| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Department of Obstetrics and Gynecology, Keio University School of Medicine (T.M., Y.Y.), Tokyo 160-0016, Japan; and the Department of Biological Responses, Institute for Virus Research, Kyoto University (T.M., H.S.), Kyoto 606-8397, Japan
Address all correspondence and requests for reprints to: Dr. Tetsuo Maruyama, Department of Obstetrics and Gynecology, Keio University School of Medicine, 35 Shinanomachi, Shinjuku-ku, Tokyo 160-0016, Japan. Fax: 81-3-3226-1667.
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
ß
heterodimeric transmembrane glycoprotein receptors, are one of the
structural components of focal adhesions that are sites of tight
adhesion to the underlying extracellular matrix (ECM) (1, 2). The
aggregation of integrins upon cell to ECM attachment activates the
focal adhesion kinase (FAK) and leads to the assembly of a
multicomponent signaling complex including vinculin, -actinin,
paxillin, cortactin, talin, and tensin (2, 3, 4, 5). Src family tyrosine
kinases such as c-Src and Fyn, which are also included in the focal
adhesion complex, participate in this integrin-mediated signal
transduction. These kinases modulate the activity and recruitment of
the focal adhesion proteins and eventually affect the actin-based
cytoskeletal organization (2, 3, 4, 5). Alterations in integrin-mediated cell
adhesion to the ECM are essential regulatory processes during
development, cell growth and differentiation, migration, apoptosis, and
oncogenesis (1). Recently, integrins have been implicated in a variety of biological processes in reproduction (6, 7, 8). Various types of integrins are temporally and spatially expressed in the endometrium during the implantation window (9, 10) and in the embryo and the surrounding trophoblast (11). Additionally, the inhibitory effects of antiintegrin antibodies on embryo attachment and outgrowth onto the endometrium (12, 13) indicate a pivotal role of integrins in implantation. Also, integrins may be involved in endometrial tissue remodeling by regulating the integrity of cell to cell and/or cell to ECM interaction (7, 8). For example, integrin ß1 is preferentially expressed in the pregnancy decidua in vivo, and its expression level is increased during in vitro decidualization of endometrial stromal cells (ESC) (8, 10, 14), suggesting the importance of integrin ß1 in decidual transformation. However, the role of integrins and focal adhesion proteins in decidualization remains unclear.
We previously reported that the kinase activity of c-Src is elevated and accompanied by changes in tyrosine phosphorylation of several cellular proteins during in vitro decidualization (15). Focal adhesion proteins, including FAK and paxillin, are deeply involved in the integrin-mediated signal transduction through focal adhesion assembly, serving as substrates for the Src family tyrosine kinases (2, 3, 4, 5). This knowledge prompted us to determine whether activation of an integrin-mediated signaling pathway(s) using FAK and paxillin is a downstream event of the decidual c-Src activation. Endometrial cells differentially express many types of integrins in the process of decidualization (8, 10, 14). Moreover, human and rodent ESC up- and down-regulate various types of ECM during in vivo and in vitro decidualization (16, 17, 18, 19, 20), thereby changing the composition of cell to cell and cell to ECM adherence. Taken together, it is likely that alterations in ECM components may activate decidualization-specific signaling cascades involving c-Src and focal adhesion proteins through ECM/integrin interaction, ultimately leading to the morphological and functional decidual transformation. Therefore, to investigate the c-Src-mediated signaling pathway responsible for decidualization, we here focused on analyzing tyrosine phosphorylation and subcellular localization of focal adhesion proteins such as paxillin and FAK.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Isolation of ESC and hormonal treatment
Human endometria were obtained as described previously (15, 22).
Three additional endometria obtained from premenopausal women without
any history of gynecological malignancy were subjected to FAK and
paxillin immunoprecipitation and immunocytochemical studies. Informed
consent was obtained from all subjects.
Isolation and hormonal treatment of ESC were performed as previously described (15, 22). In brief, isolated ESC were cultured and grown to subconfluence in the basal medium: phenol red-free RPMI 1640 supplemented with 10% charcoal-stripped FCS and 1% antibiotic-antimycotic mixture (Life Technologies, Inc., Grand Island, NY). After 2 days of preculture, the medium was replaced with fresh basal medium containing 10 nM 17ß-estradiol (E2), 1 µM progesterone (P), 10 nM E2 plus 1 µM P, or 0.1% ethanol as control vehicles. The culture medium containing the hormones was renewed every 2 days. For hormone withdrawal experiments, ESC that had been treated with E2 and P for 14 days were washed and cultured with or without E2 and P for an additional 4 days.
Immunoprecipitation and immunoblotting with antiphosphotyrosine,
paxillin, and FAK antibodies
Cell lysates were prepared with radioimmunoprecipitation
buffer as previously described (15, 22). The protein concentration of
the samples was measured by the DC protein assay kit (Bio-Rad Laboratories, Inc., Cambridge, MA). Immunoprecipitation and
Western immunoblotting were performed as previously described (15). In
brief, 100–200 µg cell lysates were immunoprecipitated by incubation
with paxillin or FAK antibody for 120 min at 4 C, followed by addition
of 20 µl protein G-Sepharose beads (Sigma Chemical Co.)
for 60 min. Immune complexes were washed three times with
radioimmunoprecipitation buffer and then resuspended in 2
x SDS sample buffer. The samples were subjected to SDS-PAGE,
transferred to Immobilon (Millipore Corp., Bedford, MA),
and detected by incubating the blots with the specified antibody.
Immunoblots were incubated with either goat antirabbit or antimouse
horseradish peroxidase-conjugated antibodies and developed using an
enhanced chemiluminescence kit according to the manufacturer’s
procedure (Amersham Pharmacia Biotech, Arlington Heights,
IL). When indicated, immunoblots were stripped in the buffer [62.5
mM Tris (pH 6.8), 2% SDS, and 100 mM
ß-mercaptoethanol] at 50 C for 30 min and reprobed with another
specific antibody.
Immunofluorescence and confocal laser microscopy
The isolated ESC, obtained as previously described (15, 22),
were seeded and grown on plastic chamber slides (Permanox, Lab-Tek
Chamber Slide, Nunc, Roskilde, Denmark) to subconfluence. Subsequently,
the cells were treated with or without E2 and/or
P for 7 days. Cells were then fixed in 3.7% formaldehyde in PBS for 20
min at room temperature and permeablized with 0.5% Triton X-100 in PBS
at room temperature. After they were blocked by 10% FCS in PBS for 30
min at room temperature, slides were incubated with either antipaxillin
(1:250) or antivinculin antibody (1:400) for 90 min at 37 C. After
being washed three times in PBS, bound antibodies were visualized using
fluorescein-conjugated antibody for 30 min at 37 C in a moist
chamber. For double staining of filamentous actin (F-actin) and either
paxillin or vinculin, fluorescein-conjugated phalloidin was included
with a secondary rhodamine-conjugated antimouse antibody. The slides
were washed extensively in PBS and mounted. The confocal images as
shown in Fig. 4
were collected with a laser scanning microscope system
(LSM-GB, Olympus Corp., Tokyo, Japan), as described
previously (23, 24), under the following conditions: object lens,
Olympus Corp. Splan-Apo x60 (numerical aperture =
1.4); zoom ratio, x 2.0. The other confocal images as shown in Fig. 5
were taken using the MRC-600 Laser Scanning Confocal Imaging System
(Bio-Rad Laboratories, Inc., Cambridge, MA) connected to a
Carl Zeiss Axiplan compound microscope (New York, NY).
|
|
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
, the whole pattern of tyrosine
phosphorylation of cellular proteins was changed during
E2- and P-induced decidualization as reported
previously (15). A thick approximately 60-kDa band was clearly observed
in ESC treated with control vehicles throughout the culture period
(Fig. 1A, upper panel, lanes 1, 3, and 13, white
arrowheads), whereas the intensity of this band was decreased as
ESC underwent decidualization (Fig. 1A, upper panel, lanes
2, 4, and 15). In contrast, an approximately 56-kDa band whose tyrosine
phosphorylation was not detected in undifferentiated ESC (Fig. 1A, upper panel, lanes 1, 3, and 13), became phosphorylated
during decidualization (upper panel, lanes 2, 4, and 15,
gray arrowheads). It remains to be solved whether each band
of several tyrosine-phosphorylated proteins, including the
approximately 60- and 56-kDa proteins represents a single protein.
However, we previously provided strong evidence suggesting that the
dephosphorylated, approximately 60-kDa band is c-Src (15).
|
, upper panel, lanes 9–12 and 19–21, black
arrows). The intensity of the phosphorylation signal appeared to
correlate with the protein amount of the corresponding
immunoprecipitate (Fig. 1A, lanes 9–12 and 19–21 in upper panel
vs. middle panel). The level of FAK expression was
clearly constant throughout the culture (Fig. 1A, middle
panel, lanes 1–4 and 13–15).
Similarly, the phosphorylation level of the 68-kDa paxillin
immunoprecipitate was almost constant throughout the culture period
(Fig. 1A, upper panel, lanes 5–8 and 16–18, gray
arrows). Although the level of paxillin expression appeared to
slightly increase during decidualization (Fig. 1A, lower
panel, compare lane 1 with 2 and lane 13 with 15), no significant
induction of paxillin was observed in the other two independent sets of
similar experiments (Fig. 1B, lower panel, lanes 9–12; data
not shown). In the lower panel of Fig. 1A, the signals of
some paxillin immunoprecipitates were weak (lanes 7, 8, and 18), which
might be caused by partial disruption of the paxillin blots from the
membrane due to the repeated stripping and reprobing procedure. To
further investigate the phosphorylation level of paxillin in hormonally
treated ESC, the other independent set of similar experiments was
performed as shown in Fig. 1B. These results clearly demonstrated that
the phosphorylation levels of the paxillin immunoprecipitates (Fig. 1B, upper panel, lanes 1–8 and 13–16, gray arrows)
and the protein amount of the immunoprecipitable paxillin (lower
panel, lanes 1–8 and 13–16) were not markedly changed upon
hormonal treatment and thereafter during decidualization.
E2 alone did not provoke any change in the whole
pattern of tyrosine phosphorylation or in the phosphorylation levels of
FAK and paxillin (Fig. 1A, lanes 14, 17, and 20; Fig. 1B, lanes 2, 6,
and 14), compared with the control treatment (Fig. 1A, lanes 13, 16,
and 19; Fig. 1B, lanes 1, 5, and 13).
No marked alterations in tyrosine phosphorylation levels of
paxillin and FAK upon steroid withdrawal
We previously reported that withdrawal of both estrogen and
progesterone from cultures of decidualized ESC reduces c-Src kinase
activity to the basal level, concomitantly reversing the tyrosine
phosphorylation pattern to the unstimulated state (15). As FAK and
paxillin are downstream targets of c-Src (2, 3, 4, 5), we further determined
whether tyrosine phosphorylation of FAK and paxillin was changed in
response to steroid withdrawal. As shown in Fig. 2, phosphorylated, approximately 64- and
56-kDa bands disappeared upon steroid withdrawal (upper
panel, lanes 1 and 2, gray arrowheads), whereas the
approximately 60-kDa band became phosphorylated (upper
panel, lanes 1 and 2, white arrowheads). These changes
in the whole pattern of tyrosine phosphorylation are consistent with
the previous results (15). Although the phosphorylation signals of FAK
and paxillin immunoprecipitates seemed to become weak upon steroid
withdrawal (Fig. 2, upper panel, lanes 3–6), the decrease
in the phosphorylation level correlated with that in the protein amount
of the corresponding immunoprecipitates (middle panel, lanes
5 and 6; lower panel, lanes 3 and 4). Thus, steroid
withdrawal also did not markedly affect the tyrosine phosphorylation
status of FAK and paxillin.
|
).
In contrast, E2- plus P-treated ESC became
rounded and densely packed, and gave rise to multicellular nodules
(Fig. 3B). As shown in the phase contrast micrographs under the lower
magnification (Fig. 3, C and D), numerous nodules were dominant in
cultures of decidualized ESC (Fig. 3D), whereas no apparent nodules
were observed in the control cultures (Fig. 3C).
|
and 5A, two right panels;
and Fig. 5B, three right panels). Although
E2 or P alone induced cell aggregation to some
extent, accumulation of paxillin in the focal adhesion plaques was not
decreased (Fig. 4, B and C, respectively). The occasional monolayer
populations of E2- and P-treated ESC, located
between the multicellular nodules, also exhibited focal adhesions
containing paxillin and vinculin (data not shown). However, in the
multicellular nodules no discrete accumulations of paxillin staining
could be detected at borders of adjoining cells or at distinct focal
adhesion plaques; rather, paxillin was diffusely localized in the
cytosol (Fig. 4E). However, in contrast, paxillin was localized in the
focal contacts in the substratum of those nodules (Fig. 4D).
Disorganization of actin-based cytoskeleton in the
decidualization-induced multicellular nodules
The orderly recruitment and complex formation of focal adhesion
proteins are crucial for actin-based cytoskeletal organization (2, 3, 4, 5).
We further investigated the subcellular distribution of actin stress
fibers in the multicellular nodules where the formation of the focal
adhesion assembly was impaired. Double labeled immunostaining of ESC
using FITC-conjugated phalloidin and antipaxillin antibody, visualized
by rhodamine-conjugated secondary antibody, showed that the
undifferentiated and flattened ESC in the control cultures clearly
possessed focal adhesion contacts containing paxillin (Fig. 5A, two right panels, yellow signals). Moreover, there were well
stretched actin stress fibers (Fig. 5A, two right panels, green
signals) linking paired focal adhesion contacts (white
arrows). However, neither the typical subcellular distribution of
paxillin nor F-actin connecting focal adhesion plaques could be
detected within the multicellular nodules (Fig. 5A, left
upper and middle panels, respectively). The merged
image showed considerable yellow signals mainly at borders of adjoining
cells (Fig. 5A, lower left panel). Although it may suggest
the codistribution of some populations of cytosolic paxillin with
F-actin, it remains to be clarified whether the merged signals were due
to a real colocalization or a to simple overlapping of both signals at
the cell periphery. Similarly, in the undifferentiated ESC, vinculin
staining was observed at the focal contacts (Fig. 5B, upper right
panel) with actin stress fibers (middle right panel)
connecting vinculin-containing focal adhesion plaques (lower
right panel, yellow signals). In the decidualization-induced
multicellular nodules, however, vinculin did not localize at distinct
focal contacts (Fig. 5B, upper left panel), but was
diffusely localized in the cytosol (lower left panel, red
signals) and also localized at the cell periphery along with
F-actin (lower left panel, yellow signals).
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
In our culture system, the levels of FAK and paxillin phosphorylation appeared to be constant during decidualization and thereafter upon steroid withdrawal. Given our previous results showing activation of c-Src during in vitro decidualization and its inactivation upon steroid withdrawal (15), our present findings suggest that tyrosine phosphorylation of FAK and paxillin may not be related to the kinase activity of c-Src in decidualized ESC. This is in agreement with the report by Schlaepfer et al., who demonstrated that the activation of c-Src can be dissociated from FAK activation (25) and that there are at least two separate downstream pathways (25). This model may account for the c-Src-independent phosphorylation of the endometrial FAK and paxillin as presented here. In addition, c-Src interaction with FAK is known to be regulated by cell adhesion (5). Given the observed breakdown of focal adhesion contacts in decidualized ESC, it seems that dissociation of the c-Src/FAK focal adhesion complex may take place during decidualization. Thus, FAK and paxillin are no longer regulated by activated c-Src.
In contrast to the c-Src-FAK interaction, the paxillin-FAK interaction has been reported to be constitutive and unaffected by integrin-ECM interaction, i.e. the formation of focal adhesions (26). FAK is believed to be a point of convergence in the actions of integrins and growth factors (27, 28). In fact, FAK is not only activated by integrins, but also by mitogenic neuropeptides (29), thrombin (30), and ligands for tyrosine kinase receptors, such as platelet-derived growth factor and hepatocyte growth factor (31, 32, 33). As ESC produce a large number of growth factors during decidualization (34, 35), it is possible that tyrosine phosphorylation of the decidual FAK and paxillin may be regulated mainly by decidualization-associated growth factors rather than by integrin/c-Src-mediated signaling pathways. In this regard, decidualization might provoke a stepwise transition from integrin mediated- to growth factor-dependent signaling pathway(s) upstream of FAK and paxillin. This hypothesis may account for the constant levels of FAK and paxillin phosphorylation during decidualization despite the disruption of focal adhesions.
We have shown for the first time that there is a disorganization of the actin-based cytoskeleton and a loss of focal adhesions within the decidualization-associated multicellular nodules. The mechanism for this disruption is unknown. In fibroblasts, activation of the cAMP/PKA-mediated signaling pathway is known to disrupt focal adhesion plaques as well as the actin-based cytoskeleton through down-regulation of myosin light chain kinase (2, 36). Given that activation of the cAMP/PKA pathway is one of the key events associated with decidualization (37, 38, 39), this pathway may contribute to the decidualization-induced disruption of the cytoskeleton. Carter et al. reported that F-actin is reversibly disrupted in human ESC expressing the temperature-sensitive simian virus 40 large T antigen (40). Moreover, coexpression together with EJ ras oncogene results in the localized reorganization of stress fibers near the cell periphery (40). This pattern of subcellular F-actin distribution appears to be similar, as observed here within decidualization-induced multicellular nodules, in that F-actin was concentrated at the cell periphery. Similar localization of actin microfilaments at the cell periphery has been reported in a mouse in vitro decidualization model (20). It is, therefore, of great interest to determine whether the process of decidualization may involve Ras or Ras-related proteins such as Rho (2), a small GTP-binding protein potentially regulating the focal adhesion assembly and organization of the actin-based cytoskeleton (2).
Recently, Kim et al. reported that cytochalasin D-induced disruption of the actin-based cytoskeleton results in increased expression of insulin-like growth factor-binding protein-1, a typical decidualization-associated marker (41). They proposed the hypothesis that alterations in the cytoskeleton are deeply involved in insulin-like growth factor-binding protein-1 regulation (41). Although treatment of ESC with cytochalasin D is artificial, our data support the idea, in that breakdown of the cytoskeleton occurring during in vitro decidualization together with the loss of focal contacts may enhance the production of many bioactive substances by ESC. The coincidence of the production of decidualization-associated markers and the initiation of morphological decidual changes (15, 22, 42) also suggests a possible involvement of the cytoskeletal organization in decidual function.
The present study is the first to follow the tyrosine phosphorylation and subcellular distribution of focal adhesion proteins in hormonally treated human ESC in vitro. Together with our previous result indicating the elevation of c-Src kinase activity during in vitro decidualization (15), our present data suggest that c-Src may activate a different signaling pathway(s), which does not involve FAK and paxillin, in decidualized ESC. Furthermore, it is possible that dissociation of the focal adhesion complex together with breakdown of the actin-based cytoskeleton might actively contribute to both functional and morphological decidualization. In this regard, it is conceivable that cytosolic free forms of focal adhesion proteins may play a functional role in decidual transformation. Thus, further investigation focusing on the endometrial integrins, focal adhesion proteins, Src family tyrosine kinases, and actin-based cytoskeleton will help us understand the molecular mechanism underlying decidualization in vivo and in vitro.\.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Present address: Department of Molecular Biology (H.S.), Osaka
Bioscience Institute, Furuedai, Suita, Osaka 565-0874, Japan.
Received May 3, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  T. Nagashima, T. Maruyama, H. Uchida, T. Kajitani, T. Arase, M. Ono, H. Oda, M. Kagami, H. Masuda, S. Nishikawa, et al. Activation of SRC Kinase and Phosphorylation of Signal Transducer and Activator of Transcription-5 Are Required for Decidual Transformation of Human Endometrial Stromal Cells Endocrinology, March 1, 2008; 149(3): 1227 - 1234. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
I. Ihnatovych, W. Hu, J. L. Martin, A. T. Fazleabas, P. de Lanerolle, and Z. Strakova Increased Phosphorylation of Myosin Light Chain Prevents in Vitro Decidualization Endocrinology, July 1, 2007; 148(7): 3176 - 3184. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. Sakai, T. Maruyama, R. Sakurai, H. Masuda, Y. Yamamoto, A. Shimizu, I. Kishi, H. Asada, S. Yamagoe, and Y. Yoshimura Involvement of Histone Acetylation in Ovarian Steroid-induced Decidualization of Human Endometrial Stromal Cells J. Biol. Chem., May 2, 2003; 278(19): 16675 - 16682. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y. Yamamoto, T. Maruyama, N. Sakai, R. Sakurai, A. Shimizu, T. Hamatani, H. Masuda, H. Uchida, H. Sabe, and Y. Yoshimura Expression and subcellular distribution of the active form of c-Src tyrosine kinase in differentiating human endometrial stromal cells Mol. Hum. Reprod., December 1, 2002; 8(12): 1117 - 1124. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |
