| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Regular Articles |
ko Ili
*
ev*
From the Departments of Stomatology,*
Anatomy,
Obstetrics, Gynecology and
Reproductive Sciences,
and Pharmaceutical
Chemistry,
University of California San
Francisco, San Francisco, California; and Quality Controlled
Biochemicals,¶
Division of BioSource
International, Hopkington, Massachusetts
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
During development of the human
placenta, the organ’s specialized epithelial cells, termed
cytotrophoblasts (CTBs), encounter many of the same obstacles that
tumor cells must overcome during clonal selection. Specifically, a
subset of these CTBs must detach from the basement membrane, where they
reside as a monolayer in the fetal component of the placenta, and
invade the uterus, where they survive only if they access a supply of
maternal blood. Because subsequent human development depends on the
tumor-like properties of this subpopulation of CTBs, the molecular
mechanisms that are involved have been carefully programmed.
Consequently, the cells acquire the ability to invade as part of a
tightly regulated, stepwise differentiation process (diagrammed in
Figure 1A
).1
Furthermore,
this subset of CTBs stops proliferating once it initiates the program
leading to an invasive phenotype.2
Therefore, studying the
process of CTB differentiation/invasion offers a unique opportunity to
identify, by drawing functional analogies, molecules particularly
likely to play important roles in promoting the invasive phenotype of
tumor cells.
|
6ß4 and
E-cadherin—and the onset of expression of adhesion receptors expressed
on endothelium—VE-cadherin, Ig family members VCAM-1 and PECAM-1, and
integrins vß3 and
1ß1.3-6
Cadherin-11 is also up-regulated on extravillus CTBs and on
decidualizing endometrial stroma.7,8
By using an in
vitro model of CTB invasion that replicates the process of
adhesion molecule switching in vivo, we assessed the
contribution of individual adhesive interactions to the invasion
process. In this regard, we found a particularly important role for
integrin cell-extracellular matrix (ECM) receptors. A number of these
molecules whose expression is up-regulated during differentiation
promote invasion (eg,
1ß1,
vß3), whereas others
are inhibitory (eg,
5ß1). Thus, this
highly regulated invasive program is governed by a system of checks and
balances. Together, our previous work suggests that the intracellular mechanisms that CTBs use to translate the information they receive via integrin receptors into signals is crucial to their ability to invade the uterus. Integrin interactions with ECM components lead both to occupancy of their ligand-binding sites and to clustering, which assembles multimolecular signaling complexes in focal adhesions, sites of cell-ECM interactions.9,10 Thus, integrin signals are connected to multiple intracellular signaling pathways.11 One important outcome of integrin clustering by ECM is the (auto)phosphorylation of focal adhesion kinase (FAK) on tyrosine 397 (pY397FAK). pY397FAK, in turn, serves as the backbone of a scaffold that recruits additional signaling proteins to focal adhesions.12 FAK has been implicated in the transduction of survival signals from ECM13-16 and in promoting cell migration.17-22 Both processes are relevant to establishing a pool of cells with a migratory/invasive (metastatic) phenotype.
Several studies have attempted to determine whether there is a correlation between the levels of FAK expression and the acquisition of metastatic potential. Although FAK was overexpressed in some types of highly invasive tumors, the degree of FAK expression was not a significant prognostic factor.23 Interestingly, comparative studies of normal breast tissue and invasive breast carcinoma revealed that FAK levels can be elevated24 or reduced25 in transformed versus normal tissue. From these studies it seems that regulation of total FAK levels is not crucial for promoting an invasive phenotype. Thus, the key might be regulation of its activity.
In the current study we have extended our work on the adhesion
phenotype switching that occurs as CTBs differentiate/invade to address
the putative role(s) of FAK in this unusual tumor-like behavior. Again,
we have used a combination of in situ and in
vitro approaches. Localization studies in placental tissues showed
that CTBs in all stages of differentiation express FAK. However, only a
subset of CTBs invade the uterine wall (Figure 1A)
. Therefore, we
hypothesized that some of these cells had high levels of
(auto)phosphorylated FAK. To test this idea, we immunolocalized
pY397FAK on tissue sections of the fetal-maternal interface. We also
tested the effect of reducing CTB expression of FAK protein on the
cells’ ability to invade a Matrigel substrate in vitro.
Finally, we asked whether conditions that inhibit CTB invasion (hypoxia
in vitro, preeclampsia in vivo) were associated
with a reduction in FAK (auto)phosphorylation, as determined by
measuring FAK kinase activity and by staining for pY397FAK. These
results support the hypothesis that FAK (auto)phosphorylation on Y397
plays a key role in promoting the invasive behavior of normal CTBs and,
by analogy, may also play an important role in tumor development.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The 7D3 rat anti-human cytokeratin monoclonal antibody (mAb) was raised in this laboratory.3 Several antibodies recognized both the phosphorylated and the nonphosphorylated forms of FAK: two mouse mAbs (Transduction Laboratories, Inc., San Diego, CA, and Chemicon, Temecula, CA); rabbit polyclonal antibodies A-17, C-20 (Santa Cruz Biotechnology, Inc., Santa Cruz, CA), and JF1;17 and goat polyclonal antibodies A-17 and C-20 (Santa Cruz Biotechnology Inc.). Anti-paxillin mAb was purchased from Zymed (South San Francisco, CA) and anti-phosphotyrosine 4G10 from UBI (Lake Placid, NY).
An antibody that recognized only FAK (auto)phosphorylated on Tyr397
(pY397FAK) was obtained from BioSource International Inc. (Camarillo,
CA). To validate its use in immunolocalization studies, the specificity
of this antibody was rigorously tested before initiation of the
experiments reported here. Briefly, tissue sections cut from several
organs of fak knockout mice and the parental wild-type
animals were analyzed along with cultured cells from both
strains.17
The samples from fak-null animals,
which served as negative controls, also had enhanced levels of
potentially cross-reactive antigens, including proteins phosphorylated
on tyrosine and the homologous (
50%) FAK family member
Pyk2.26
Initial experiments showed that anti-pY397FAK
antibody at a concentration of 2.5 µg/ml stained tissue sections and
cells from wild-type, but not fak-null, animals. Staining
was completely abolished when the antibody was preincubated with 100
µg/ml of the phosphopeptide used to generate the antibody, whereas
addition of 100 µg/ml of the nonphosphorylated peptide had no effect.
Preincubation with a phosphopeptide (100 µg/ml) that corresponds to
the homologous sequences in Pyk2 had also no effect on the staining
specificity of anti-pY397FAK antibody.
Tissue Sources for Immunolocalization Experiments
Control samples of chorionic villi and the portion of the uterine wall to which they attached were obtained from pregnancy terminations for nonmedical reasons and from uncomplicated spontaneous deliveries. Because of their relatively small size, samples from each first trimester placenta were processed in toto. Second and third trimester samples were from three to five randomly chosen sites. None of the samples had abnormalities that could be detected either grossly or histologically. Tissues were obtained from women whose pregnancies were terminated during the first trimester (52 samples), second trimester (7 samples), or at the time of delivery (8 samples). Conclusions were based on analysis of all samples in each group.
Samples of chorionic villi and the portion of the uterine wall to which
they attached were also obtained by placental bed biopsy of women who
received prenatal care at the University of California San Francisco
(Moffitt/Long Hospital). Informed consent for the procedure was
obtained during one of the late second/early third trimester clinic
visits. The consent form and biopsy procedure were approved by the
University of California San Francisco Committee on Human Research.
Control samples were obtained from eight women who had no underlying
medical conditions at 33 (n = 1), 35
(n = 4), and 36 (n = 3)
weeks of gestation. Eight samples were obtained from preeclamptic
patients at 32 (n = 1), 33
(n = 4), and 35 (n = 3)
weeks of gestation. Preeclampsia was diagnosed according to the
following criteria, recommended by Chesley27
: nulliparity;
no history of hypertension before pregnancy; increase in diastolic
pressure of 15 mm Hg or systolic pressure of 30 mm Hg compared with
blood pressure obtained before 20 weeks of gestation; proteinuria 
0.3
g/24 hours (or 1+ on urine dipstick) in a catheterized specimen;
hyperuricemia >5.5 mg/dl (or 1 SD greater than the normal mean value
before term); return to normal blood pressure and resolution of
proteinuria by 12 weeks postpartum. Severe preeclampsia was diagnosed
according to the following criteria recommended by the American College
of Obstetricians and Gynecologists: systolic blood pressure 160 mm Hg
and/or diastolic pressure 110 mm Hg; proteinuria of 5 g in a
24-hour period or 3+ on urine dipstick; presence of cerebral or visual
disturbances. Two patients were diagnosed with severe preeclampsia and
were delivered by Cesarean section (32 and 33 weeks); six patients with
preeclampsia had vaginal deliveries [33 (n = 3)
and 35 (n = 3) weeks].
Immunolocalization
Placental tissues were processed for double indirect immunolocalization as previously described.2,3 Briefly, tissues were fixed in 3% paraformaldehyde for 30 minutes, washed three times in phosphate-buffered saline (PBS), infiltrated with 5 to 15% sucrose followed by optimal cutting temperature (OCT) compound (Miles Scientific, Naperville, IL), and frozen in liquid nitrogen. Sections (6 µm) were prepared using a cryostat (Slee International, Inc., Tiverton, RI). Nonspecific antibody reactivity was blocked by using a commercially available kit (Vector, South San Francisco, CA). Sections were then incubated for 2 days at 4°C with a mixture of the three antibodies: 1) 2.50 µg/ml of ChromPure donkey whole IgG (Jackson ImmunoResearch, West Grove, PA); 2) 1 µg/ml of 7D3 anti-human cytokeratin; and 3) 1 µg/ml of either anti-FAK (JF1) or anti-pY397FAK. Then, the sections were rinsed three times in PBS for 10 minutes and incubated for 30 to 60 minutes at room temperature with the appropriate species-specific secondary antibodies conjugated to rhodamine (donkey anti-rat IgG) or biotin (donkey anti-mouse or anti-rabbit IgG), and washed three times in PBS for 10 minutes. Sections were then incubated for 15 minutes at room temperature with streptavidin/fluorescein isothiocyanate conjugate (Vector) and 10 µg/ml Hoechst 33342 (Molecular Probes, Eugene, OR), washed three times in PBS for 10 minutes, and mounted with Vectashield medium (Vector). Samples were examined with a Zeiss Axiophot Epifluorescence microscope (Thornwood, NY) equipped with filters to selectively view the rhodamine, fluorescein, and Hoechst 33342 fluorescence. When FAK was immunolocalized in cultured cells, fixation was followed by permeabilization (10 minutes in cold methanol). All other steps were performed using the protocol described for sections.
Several types of control incubations were included. Sera (preimmune or isotype-matched nonimmune) and PBS were substituted for the primary antibodies. In addition, fluorescein isothiocyanate-donkey anti-mouse and fluorescein isothiocyanate-donkey anti-rabbit IgG were substituted for biotin-conjugated reagents to detect binding of the primary antibody. Finally, the effects on blocking and control peptides were assessed as described above. Staining of control tissue sections and cells was never observed.
Cell and Tissue Culture
CTBs were isolated from pools of multiple placentas as described.28 The placentas were obtained immediately after first trimester terminations and third trimester deliveries. Remaining leukocytes were removed by using a mAb to CD45 coupled to magnetic beads. The purified CTBs were cultured in DME H21 minimal essential medium containing 2% Nutridoma (Roche Molecular Biochemicals, Indianapolis, IN) and 50 µg/ml of gentamicin on two substrates. Routinely the cells were plated on Matrigel-coated tissue culture wells (Collaborative Biomedical Products, Bedford, MA). Co-cultures were also established with decidua (n = 4, 10 to 12 weeks of gestation).
Anchoring villi were prepared for culture as previously described.29,30 Briefly, small fragments of placental tissue from the fetal-maternal interface were teased apart until they had the characteristic tree-like appearance of chorionic villi as viewed in a stereomicroscope. Anchoring villi were identified by the attached remnants of cell columns. Light microscopic examination of hematoxylin-stained sections of villus tissue preparations consistently showed the presence of floating and anchoring villi and the absence of endometrial contamination. Anchoring villi (wet weight 5 to 10 mg) were cultured on the surface of Matrigel-coated 12-mm Millicell-CM culture dish inserts (0.4 µm pores; Millipore Corp., Bedford, MA). The inserts were then placed into 24-well dishes in a mixture of F12 HAM/Dulbecco’s modified Eagle’s medium (1:1/v:v) culture medium (Sigma Chemical Co., St. Louis, MO) supplemented with an antibiotic/antimycotic mixture (100 U/ml penicillin, 100 µg/ml streptomycin, 1.25 µg/ml amphotericin B; Sigma) and 10% fetal calf serum. Each group (control and experimental) consisted of five explants. In toto, 12 placentas were individually analyzed.
Isolated CTBs, co-cultures and anchoring villi were maintained either under standard tissue conditions (5% CO2/95% air) or placed in a Bactron anaerobic incubator (Sheldon Manufacturing Inc., Cornelius, OR), where they were maintained in a 2%O2/93% N2/5% CO2 environment. Dissolved O2 at the cell-medium interface, measured using micro-oxygen electrodes (MI-730; Microelectrode Inc., Londonderry, NH), was 20% under standard tissue conditions and 2% in the Bactron anaerobic incubator.
BrdU Incorporation
Anchoring villi from 6- to 8-week-old placentas were cultured on Matrigel in either 20% O2 or 2% O2. After 48 hours, the original medium was replaced with fresh medium containing 1 µmol/L 5-bromo-2'-deoxyuridine (BrdU, Roche Molecular Biochemicals). Twenty-four hours later tissue was washed, fixed, embedded in OCT, and frozen in liquid nitrogen. Sections cut from these blocks were stained with fluorescein isothiocyanate-labeled anti-BrdU monoclonal antibody as recommended by the manufacturer (Roche Molecular Biochemicals).
Adenovirus Construction and Transduction
Adenoviruses were generated by using published
methods.31
The antisense FAK vector (pcDNA3 antisense FAK)
was the kind gift of Drs. Y. Takeuchi and M. Suzawa from Kyoto
University, Japan.32
pEGFP-C1 was purchased from Clontech
(Palo Alto, CA). NdeI-BamHI fragments of pcDNA3
antisense FAK and pEGFP-C1 were inserted into the pAdlox shuttle
vector. The viruses were established by transfecting the ligated
adenoviral genome constructs with replication-defective 
5 virus
into CRE8 cells that stably overexpressed Cre-recombinase. Positive
clones were expanded in HEK293 cells. Adenoviruses that expressed
either a FAK antisense oligonucleotide (AdASFAK) or GFP (AdcGFP) were
plaque purified (1 x 1010
pfu/ml) and
isolated as described.31
For transduction, explants or
cells were infected with wild-type replication-defective 5
adenovirus, AdASFAK, or AdcGFP. Before plating, isolated CTBs and
anchoring villus explants were incubated for 2 hours in medium
containing 20 to 150 pfu/cell of virus. After one washing with PBS, the
cells and villi were cultured as described above. At the end of the
experiment FAK expression was monitored by immunostaining and
immunoblotting.
Invasion Assays
Two assays were used. First, to quantify invasion, 2 x 105 CTBs were cultured in Matrigel-coated Transwell inserts (6.5 mm; Costar Corp.) containing polycarbonate filters with 8-µm pores. After 72 hours the cultures were rinsed in PBS, fixed, and stained with anticytokeratin antibodies as described below. To assay invasion, the filters were dissected from the inserts with a scalpel blade and mounted on poly-L-lysine-coated slides in a drop of mounting medium such that the underside of the filters faced upward. The number of cytokeratin-positive cell processes that penetrated the Matrigel and appeared on the underside of the filters was counted. In each of three independent experiments, three filters were used for each experimental condition. Second, CTBs were cultured with decidual explants.33 Briefly, portions of the decidua parietalis, which does not contain CTBs before culture, were obtained at the time of pregnancy termination (10 to 12 weeks). The tissue was washed three times in ice-cold PBS and cut into 2-mm3 pieces. After another three washes, five decidual pieces in 500 µl of medium (DME/2% Nutridoma) were plated on each of several Millicell inserts (12-mm diameter) coated with 200 µl of Matrigel. The inserts were transferred to 16-mm culture dishes containing 600 µl of medium and incubated for 6 hours, the amount of time required for the explants to attach to the ECM substrate. Then the medium was aspirated from the upper chamber, and 106 CTBs in 500 µl of medium was added. The cultures were maintained for 4 days in either a 2% or a 20% oxygen atmosphere, and medium was changed daily. Three co-cultures were set up for each experimental condition, and the experiments were repeated three times. The samples were fixed in 3% paraformaldehyde for 40 minutes, washed three times in PBS (4°C, 10 minutes), infiltrated with sucrose, embedded in OCT compound, and frozen in liquid nitrogen. Sections (6 µm) were cut using a cryostat (Slee International Inc., Tiverton, RI) and collected on poly-L-lysine-coated microscope slides. The sections were stained as described below. Samples were examined with a Zeiss Axiophot epifluorescence microscope equipped with proper filters and photographed with Kodak T-Max 400 film (Kodak, Rochester, NY).
Western Blots and FAK (Auto)phosphorylation Assay
To examine the efficiency of FAK antisense in reducing levels of the endogenous FAK protein, cells were lysed in modified RIPA buffer [1% sodium deoxycholate, 0.1% sodium dodecyl sulfate (SDS), 1% Nonidet P-40, 150 mmol/L NaCl, 10 mmol/L Tris/HCl, pH 7.4, 1 mmol/L ethylenediaminetetraacetic acid] containing freshly added protease inhibitors (1 µg/ml aprotinin, 1 µg/ml leupeptin, 1 mmol/L phenylmethyl sulfonyl fluoride). Lysates were precleared with protein G-Sepharose (Amersham Pharmacia Biotech, Piscataway, NJ) and separated by SDS-polyacrylamide gel electrophoresis (PAGE) using 8% gels. Gels were transferred to nitrocellulose (Schleicher & Schuell, Keene, NH) and blotted using monoclonal anti-FAK antibody (Transduction Laboratories).
For kinase assays FAK was immunoprecipitated from cells lysed in
modified RIPA buffer using a combination of two goat polyclonal
anti-FAK antibodies, A-17 and C-20 (Santa Cruz Biotechnology).
Immunoprecipitates were washed and then incubated in universal kinase
buffer (10 mmol/L PIPES, pH.7.4, 10 mmol/L MnCl2)
in the presence of [
-32P]ATP for 30 minutes
at 30°C.34
Samples were separated using 8% SDS-PAGE and
transferred to nitrocellulose. The membrane was blotted first with
anti-FAK mAb (Transduction Laboratories). The bands were visualized
using alkaline phosphatase-conjugated secondary antibodies and BCIP/NBT
dye substrate (Roche Molecular Biochemicals). Dried membranes were
exposed to BioMax MS film (Kodak) to detect incorporated radioactivity.
To assess total protein tyrosine phosphorylation, whole cell/Triton soluble lysates were prepared by solubilizing cells in a buffer containing 1% Triton X-100, 150 mmol/L NaCl, 10 mmol/L Tris/HCl, pH 7.4, 1 mmol/L ethylenediaminetetraacetic acid, and freshly added protease inhibitors (1 µg/ml aprotinin, 1 µg/ml leupeptin, 1 mmol/L phenylmethyl sulfonyl fluoride). Lysates were spun at maximum speed in a microfuge for 10 minutes at 4°C. Dissolving the pellets in modified RIPA buffer yielded a whole cell/Triton insoluble/RIPA soluble lysate. Lysates were spun at maximum speed in a microfuge for 10 minutes at 4°C. Dissolving the pellets in SDS-PAGE sample buffer yielded whole cell/Triton-insoluble/RIPA-insoluble lysates.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Initially, we used immunolocalization techniques to evaluate FAK
expression during human CTB invasion in situ. These studies
focused on the fetal-maternal interface, which includes floating villi
together with anchoring villi, their sites of uterine attachment, and
the portion of the uterine wall that CTBs invade. The histological
organization of this region is diagrammed in Figure 1A
. A
photomicrograph of a tissue section stained with anti-cytokeratin
(CK), which labels syncytiotrophoblasts (STBs) and all CTB
subpopulations, is shown in Figure 1B
. In both panels the long arrow
denotes the direction of migration/invasion.
CTBs in all locations showed high levels of FAK immunoreactivity
(Figure 2)
. In mononuclear villus CTBs
(vCTB), FAK localized to the plasma membrane region (Figure 2A
, short
arrows). In multinucleate STBs, staining of the cytoplasm in a punctate
pattern was often detected (Figure 2A
, long arrows). However, the
brightest immunoreactivity was found at the vCTB-STB border, suggesting
that FAK expression is associated with the apical surfaces of the vCTB
plasma membranes, the basal surfaces of STB plasma membranes, or both
areas (Figure 2A
, arrowheads). CK-positive invasive CTBs (iCTB) within
the uterine stroma also stained for FAK in a pattern that was primarily
plasma membrane-associated (Figure 2C
, short arrows). All of the
antibodies against FAK protein that were tested, including those
recognizing epitopes in the C-terminal or N-terminal region, gave
similar staining patterns.
|
. vCTBs and STBs showed little or no
staining with anti-pY397FAK (not shown). Interestingly, this antibody
distinguished two CK-positive iCTB populations within the
uterine wall. Strong staining for pY397FAK was seen in sites of initial
invasion, CTBs near the surface of the uterine wall that borders the
intervillous space, whereas CTBs that had penetrated more deeply
did not stain (Figure 3B)
. In pY397FAK-positive cells,
immunoreactivity was either diffuse throughout the cytoplasm (Figure 3, D–F
; arrowheads) or localized to a very specific region (Figure 3, D
-F; short arrows).
|
Next, we assessed the role of FAK in CTB migration/invasion in vitro. In the absence of a specific inhibitor of FAK (auto)phosphorylation on Y397, we used an adenovirus approach to deliver antisense oligonucleotides that were designed to lower the cells’ expression of FAK protein (AdASFAK). Two in vitro models were used. The first was an organ culture system in which explanted anchoring villi were cultured on Matrigel-coated wells.29,30 In this model, villi attach to the Matrigel via the remnants of severed CTB columns. Subsequently, the area of the column that is in contact with the Matrigel gives rise to invasive CTBs that migrate out of the explant.
Initially, we used an adenovirus that was engineered to encode green
fluorescent protein (AdcGFP) to determine which cells, among the many
types found in the villus explant (see Figure 1
), were infected. GFP
was expressed primarily by CTBs in the cell columns (CC) (Figure 4A)
. This finding is consistent with our
previous observation that CTBs in this location up-regulate expression
of vß3 integrin,6
which is an adenovirus
receptor.35
GFP was expressed throughout the column,
including at the expanded tips that contain outgrowths of
migrating/invading CTBs (Figure 4, B and C
; arrow). In contrast,
explants that were infected with AdASFAK (Figure 4D)
showed a
consistent reduction in CTB migration/invasion from the cell columns
(Figure 4D
, arrowhead). At the end of the experiment, tissue sections
were prepared from control explants that expressed GFP and experimental
explants that expressed antisense FAK. Immunolocalization showed that
cell columns from the experimental explants stained only weakly for FAK
as compared to cells in comparable regions of the control cultures
(compare Figure 4E and 4F
).
|
We used a second tissue culture model to confirm and extend the
results obtained in the villus explant system. When isolated vCTBs are
plated on a Matrigel substrate, they execute the differentiation
pathway that gives rise to iCTBs in vivo, including the
adhesion molecule switching program.1
Typically, this
process takes 48 to 72 hours. Four different preparations of purified
vCTBs were transduced with AdASFAK or control virus. Effects on FAK
protein levels, FAK localization, and CTB invasion were quantified for
each preparation. An immunoblot from one of the preparations comparing
FAK protein levels in lysates of CTBs transduced with AdASFAK or
control virus is shown in Figure 5A
.
After 72 hours in culture, densitometric analyses showed that
transduction with AdASFAK reduced FAK expression by 48% relative to
control CTBs that were either transduced with the parental, wild-type
virus (AdWT) or not exposed to virus. As an additional control, the
lower molecular mass region of the same blot was probed with an
antibody specific for paxillin (<70 kd). Bands of equal intensity
visible in all of the lanes were an independent confirmation that
samples containing equal amounts of protein were loaded. Immunoblot
analysis of two of the other preparations transduced with AdASFAK
showed a reduction in FAK expression of 42 and 54%, respectively,
relative to FAK levels in the control cells. In the fourth experiment,
FAK protein was not detected after transduction with AdASFAK, whereas
strong signals were observed in the control cell lysates.
|
shows the results of the experiment in which transduction
with AdASFAK reduced FAK expression by 48% as assessed by
immunoblotting. Staining with anti-CK was used to image all of the
CTBs. Control cells on the Matrigel substrate were well spread and
extended numerous protrusions. In comparison, CTBs that were transduced
with AdASFAK had a more rounded appearance. When control cells were
stained with an antibody that recognizes all forms of FAK, high levels
of immunoreactivity were detected throughout the cytoplasm and in
association with the plasma membrane. In the AdASFAK-transduced
samples, FAK staining was weak or undetectable in most CTBs.
The remaining CTBs from the four preparations were used to assess the
effects of reducing FAK expression on their ability to invade. The
assay is diagrammed in Figure 5C
. Isolated CTBs are plated on
Matrigel-coated Transwell filter membranes (Figure 5C, I)
. As the cells
differentiate, they invade the Matrigel substrate (Figure 5C, II)
, pass
through the filter pores and emerge on the opposite side of the
membrane (Figure 5C
, II and III). Invasion is quantified after 48 to 72
hours by counting the number of cell processes on the underside of the
filter.
The data from four experiments are summarized in Figure 5D
. In this
analysis, the level of CTB invasion in cultures that were not exposed
to virus was considered to be 100%. Transduction of the cells with
wild-type adenovirus lowered invasion by 50% without affecting the
number of cells that attached, the expression of several
differentiation markers (eg, 5ß1, HLA-G), or the rate of apoptosis
(data not shown). Thus, transduction with adenovirus affected invasion
by an unexplained mechanism, possibly by interfering with normal
function of the Vß3
integrin in adhesion. Nevertheless, transduction of CTBs with AdASFAK
resulted in an additional four-fold reduction in the cells’ ability to
invade, again without affecting attachment, differentiation, or
apoptosis. Finally, we also assayed CTB migration on fibronectin-coated
Transwell filters using methods we described previously.36
The results showed that migration was inhibited to the same extent as
invasion (data not shown).
Conditions that Restricted CTB Invasion in Vitro and in Vivo also Reduced pY397FAK Levels
Because CTBs that penetrate the uterine wall in vivo
must assume an invasive phenotype, we also examined FAK expression in
situations in which their invasion is impaired. First, we investigated
the effects of hypoxia, an important regulator of CTB invasion, on
pY397FAK immunoreactivity. The morphology of control explants that were
cultured for 72 hours under standard conditions (20% oxygen) is
illustrated in Figure 6, A and B
. As we
showed previously, under these conditions CTBs in column remnants exit
the mitotic cycle (Figure 6, C and D)
and switch on the expression of
stage-specific antigens that mediate migration and invasion (data not
shown).37
Here we found that CTBs at the edges of these
explants stained brightly with antibodies that recognize all forms of
FAK (data not shown) and pY397FAK. Particularly strong immunoreactivity
for pY397FAK was noted in association with protrusions at the leading
edge of cells at the periphery of explants (Figure 6, E and F)
. The
morphology of experimental explants that were cultured under hypoxic
conditions (2% oxygen) is illustrated in Figure 6, G and H
. As we
showed previously, under these conditions CTBs proliferate as indicated
by BrdU incorporation (Figure 6, I and J)
.37
Thus, the
columns are larger than in control explants (compare Figure 6, A–D
,
with Figure 6, G–J
). Also in hypoxia, CTB invasion, assayed as
diagrammed in Figure 5C
, is impaired.33,37,38
Here we
found strong FAK staining in CTBs that were cultured in both 2% and
20% oxygen (data not shown). However, staining for pY397FAK was
greatly reduced in CTBs cultured in hypoxia. In addition, the staining
pattern changed. Much of the immunoreactivity was diffuse and
cytoplasmic rather than plasma membrane-associated (compare Figure 6K to 6E
).
|
. To determine whether this
decrease included FAK (125 kd) phosphorylation, we examined levels of
total tyrosine phosphorylation and (auto)phosphorylation in FAK
immunoprecipitates. Interestingly, there was very little difference in
overall tyrosine phosphorylation of FAK, whereas
(auto)phosphorylation as shown by [32P]
incorporation was lower (Figure 7B)
. Consequently, we examined tyrosine
phosphorylation of paxillin, a putative FAK substrate. Lower levels of
paxillin phosphorylation were observed in conditions associated with
lower levels of FAK (auto)phosphorylation. Finally, this did not
change the amount of FAK that co-immunoprecipitated with paxillin.
However, either the amount or the phosphorylation of an unknown protein
that associated with paxillin (asterisk) increased in hypoxia (Figure 7C)
.
|
. iCTBs in all
locations showed strong staining for pY397FAK, primarily in association
with the plasma membrane (Figure 8C)
. Also in accord with our published
work, we found that in hypoxia (2% oxygen) the CTB aggregates, which
attached to the decidua, failed to invade below the surface (Figure 8D)
. Staining with an antibody specific for pY397FAK showed only weak
immunoreactivity in the cytoplasm of CTBs throughout the aggregates
(Figure 8F)
. This difference was noted despite the fact that hypoxia
did not affect the staining patterns of antibodies that recognized all
forms of FAK (data not shown).
|
. In the
control samples, staining for all forms of FAK was detected in
association with the plasma membrane of CTBs found near the surface of
the uterine wall (Figure 9A)
. A very similar staining pattern was
observed in preeclampsia, except that the immunoreactivity was often
more intense than in control samples (Figure 9B)
. In contrast, we noted
dramatic differences between the two groups in the immunolocalization
of pY397FAK. Whereas bright staining was detected in association with
the plasma membranes of CTBs in the control samples (Figure 9E)
, little
or greatly reduced immunoreactivity was detected in CTBs when the
pregnancy was complicated by preeclampsia (Figure 9F)
.
|
. Control CTBs
were isolated from the placenta of a patient who had a clinically
normal pregnancy, and experimental CTBs were isolated from the placenta
of a woman with preeclampsia. Then the cells were cultured under
standard conditions for up to 48 hours. To assess
(auto)phosphorylation, cell lysates were prepared at three time points:
before plating, and after either 24 or 48 hours in culture. Then we
assayed for incorporation of []32P in FAK
immunoprecipitates (see Material and Methods). Afterward the products
were separated by SDS-PAGE and transferred to nitrocellulose.
Immunoblotting with an antibody that recognized all forms of FAK showed
that each of the lysates contained very similar levels of FAK protein
(Figure 10A)
. However, as compared to control cells, FAK
(auto)phosphorylation on Y397 was not detected in CTBs isolated from
the placenta of a patient diagnosed with preeclampsia (Figure 10A)
.
|
. To image the cell processes that make the
initial tracks through the Matrigel, photomicrographs were taken from
the underside of the filter. In control cultures, numerous
cytokeratin-positive cell processes were visible. These stained
brightly for pY397FAK (Figure 10B)
. When the pregnancy was complicated
by preeclampsia, no cell processes had emerged on the underside of the
filter by 72 hours. Instead, anti-cytokeratin staining revealed only
halos of CTBs that remained on the upper, Matrigel-coated surface
(Figure 10B)
. Similar halos of anti-pY397FAK immunoreactivity were not
observed in association with these cells. |
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
. Our previous work showed that
molecules whose expression is up-regulated coincident with uterine
invasion in situ play important functional roles in culture
models that replicate this process in
vitro.28,41,42
Subsequent gene deletion studies have
confirmed that a number of these molecules play important roles
in vivo during murine placentation.43,44 Here we exploited the combination of approaches we have developed to study human placentation to understand the role of FAK during CTB invasion. In situ, FAK was highly expressed, primarily in a plasma membrane-associated pattern, at all stages of CTB differentiation. This observation was somewhat unexpected given our previous work that strongly suggested the importance of integrin function to the cells’ acquisition of a migratory and invasive phenotype.41 This result prompted us to assess the staining pattern of an antibody that recognized only the form of the molecule that is phosphorylated on tyrosine 397, pY397FAK, the primary (auto)phosphorylation site that initiates further FAK activation. Although vCTBs and STBs showed little or no reactivity with this antibody, strong staining was observed in association with a subset of CTBs near the surface of the uterine wall. In vitro function-perturbation experiments showed that reducing FAK expression diminished CTB invasion. The inverse correlation also held: situations in which CTB invasion was down-regulated, such as hypoxia in vitro and preeclampsia in vivo, were associated with a dramatic reduction of pY397FAK levels in the cells especially in membrane-associated areas. Together our results suggest that (auto)phosphorylation of Y397 on FAK, rather than absolute FAK levels, is a critical component of the signaling pathway that mediates migration/invasion.
Our findings have interesting implications that are particularly
relevant to current theories regarding tumor cell invasion, a
prerequisite for metastatic spread, the most common cause of morbidity
and mortality in various neoplastic diseases.45
Invasion,
the movement of cells beyond their normally circumscribed boundaries,
is a complex process that requires coordination of many extracellular
signals, including those generated by soluble factors, as well as
contact with the ECM and other cells. Likewise, intracellular signals,
such as those that lead to assembly and disassembly of the actin
cytoskeleton, are also crucial.46
The strongest evidence
that FAK is a key point of convergence of both ECM- and growth
factor-regulated signals that influence cell migration comes from
studies of FAK-null mouse embryos. Specifically, the cells without FAK
exhibit a reduced migration rate because of impaired focal adhesion
turnover.17,18,22
Conversely, overexpression of FAK leads
to an increase in cell migration.19
In light of the
results of both types of experiments, it is not surprising that a
number of groups report that FAK expression is up-regulated in cancer
cells as compared to their normal counterparts.24,47-51
Our results strongly suggest that (auto)phosphorylation of Y397FAK
could be a critical determinant of the metastatic potential of tumor
cells. This is supported by our analysis of a rare malignant iCTB tumor
(placental site tumor) that we fixed according to the protocol that
allowed immunolocalization of FAK and pY397FAK (Ili and
Genbaev, unpublished data).
Recent work from three laboratories shows that FAK also plays a role in
cell proliferation, suggesting additional tie-ins to other cellular
processes that are critical to tumor progression.52-54
Prompted by these findings, we studied the relationship of FAK
expression to migration versus proliferation in several cell
lines at the population level. Immunoblotting showed that the cells,
including the JAR choriocarcinoma (malignant vCTB) cell line, with the
highest invasive potential and lowest proliferative rate had relatively
lower levels of FAK in toto, but higher levels of pY397FAK.
In comparison, other lines, including JEG (malignant vCTB) cells, with
the lowest invasive potential and highest proliferation rate had higher
levels of total FAK but less FAK (auto)phosphorylated on Y397
(Ili and Genbaev, in preparation). In the tissue sections
and culture models analyzed here, we failed to detect immunostaining
for pY397FAK in proliferating cells—villus and column CTBs in
situ and CTBs in explanted anchoring villi cultured under hypoxic
conditions in vitro. Instead, expression was detected in
migrating CTBs that had withdrawn from the cell cycle.2
Together, our results suggest that it is possible to separate FAK
signals that lead to proliferation from those that promote
migration/invasion.
In conclusion, our studies of human placental CTB invasion both in situ and in vitro suggest that regulation of FAK (auto)phosphorylation, rather than FAK levels, is crucial to the cells’ ability to invade. This finding opens several interesting avenues of investigation, particularly with regard to tumor biology. One area is forecasting the outcome of premalignant conditions. For example, it will be interesting to determine whether increased staining for pY397FAK correlates with specific stages in tumor progression such as initial penetration of the cells’ own basement membrane. Additionally, some chemopreventive drugs may modulate integrin-mediated signals and reduce tyrosine (auto)phosphorylation of FAK,55 either directly or via a myriad of interconnected pathways. In this case evaluation of pY397FAK levels could help determine treatment efficacy. These types of studies will add valuable information to our understanding of the consequences of increases in FAK activity and the relationship of this phenomenon to both normal and abnormal processes that involve the migration and invasion of cells.
|
Acknowledgements |
|---|
|
Footnotes |
|---|
Supported by the University of California at San Francisco Academic Senate and the National Cancer Institute Howard Temin Award (KO1 CA87652–01) (to D. I.); and by National Institutes of Health grants HD 30367, HD 26732, and HL 64597.
D. I. and O. G. contributed equally to this work.
Present address of F. G.: SUGEN, Inc., Oncology Targets, 230 E. Grand Ave., South San Francisco, California 94080.
Accepted for publication April 10, 2001.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
ev O, McMaster MT, Fisher SJ: A repertoire of cell cycle regulators whose expression is coordinated with human cytotrophoblast differentiation. Am J Pathol 2000, 157:1337-1351ovac L, Jones CJ, Aplin JD: Trophoblast differentiation during formation of anchoring villi in a model of the early human placenta in vitro. Placenta 1995, 16:41-56[Medline]
ev O, Dejana E, Wheelock M, Damsky CH: Human cytotrophoblasts adopt a vascular phenotype as they differentiate. A strategy for successful endovascular invasion? J Clin Invest 1997, 99:2139-2151[Medline]
D, Almeida EA, Schlaepfer DD, Dazin P, Aizawa S, Damsky CH: Extracellular matrix survival signals transduced by focal adhesion kinase suppress p53-mediated apoptosis. J Cell Biol 1998, 143:547-560 D, Han Q, Hauck CR, Jin F, Kawakatsu H, Schlaepfer DD, Damsky CH: Matrix survival signaling: from fibronectin via focal adhesion kinase to c-Jun NH(2)-terminal kinase. J Cell Biol 2000, 149:741-754 D, Furuta Y, Kanazawa S, Takeda N, Sobue K, Nakatsuji N, Nomura S, Fujimoto J, Okada M, Yamamoto T, Aizawa S: Reduced cell motility and enhanced focal adhesion contact formation in cells from FAK-deficient mice. Nature 1995, 377:539-544[Medline]
D, Kanazawa S, Furuta Y, Yamamoto T, Aizawa S: Impairment of mobility in endodermal cells by FAK deficiency. Exp Cell Res 1996, 222:298-303[Medline]
vß3 integrin is modulated by a focal adhesion kinase pathway. Cancer Res 1999, 59:1655-1664 D, Klingbeil CK, Schaefer E, Damsky CH, Schlaepfer DD: FAK integrates growth-factor and integrin signals to promote cell migration. Nat Cell Biol 2000, 2:249-256[Medline]
D, Jones KC, Damsky CH, Hunter T, Schlaepfer DD: Pyk2 and Src-family protein-tyrosine kinases compensate for the loss of FAK in fibronectin-stimulated signaling events but Pyk2 does not fully function to enhance FAK-cell migration. EMBO J 1998, 17:5933-5947[Medline]
ev O, Jensen KD, Powlin SS, Miller RK: In vitro differentiation and ultrastructure of human extravillous trophoblast (EVT) cells. Placenta 1993, 14:463-475[Medline]
ev O, Miller RK: Post-implantation differentiation and proliferation of cytotrophoblast cells: in vitro models. Placenta 2000, 21:S45-S49
2ß1 integrin interaction is mediated by focal adhesion kinase and its downstream signals in murine osteoblastic cells. J Biol Chem 1997, 272:29309-29316ev O, Joslin R, Damsky CH, Polliotti BM, Fisher SJ: Hypoxia alters early gestation human cytotrophoblast differentiation/invasion in vitro and models the placental defects that occur in preeclampsia. J Clin Invest 1996, 97:540-550[Medline]
D, Hashiyama M, Noumura T, Yamamoto T, Suda T, Aizawa S: p59fyn-p125FAK cooperation in development of CD4+CD8+ thymocytes. Blood 1996, 87:865-870ev O, Zhou Y, Dong J, Cross JC, Israel MA, Fisher SJ: Id-2 regulates critical aspects of human cytotrophoblast differentiation, invasion and migration. Development 2000, 127:549-558[Abstract]
ev O, Zhou Y, Ludlow JW, Fisher SJ: Regulation of human placental development by oxygen tension. Science 1997, 277:1669-1672ev O, Damsky CH, Fisher SJ: Oxygen regulates human cytotrophoblast differentiation and invasion: implications for endovascular invasion in normal pregnancy and in pre-eclampsia. J Reprod Immunol 1998, 39:197-213[Medline]
