cellularity after defined periods of culture. Cells were harvested using a 0.05% trypsin-PBS/EDTA solution and resuspended in serum-reduced medium (RPMI + 1% FBS). Cell suspensions of 2.5 x 104 cells/100 µL were plated into 96-well plates in serum-reduced media with two concentrations of recombinant mammalian IGFBP-1 (0.5 nmol/L and 1.0 nmol/L). Ninety-six-well plates were then incubated for 48 h, without any additional supplementation, at 37 C in a 5%-CO2 incubator. At the end of the incubation, each well received 25 µL of 0.5% MTT solution. The plates were then returned to the incubator for a period of 2 h. At the completion of this second incubation, 100 µL extraction buffer (20 µg SDS in 80 mL N,N-dimethyl formamide and water, to a vol of 100 mL, pH 4.7) was added and mixed thoroughly. The plates were then incubated overnight at 37 C in a 5%-CO2 incubator. The plates were read on a spectrophotometer at an absorbance of 540 nm.
Transwell migration assay
Migration assays were conducted in transwells fitted with
Millipore Corp. membranes (6.5-mm filters, 8-µm pore
size; Costar, Toronto, ON); 2.5 x
104 HTR-8/svneo cells·100 µL serum-reduced
medium (RPMI 1640 with 1% FBS) were plated in upper wells of transwell
chambers containing either 200 µL serum-reduced medium, or
serum-reduced medium with various concentrations (0.1, 0.5, 1.0, 10
nmol/L) of IGFBP-1. Use of serum-reduced medium was found to be a
prerequisite for a full detection of IGFBP-1 effects (4).
After the optimal concentrations and the temporal kinetics of IGFBP-1
stimulation of migration were established, subsequent migration assays
were done, mostly at 48 h, in the presence of IGFBP-1 (1.0
nmol/L) ± polyclonal 1:50 or 1:100 dilution of 5ß1 antibody,
200 µmol/L GRGDSP and GRGESP hexapeptides, 1.0 nmol/L mutated
IGFBP-1, 0.5 µg/mL Herbimycin A, and 10 and 30 µmol/L MEK inhibitor
PD 98059. When cells were pretreated with drugs or antibodies, IGFBP-1
was always added to both the upper and lower wells. Lower wells
contained 800 µL complete media, or 800 µL complete media with 1.0
nmol/L IGFBP-1. Chambers were assembled and incubated for 24, 48, or
72 h in a humidified environment (5% CO2)
at 37 C. After incubation, cells from the upper surface of
Millipore Corp. membranes were completely removed with
gentle swabbing; remaining migrant cells were fixed and stained using
Diff-Quik Stain Set (Dada AG, Dudingen, Switzerland). Membranes were
then rinsed with distilled water, cut from transwells, and mounted onto
glass slides. Cellular migration indices were determined by counting
the number of stained cells on the membrane, in five randomly selected,
nonoverlapping fields, at 400x magnification, under a light microscope
(researcher blind to experimental conditions).
Immunoprecipitation and immunoblotting of FAK and MAPK proteins
Cells were grown on poly-L-lysine-coated dishes and serum starved overnight. After treatment for 0, 5, 10, 30, or 60 min with 1.0 nmol/L IGFBP-1 in fresh medium, the cells were rinsed twice with cold PBS and lysed with RIPA buffer (150 mM NaCl; 50 mM Tris-HCl, pH 7.5; 1% Triton X-100; 1% deoxycholate; 0.1% SDS; and 2 mM EDTA) containing phosphatase inhibitors (50 mM NaF and 1 mM Na3VO4, including a Complete Mini tablet; Boehringer Ingelheim GmbH, Mannheim, Germany) for 1/2 h at 4 C. The lysates were centrifuged at 15,000 rpm for 15 min, to remove any insoluble material. The amount of total protein was determined using the BCA protein assay reagent (Pierce Chemical Co., Brockville, ON). The protein samples were normalized before being used for either direct immunoblotting of total cell proteins or immunoprecipitation of equivalent amounts of proteins with FAK antibody (2 µg antibody/500 µg protein, for 2 h at 4 C) bound to GammaBind G-Sepharose beads (30 µL for 2 h at 4 C; Pharmacia Biotech AB, Uppsala, Sweden). The resultant immunocomplexes were rinsed three times with lysis buffer to remove unbound proteins and then resuspended in reducing sample buffer. Samples were heated to 100 C for 5 min, analyzed by SDS-PAGE, and then transferred to nitrocellulose and blocked with either nonfat milk (5%) or 3% BSA before immunoblot analysis using antiphosphorylated ERK or antiphosphotyrosine antibodies and horseradish-peroxidase-conjugated goat antimouse secondary antibody. The blots were then visualized using an enhanced chemiluminescence system (ECL Plus Western Blotting Detection System, Amersham Pharmacia Biotech, Oakville, ON). For control purposes, some of the blots were stripped and reprobed for total proteins, using the appropriate antibodies, and subjected to the same chemiluminescence system. Adobe Photoshop, Version 4.0 was used to prepare digital images.
Immunofluorescence microscopy
Cells were allowed to spread and adhere, in complete media, onto 12-mm glass coverslips. The cells were then subjected to serum starvation overnight. Cells were treated for 0, 5, 10, 30, or 60 min with 1.0 nmol/L IGFBP-1 and washed with PBS. Cells were then fixed with 100% cold methanol. After three washes with PBS, cells were treated with 10% normal goat serum to block nonspecific protein binding sites. FAK was visualized by incubating first with rabbit antihuman FAK antibody (1/100 dilution, for 1 h at 4 C) and then with fluorescein isothiocyanate-conjugated goat antirabbit IgG (1/100 dilution, for 1 h at room temperature, in the dark). A fluorescent mounting medium (DAKO Corp., Carpinteria, CA) containing Hoechst stain (Sigma) was used to mount the coverslip for visualization.
Statistics
Data were analyzed using one-way (for single treatments) and two-way ANOVA (for multiple treatments), followed by the Tukey test. Because data were not normally distributed, the Mann-Whitney rank sum test was employed to determine the level of significance of differences in pairs of various treatment groups. Differences of P < 0.05 were considered significant.
Results
IGFBP-1 stimulates migration of HTR-8/svneo cells
We had previously reported that IGFBP-1 stimulates migration
of EVT (HTR-8) cells, from which HTR-8/svneo cells were derived, when
tested with a monolayer wound assay (7) or a transwell
migration assay, in which radioactivity was used as a measure of
migration of cells prelabeled with tritiated thymidine
(4). In the present study, using visual counting of
migrant cells in a transwell assay, we have followed the temporal
kinetics of migration of HTR-8/svneo cells in the presence or absence
of IGFBP-1, and TGF-ß (employed as a positive control for migration
inhibition) to show that the normal migration, as well as
migration-stimulation by IGFBP-1, peaked at 48 h (Fig. 1
). A further dose-response study (using
0.1–10 nmol/L IGFBP-1) established that, with the present assay,
the stimulating effects of IGFBP-1 reach a plateau at 1.0 nmol/L
concentration (data not shown), which was employed in all subsequent
studies, unless specified.
|
In our earlier studies employing tritiated thymidine uptake as a
measure of proliferation of HTR-8 cells, no significant effect on
proliferation was noted with IGFBP-1 concentrations ranging between
0.1–500 nmol/L (4). In the present study, we tested again
whether migration-stimulating effects of IGFBP-1 could be
explained, at least in part, by a stimulation of proliferation, as
measured with a 48-h MTT assay (Table
1). At concentrations of 0.5 and 1.0
nmol/L, which stimulated migration, no effect on cellularity was
observed, thus excluding the possibility that the migration stimulation
was secondary to proliferation stimulation by IGFBP-1.
|
5ß1 antibody
To examine the requirement of access to the 5ß1
integrin for normal migration or IGFBP-1 stimulation of migration,
HTR-8/svneo cells were preincubated for 1 h with various dilutions
of a polyclonal anti-5ß1 integrin-blocking (29)
antibody. In a 48-h transwell migration assay, all dilutions (1/200,
1/150, 1/100, and 1/50) of the antibody alone caused a significant
decrease in cell migration, compared with control (P <
0.001). Further experiments showed that preincubation with the antibody
(1/50 and 1/100 dilutions), followed by treatment with IGFBP-1 (1.0
nmol/L), did not abrogate the inhibitory effects of the antibody. In
fact, the levels of migration were similar to the completely inhibited
levels observed by incubation with the integrin antibody alone (Fig. 2). These findings demonstrate the
requirement of the 5ß1 integrin for basal EVT cell migration, as
well as migration-stimulating effects of IGFBP-1, but do not
necessarily provide evidence as to whether the IGFBP-1 effects were
dependent on the binding of its RGD sequence to the RGD binding domain
of the 5ß1 integrin on the EVT cell surface.
|
5ß1 integrin is
required for IGFBP-1 stimulation of migration in HTR-8/svneo cells
The significance of the RGD domain on IGFBP-1 and its
interaction with the RGD binding site on the 5ß1 integrin,
expressed by the EVT cells, was investigated by pretreating the cells
for 1 h with excess (200 µmol/L) of a GRGDSP hexapeptide and
then using the cells inclusive of the hexapeptide in the transwell
migration assay in the presence or absence of IGFBP-1. This treatment
alone had no effect on basal migration in most experiments, but it
occasionally caused a minor stimulation (as shown in Fig 3). However, a preincubation with this
hexapeptide completely abrogated the migration-stimulating effects of
IGFBP-1 (P < 0.001). Pretreatment of cells with
the same concentration of a GRGESP hexapeptide (used as negative
control) showed no effect on either the basal migration or
IGFBP-1-mediated stimulation of migration (Fig. 3). Because the GRGDSP,
and not the GRGESP hexapeptide, has the ability to occupy the RGD
binding domain on the 5ß1 integrin, these results demonstrate the
requirement of access to this domain for IGFBP-1 action. Further
experiments employed mutated IGFBP-1 to substantiate these
results.
|
). These results again demonstrate the
requirement of the RGD domain of IGFBP-1 for stimulation of
migration.
|
Detection and localization of FAK was examined after various
time periods of treatment with IGFBP-1 (0.5 nmol/L). At 5
and 10 min, there was an apparently greater intensity of fluorescence
than at 0 min (control) or at later time points (30 and 60 min) (Fig. 5; A, B, C, D, and E) most likely because
of a higher degree of cell spreading, given that there was no change in
the total level of FAK protein identified in Western blots (see Fig. 7A). Addition of secondary antibody, after exclusion of the primary
antibody or exposure to rabbit IgG, showed no detectable fluorescence
(not shown in figure plate). After 10 min of treatment with IGFBP-1,
focal localization of strong fluorescence became apparent in the
lamellipodial extensions of the cells (as evident at a higher
magnification, Fig. 5F). These results indicate a possible
redistribution of FAK to cellular attachment sites in response to
IGFBP-1.
|
|
HTR-8/sveo cells that were treated for 5 and 10 min with
IGFBP-1 (0.5 nmol/L) showed an increase in ERK-1 and ERK-2
phosphorylation, compared with control. The greatest effect was
observed after 10 min of treatment. At 30 and 60 min, the
phosphorylation level returned to control levels (Fig. 6A). Because IGFBP-1 was included in
replenished medium, in control experiments we tested whether medium
replenishment alone had any effect on MAPK phosphorylation at similar
time intervals. Because no effect was observed (data not shown), the
IGFBP-1 stimulation was considered real. These results for IGFBP-1
were reproducible and were substantiated by densitometric
measurements, relative to total ERK-1 and ERK-2. They demonstrate that
MAPK pathway is activated in HTR-8/svneo cells in response to IGFBP-1.
Preincubation of HTR-8/svneo cells for 30 min with 10 µmol/L and
30 µmol/L of PD 98059, a well-characterized MEK inhibitor
(33), blocked basal and IGFBP-1-induced MAPK
phosphorylation in duplicate experiments (results not presented).
|
). PD 98059 did not
have any effect on cell viability, as measured by a trypan blue
exclusion assay. These results demonstrate the requirement of the MAPK
pathway for IGFBP-1-mediated stimulation of migration. IGFBP-1 treatment stimulates phosphorylation of FAK, and herbimycin A abrogates IGFBP-1 stimulation of migration in HTR-8/svneo cells
HTR-8/svneo cells that were treated for 5 and 10 min with IGFBP-1 (0.5 nmol/L) showed a significant increase in phosphorylation of FAK, compared with control (Fig. 7A). The greatest effect was observed after 10 min of treatment. This finding shows that there is a stimulation of FAK phosphorylation in response to IGFBP-1, and perhaps this stimulation is required for promotion of EVT cell migration.
To test this possibility, we introduced the drug Herbimycin A, a
known tyrosine kinase inhibitor (34), because a specific
FAK inhibitor is not available. Herbimycin A pretreatment of cells (0.5
mg/mL for 1 h) inhibited both basal and IGFBP-1-induced FAK
phosphorylation in HTR-8/svneo cells (data not shown). A similar
pretreatment did not have any effect on the basal migration of
HTR-8/sveno cells in a 48-h transwell migration assay. However, this
pretreatment caused a significant abrogation (P <
0.001) of IGFBP-1-mediated stimulation of migration (Fig. 7B).
These results demonstrate the role of tyrosine kinase activity
for the mediation of migration-stimulating responses to IGFBP-1.
Discussion
An adequate perfusion of the placenta by maternal blood is
dependent on optimal migration and invasion of EVT cells into the
decidua, followed by a remodeling of the uteroplacental arteries
(1). Poor EVT cell migration and invasiveness is a key
pathological feature of preeclampsia (35). IGFBP-1 is a
major decidual cell product (36), which was shown by us to
promote EVT cell invasiveness by stimulating migration; this
stimulation did not require the presence of IGFs (4, 7).
The present study used an immortalized human EVT cell line with
phenotypic and functional characteristics of EVT cells in
situ to investigate the nature of the receptor-ligand interaction,
as well as signal transduction events underlying IGFBP-1
stimulation of migratory function. We show that IGFBP-1 stimulation
of EVT cell migration occurs possibly by binding of its RGD domain to
the 5ß1 integrin, leading to activation of FAK and stimulation of
MAPK pathway.
Using visual counting of migrant cells in a transwell migration assay, we have confirmed our earlier findings of IGFBP-1 stimulation of nonimmortalized EVT cell migration, as measured with a monolayer wound assay (7) and a radioactive method of transwell migration assay (4). Using a cell proliferation assay in the presence of similar concentrations of IGFBP-1, which were shown to stimulate migration, we excluded the possibility that the observed increase in migration was attributable to an increase in proliferation of the EVT cells. Our present and earlier (4) data failed to support the notion of an invasion-blocking role for IGFBP-1 (37). This suggestion was made on the basis of the findings of enhanced trophoblast penetration of an in vitro decidualized endometrial stromal cell layer in the presence of a high dose of insulin (5 µg/mL), which reduces IGFBP-1 production by decidual cells and a reversal of the effects with a high dose (1 µg/mL) of IGFBP-1. In the above study, other effects of insulin on the decidua or the trophoblast were not ruled out, and the physiological significance of a relatively high concentration (1 µg/mL, i.e. 40 nmol/L) of IGFBP-1 used remains in question. Although IGFBP-1 concentration in the pericellular environment of placental EVT at different gestational periods is not known, its concentration in the maternal serum, between 16 and 36 weeks of normal pregnancy, ranges between 80–360 mg/L or 3.2–12 nmol/L (38). Therefore, the concentration (1 nmol/L) of IGFBP-1 at which we obtain highest migration stimulation, reaching a plateau between 1–10 nmol/L, seems to be highly physiological.
The discovery that EVT cells that migrate out of chorionic villi
in situ selectively express 5ß1 integrin (5, 6) led to an increased interest in the possible functional
relationship of this receptor to the migratory phenotype of EVT cells.
Integrins were traditionally believed to be mediators of cell-to-ECM
and cell-to-cell adhesion. However, more recent studies indicate that
they are capable of functioning as true receptors when occupied by ECM
components (19). In our study, treatment of EVT cells with
an 5ß1 integrin blocking antibody (29) showed an
almost complete inhibition of basal migration, which could not be
rescued by later exposure to IGFBP-1. These results confirm earlier
findings from this laboratory using the monolayer wound assay
(7) and demonstrated the requirement of access to this
receptor for EVT cell migration. However, they do not necessarily prove
that IGFBP-1 effects were dependent on binding to the 5ß1
integrin. It is likely that fibronectin, which is produced by EVT cells
in situ, is used by these cells for anchorage via 5ß1
integrin receptor before migration (7). Our results
are in discordance with the reported invasion-stimulating effects of an
5-blocking antibody tested in another system using villous
cytotrophoblast cells placed on Matrigel (39). Differences
in the cell system and experimental protocols used in this study do not
permit an adequate comparison with our studies. However, another study,
using a villus explant organ culture assay, revealed that treatment
with the 5-blocking antibody interfered with the outgrowth of EVT
cell columns and caused rounding of the spindle-shaped EVT cells
(40). We interpret the latter findings as indication of
loss of cell motility.
The presence of an RGD sequence on IGFBP-1 protein was first recognized
on the basis of the original cloning and sequencing of IGFBP-1
(12). The functional significance of this sequence became
evident with the demonstration that this sequence was present in both
IGFBP-1 and IGFBP-2 and that RGD binding sites were present in several
integrins, including 5ß1 (41). A number of
observations by Clemmons and his associates (13, 42, 43), using CHO cells expressing cell surface 5ß1,
established that IGFBP-1 stimulated their migration in an
IGF-independent manner by binding to the RGD binding domain of the
5ß1 integrin. First, IGFBP-1 bound to 5ß1 on CHO cells and
stimulated their migration. Second, recombinant IGFBP-1, mutated at its
RGD site (to WGD), lost this ability (in spite of retaining IGF binding
function). Third, migratory capability of CHO cells transfected with a
mutated IGFBP-1 construct was poorer than those transfected with the
wild-type construct. Migration-promoting effects of IGFBP-1 were
subsequently shown with a more physiologically relevant model using
vascular smooth muscle cells that expressed 5ß1 integrin, as well
as other integrin receptors (14). In the present study, we
demonstrated that IGFBP-1 effects were caused by the binding of the RGD
sequence to the RGD binding domain on the EVT cell surface.
The migration-promoting role of IGFBP-1, produced in situ at the choriodecidual interface, seems to be important for trophoblast physiology. A similar role is suggested for cutaneous wound healing. Indeed, IGFBP-1 in combination with IGF-1 has been used as a therapeutic agent to accelerate cutaneous wound healing in vivo (44). This therapeutic effect was abolished by mutation of the RGD domain of IGFBP-1 (43).
The mechanisms underlying IGFBP-1-mediated stimulation of migration in
various cell species (7, 13, 14) largely remained unknown.
It was proposed by Jones et al. (13) that
IGFBP-1 may be competing with the natural ligand fibronectin, which may
promote cell attachment but impede cell motility. IGFBP-1 would occupy
the fibronectin binding sites and thus improve cell detachment. Though
this hypothesis has not been tested in the present model, it is evident
from our experiments that IGFBP-1 effects are attributable to direct
signaling through its receptor 5ß1, because IGFBP-1 induces rapid
phosphorylation of some important signaling molecules, such as FAK and
MAPK. It remains to be established whether the two phosphorylation
events are linked.
Cytoplasmic domains of integrin receptors are extremely short and lack kinase activity; however, cytoplasmic tails of ß1 and ß3 chains have been shown to associate with FAK, as well as other proteins, like talin and paxillin (45, 46, 47), the activation of which may be important for integrin-mediated signal transduction. This is illustrated by the enhancement of FAK phosphorylation after binding of cells to the ECM proteins (47). When EVT cells were treated with IGFBP-1, an increase in phosphorylated FAK was observed, with the highest levels noted at 5 and 10 min after treatment. However, IGFBP-1, at much higher concentrations, has been shown to dephosphorylate FAK in a human breast cancer cell line (48). We have shown that 1.0 nmol/L IGFBP-1 in the same cell line caused phosphorylation of FAK (unpublished data) in the same manner that we have observed in EVT cells. Microscopic examination of the IGFBP-1-treated EVT cells, by immunofluorescent confocal microscopy, showed an accumulation of FAK in what seemed to be focal adhesions (i.e. at the tips of the cellular lamellipodia) most evident at 10 min. These results, taken together, suggest that focal adhesion formation, along with selective accumulation of activated FAK in the adhesion points, was in preparation for cell migration. FAK phosphorylation can also be induced by growth factors and hormones; in which case, integrins may play an intermediary role. Platelet-derived growth factor stimulates tyrosine phosphorylation of paxillin and FAK in adherent Swiss 3T3 fibroblast cells (49). Furthermore, after only minutes of exposure to hepatocyte growth factor, fibroblast cells showed an increase in FAK tyrosine phosphorylation (50). In this study, microscopic examination of cells revealed a two-step response: initially cells spread rapidly and formed focal adhesions; after which, the focal adhesions disassembled, and increased cell locomotion occurred. These authors also showed that the drug Herbimycin A, which is a known inhibitor of tyrosine kinases, inhibited hepatocyte growth factor-mediated tyrosine phosphorylation of FAK and blocked their migratory response to the growth factor. Treatment of our cells with Herbimycin A, which blocks phosphorylation of tyrosine kinases (including FAK), showed a significant abrogation of IGFBP-1 stimulation of cellular migration, indicating the functional role of tyrosine kinases in this event.
The functional role of FAK in cell migration has been convincingly established with genetic studies (51, 52). Although FAK-deficient embryos were able to implant, they showed generalized defects in the development of the mesoderm (the germ layer consisting of cells with the highest migratory ability). In addition, cells from both the mesoderm and the endoderm exhibited impaired migratory ability in vitro, when cultured on fibronectin. Conversely, FAK overexpression was shown to increase migratory ability of fibroblasts on fibronectin (53). FAK has been shown to be elevated in migrating keratinocytes, which are involved in burn repair (54). Furthermore, FAK overexpression has been reported in invasive and metastatic colon and breast tumors (55) and in rapidly migrating melanoma cell lines (56).
In addition to stimulating FAK, integrin ligation can also lead to activation of the extracellular-signal regulated kinase (ERK) pathway, as demonstrated by the transient phosphorylation of p42 and p44 MAP kinases, ERK1 and ERK2 (21, 24). In the present study, treatment of EVT cells with IGFBP-1 caused a rapid increase in phosphorylation of ERK-1 and ERK-2, with the highest levels being at 5 and 10 min after treatment, indicating that IGFBP-1 stimulation uses the MAPK pathway. This notion was confirmed by treatment with a specific MEK inhibitor, PD 98059 (33). This drug showed an inhibitory effect on ERK phosphorylation as well as on EVT cell migration, which could not be reversed in the presence of IGFBP-1. The fact that a high concentration (30 µmol/L) of the MEK inhibitor decreased basal migration of EVT cells in our study suggests that MEK is an important intermediary for the migratory signals in EVT cells. IGFBP-1 may represent one of several molecules that provide migration-promoting signals via the MAPK pathway, after binding to their respective receptors on the EVT cell surface. Indeed, studies in our laboratory have shown that IGF-II is another molecule in this category that stimulates EVT cell migration, independent of IGFBP-1, by using the MAPK pathway (57). MAPK may represent the common final pathway for numerous migration promoting signals (58) because of its ability to activate myosin light chain kinase, leading to the phosphorylation of myosin light chains, thus promoting cytoskeletal contraction.
We have not tested whether FAK phosphorylation was obligatory for the
activation of the MAPK pathway after IGFBP-1 treatment of EVT cells.
Both the presence (59) and absence (25) of a
link between FAK and MAPK signaling pathways have been reported in
other cells. Recent studies of 5ß1 integrin interaction with
fibronectin, using CHO cells, indicate that signaling via both FAK and
ERK-2 was proportional to the number of receptor-ligand bonds
(60).
IGFBP-1 seems to be the major paracrine (decidua-derived), and IGF-II the major autocrine (trophoblast-derived), molecule that can independently up-regulate the migratory function of EVT cells, an essential step for their invasive function. Indeed, blocking type 2 IGF receptor, which is responsible for IGF-II action on EVT cells, with a blocking antibody did not abrogate the action of IGFBP-1 (T. McKinnon, unpublished). A down-regulation of trophoblast invasiveness (61) or migratory ability (7) is provided in situ by TGF-ß, which is primarily a paracrine (decidua-derived) factor (61). The production of both of these molecules (IGFBP-1 and TGF-ß) by the decidual cells that have opposing regulatory roles on trophoblast migration and invasion may seem paradoxical. However, TGF-ß is secreted by the decidua in its latent form (61). It is possibly activated by plasmin in the presence of trophoblast-derived urokinase-type plasminogen activator to provide a feedback control against excessive invasion at the invasion front (3). Further studies are needed to investigate whether the intricate balance between migration/invasion-promoting and inhibiting molecules, and trophoblast response to these regulatory molecules, are deranged in preeclampsia, a trophoblast disorder of hypoinvasiveness.
In this regard, a recent longitudinal study (38), comparing IGFBP-1 levels in the maternal plasma during the course (16–36 weeks gestation) of normal pregnancy and pregnancies complicated with preeclampsia, is highly relevant. It was shown that the levels were significantly lower in women destined to be preeclamptic at earlier gestational ages (16, 20, and 24 weeks). The levels in these women increased progressively, to reach control levels at 28 and 32 weeks and then exceed control levels at 36 weeks. On the other hand, the levels remained essentially steady during the course of normal pregnancy. These findings, taken together with the present data, suggest that low IGFBP-1 levels during early gestation may serve as a diagnostic tool and possibly a pathobiological marker for the development of preeclampsia.
Footnotes
1 Supported by a grant from the Medical Research Council of Canada
(Canadian Institutes for Health Research). 
Received June 7, 2000.
Revised October 24, 2000.
Accepted November 30, 2000.
References
64 in
human trophoblast cells and its loss from extravillous trophoblast
cells. Placenta. 14:203–215.
5ß1 integrin by means of its
Arg-Gly-Asp sequence. Proc Natl Acad Sci USA. 90:10553–10557.5ß1
integrin and inhibits cytotrophoblast invasion into decidualized
endometrial stromal cultures. Growth Horm IGF Res. 8:21–31.[Medline]
5ß1. Biol Reprod. 60:828–838.This article has been cited by other articles:
 |
|
 |
|
  C. Nicola, P. K. Lala, and C. Chakraborty Prostaglandin E2-Mediated Migration of Human Trophoblast Requires RAC1 and CDC42 Biol Reprod, June 1, 2008; 78(6): 976 - 982. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Nicola, A. Chirpac, P. K. Lala, and C. Chakraborty Roles of Rho Guanosine 5'-Triphosphatase A, Rho Kinases, and Extracellular Signal Regulated Kinase (1/2) in Prostaglandin E2-Mediated Migration of First-Trimester Human Extravillous Trophoblast Endocrinology, March 1, 2008; 149(3): 1243 - 1251. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Fu, J. A. Thompson, and L. A. Bach Promotion of Cancer Cell Migration: AN INSULIN-LIKE GROWTH FACTOR (IGF)-INDEPENDENT ACTION OF IGF-BINDING PROTEIN-6 J. Biol. Chem., August 3, 2007; 282(31): 22298 - 22306. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L.E. Diaz, Y-C. Chuan, M. Lewitt, L. Fernandez-Perez, S. Carrasco-Rodriguez, M. Sanchez-Gomez, and A. Flores-Morales IGF-II regulates metastatic properties of choriocarcinoma cells through the activation of the insulin receptor Mol. Hum. Reprod., August 1, 2007; 13(8): 567 - 576. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. Mukherjee and C. Guidry The Insulin-Like Growth Factor System Modulates Retinal Pigment Epithelial Cell Tractional Force Generation Invest. Ophthalmol. Vis. Sci., April 1, 2007; 48(4): 1892 - 1899. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
N. Ben Lagha, D. Seurin, Y. Le Bouc, M. Binoux, A. Berdal, P. Menuelle, and S. Babajko Insulin-Like Growth Factor Binding Protein (IGFBP-1) Involvement in Intrauterine Growth Retardation: Study on IGFBP-1 Overexpressing Transgenic Mice Endocrinology, October 1, 2006; 147(10): 4730 - 4737. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Jasinska, Z. Strakova, M. Szmidt, and A. T. Fazleabas Human Chorionic Gonadotropin and Decidualization in Vitro Inhibits Cytochalasin-D-Induced Apoptosis in Cultured Endometrial Stromal Fibroblasts Endocrinology, September 1, 2006; 147(9): 4112 - 4121. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. L. Martin and S. Jambazov Insulin-Like Growth Factor Binding Protein-3 in Extracellular Matrix Stimulates Adhesion of Breast Epithelial Cells and Activation of p44/42 Mitogen-Activated Protein Kinase Endocrinology, September 1, 2006; 147(9): 4400 - 4409. [Abstract] [Full Text] [PDF]
|
|
) or a pretreatment with the
antibody followed by treatment with IGFBP-1 (
) on HTR-8/svneo cell
migration at 48 h. Data are presented as a percentage of control
(medium alone, normalized to 100%). Error bars,
SEM, n = 10/group; bars not sharing the same
subscript, data that are significantly different. A,
Pretreatment with the antibody alone (at all dilutions 1/50 to 1/200)
caused a significant (P < 0.001) reduction in
migration. B, Significant increase (P < 0.01) in
migration was noted in the presence of IGFBP-1 (1.0 nmol/L). However,
preincubation with anti-
