| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Departments of Physiology and Obstetrics and Gynaecology, University of Western Ontario, London, Ontario, Canada N6A 5C1
Address all correspondence and requests for reprints to: Dr. T. G. Kennedy, Department of Physiology, University of Western Ontario, London, Ontario, Canada N6A 5C1. E-mail: tkennedy{at}physiology.uwo.ca
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
During decidualization, the extracellular matrix (ECM) of the endometrium undergoes extensive remodeling, whereby the fibrillar collagen underlying the undifferentiated stromal cells becomes less fibrous due to the removal of several collagen components from the matrix (8). Subsequently, the newly differentiated decidual cells produce several basement membrane components, including laminin, entactin, and collagen type IV (9, 10). An identical change in endometrial matrix composition occurs during artificially induced decidualization in the absence of an embryo, suggesting maternal regulation of ECM remodeling (1, 8).
The process of ECM remodeling involves a balance between the activities of matrix metalloproteinases (MMPs), enzymes that degrade components of the ECM, and their natural inhibitors, the tissue inhibitors of metalloproteinases (TIMPs) (11, 12). Members of the MMP family include collagenases, gelatinases, stromelysins, metalloelastases, and the membrane-type MMPs, which together are able to degrade the numerous components of the ECM. There are currently four known TIMPs, TIMP-1, -2, -3, and -4, which exert their inhibitory action by binding to the active site of the MMPs.
ECM remodeling occurs within several physiological and pathological systems (11), and for several of these systems in vitro experiments suggest that production of MMPs is regulated by growth factors. EGF increases the production of collagenases and stromelysins in cultured human fibroblasts (13), and bFGF increases collagenase-3 and gelatinase B production within sarcoma cells (14, 15). TGFß promotes stability of the ECM by decreasing the production of proteinases and increasing production of TIMPs (16).
We have developed an in vitro model of decidualization (17) in which stromal cells are isolated from the rat uterus and allowed to undergo decidualization in culture. Using this model we have shown that these cells produce gelatinases A and B and TIMP-1, -2, and -3 and that the production of gelatinase A and TIMP-1 is regulated by PGE2 (18). In the present study we continued our work with this model to test whether isolated stromal cells produce additional MMPs and whether the expression of MMPs and TIMPs is regulated by specific growth factors.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-32P]deoxy-CTP were purchased from
Amersham Pharmacia Biotech (Baie d’Urfé, Canada).
Triton X-100 was purchased from BDH Laboratories (Toronto, Canada).
Twenty-four-well plates were purchased from Becton Dickinson and Co. (Lincoln Park, NJ).
Animals
All animals were handled in accordance with the guidelines of
the Canadian Council on Animal Care and the University Council on
Animal Care at the University of Western Ontario. Female Sprague Dawley
rats (200–225 g BW; Harlan Sprague Dawley, Inc.,
Indianapolis, IN) were housed in temperature- and light-controlled
conditions (lights on from 0500–1900 h), with free access to food and
water. Two days after arrival, animals were ovariectomized under ether
anesthesia and allowed at least 4 days to recover. To obtain rats with
uteri sensitized for decidualization, estradiol and progesterone were
administered s.c. in sesame oil as described previously (17).
Endometrial stromal cell isolation
Rats with uteri sensitized to undergo the decidual cell reaction
(the equivalent of day 5 of pseudopregnancy) were killed by
decapitation, and endometrial stromal cells were obtained from uterine
horns by enzymatic dispersion as described in detail previously (17).
Stromal cells were suspended in DMEM:F12 containing 10%
heat-inactivated charcoal-stripped FCS, penicillin (50 IU/ml),
streptomycin (50 µg/ml), and fungizone (1.25 µg/ml). The cell
suspension was filtered through a nylon mesh (70 µm pore size) to
remove glands and clumps of epithelial cells and plated at 5 x
105 cells/well in 500 µl DMEM:F12 with 10% FCS
in 24-well plates. After 2-h incubation at 37 C under 5%
CO2-95% air to allow for differential attachment
of stromal cells, the medium and nonattached cells were removed and
replaced with serum-free DMEM:F12; this time period was designated
0 h. All experiments were conducted in the absence of serum.
Endometrial stromal cells were cultured for 24 h in the absence of any growth factor, after which they were cultured for 48 h in the either the absence or presence of a single growth factor, with medium changed at 24 h. For concentration-response experiments, concentrations of growth factors were based on previous studies from our laboratory and others: EGF, 10–80 ng/ml (19); bFGF, 12.5–100 ng/ml (6); and TGFß, 0.25–2 ng/ml (20). After 72 h of culture, the media were removed for substrate-gel electrophoresis, and stromal cells and ECM were suspended in 200 µl 0.25% deoxycholate, pH 8.0, for protein assay.
For RNA analysis, a single concentration of each factor was selected: EGF, 40 ng/ml; bFGF, 50 ng/ml; and TGFß, 1 ng/ml. After 72 h of culture, media were removed, and cells were harvested in guanidine thiocyanate buffer (21).
Protein assay
The total protein content within the stromal cells and ECM was
determined using the Bio-Rad Laboratories, Inc. DC Protein
Assay kit, with BSA as the standard. Fifty microliters of sample
suspended in deoxycholate were used in the assay.
Substrate-gel electrophoresis
To detect proteolytic activity within the conditioned media
(CM), substrate-gel electrophoresis was used, with gelatin as the
substrate to detect both gelatinase (22) and collagenase (23) activity,
and casein as the substrate to detect stromelysin activity (24). CM (15
µl) and mass marker (10 µl) were diluted with 4 x sample
buffer (8 mM urea, 8% SDS in 0.5 M Tris, pH
6.8 containing bromophenol blue without 2-mercaptoethanol) and
subjected to electrophoresis through a 10% polyacrylamide gel
containing either gelatin (60 µg/ml) or ß-casein (500 µg/ml).
After electrophoresis, gelatin gels were washed twice in 2.5% Triton
X-100 for 15 min, whereas casein gels were washed twice in 50
mM Tris-HCl, pH 7.5, and 2.5% Triton X-100. Gelatin gels
were incubated for 24 h at 37 C in 50 mM Tris, 0.2
M NaCl, 3 mM CaCl2, 0.5
mg/ml Brij-35, and 0.2 mg/ml NaN3, pH 7.2; to
detect collagenase activity, gels were incubated for 48 h. Casein
gels were incubated for 48 h at 37 C in 50 mM Tris,
150 mM NaCl, 10 mM CaCl2,
1 µM ZnCl2, 0.1% Triton X-100, and
0.2 mg/ml NaN3, pH 7.6. Gels were then stained
with 0.1% Brilliant Blue R for 10 min, followed by destaining. Clear
bands indicated proteolytic activity.
Northern blot analyses
To elucidate the identity of MMPs that are affected by growth
factor treatment, we used Northern blot analysis to detect changes in
mRNA levels. Total RNA was isolated by a single step guanidium
thiocyanate method (21) as described previously (18). Total RNA (10
µg) was denatured and subjected to electrophoresis in a denaturing
gel and then blotted by capillary transfer onto Hybond-N membranes
(25). RNA was then cross-linked to the membrane by exposure to 1.2
x 105 µJ/cm2 UV energy
on a cross-linker (Hoefer, San Francisco, CA).
Candidate MMPs to be examined were selected based on the size of
clearing and the specific substrate used in the substrate-gel
electrophoresis. Mouse complementary DNA (cDNA) probes for gelatinase A
(26), collagenase-3 (27), stromelysin-1 (28), and TIMP-1 (29) were
435-, 2700-, 600-, and 800-bp fragments, respectively, from plasmids
provided by Dr. R. Khokha. Mouse cDNA probes for gelatinase B (30) and
TIMP-3 (31) were 3000- and 1800-bp fragments, respectively, from
plasmids provided by Dr. D. Edwards. The mouse cDNA for TIMP-2 (32) was
a 700-bp fragment of a plasmid provided by Dr. X. Zhang. cDNA probes
(25 ng) were labeled by the random priming technique in the presence of
[-32P]deoxy-CTP using the Random Primers
Labeling System. Northern blot analyses were performed as described by
Church and Gilbert (33) with some modifications. Briefly, membranes
were prehybridized in Church buffer [7% SDS, 0.25 M
Na2HPO4 (pH 7.2), 1
mM EDTA, and 1% BSA] at 60 C for 30 min. Hybridizations
were carried out at 65 C for approximately 20 h. Membranes were
then washed three times (15 min each) in 20 mM
Na2HPO4 (pH 7.2) with 4%
SDS at 60 C and subjected to autoradiography at -70 C with
intensifying screens. Between hybridizations, the blots were stripped
in 1 mM Tris, 1 mM EDTA, 0.1 x
Denhardt’s reagent (1 x Denhardt’s = 2% BSA, 2% Ficoll,
and 2% polyvinylpyrrolidone, pH 8.0) for 2 h at 75 C. Finally,
blots were probed with a radiolabeled cDNA for 18S ribosomal RNA to
determine the relative amounts of RNA loaded into each lane and
transferred onto the membrane (34).
Statistical analysis
All experiments were performed three times with different
endometrial stromal cell preparations. The areas of clearing from the
zymograms and the relative intensities of the mRNA signals on the
autoradiograms were quantified by densitometry (Image Master VDS,
Pharmacia Biotech, Piscataway, NJ). Concentration-response
experiments were analyzed on the raw data by ANOVA using orthogonal
polynomial comparisons. To determine which concentrations had a
significant effect on substrate degradation, ANOVA within blocks was
performed, with individual cell preparations being considered
blocks, followed by Duncan’s multiple range comparison test. For the
Northern blot experiments, ratios of target mRNA/18S ribosomal RNA were
determined, and these ratios were analyzed by ANOVA within blocks, with
individual cell preparations being considered blocks. P
< 0.05 was considered significant.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Substrate gel electrophoresis
To determine whether the growth factors affected the levels of
proteolytic enzymes, CM from the endometrial stromal cell cultures was
analyzed by substrate gel electrophoresis. With gelatin as a substrate
and incubating the gels for 24 h, we were able to detect several
bands of proteolytic activity, with major bands occurring at
approximately Mr of 92K and 72K (Fig. 1
). For the 92K Mr
bands, ANOVA of the areas of clearing revealed for both EGF and bFGF
that the effect of concentration (linear) was significant
(P < 0.05), indicating that both growth factors caused
a log2 concentration-dependent linear increase in
the areas. Multiple comparisons revealed that there was a significant
increase in the area of clearing for CM from cells cultured in the
presence of 40 ng/ml (P < 0.05) and 80 ng/ml
(P < 0.01) EGF and 50 ng/ml (P <
0.01) and 100 ng/ml (P < 0.01) bFGF compared with that
from cells cultured in the absence of a growth factor. Neither EGF nor
bFGF had a significant effect on the cleared area of the 72K
Mr band. TGFß did not have a significant effect
on the cleared area of either the 92K or 72K Mr
bands (data not shown).
|
). For CM
from EGF-treated cells, ANOVA revealed significant (P
< 0.05) effects of concentration (linear) and concentration
(quadratic), indicating that EGF caused a log2
increase in the area of clearing. Multiple comparisons revealed that
the areas of clearing for CM from cells cultured in the presence of 20,
40, and 80 ng/ml EGF were significantly larger (P <
0.01) than that from cells cultured in the absence of EGF. As indicated
by a significant (P < 0.05) concentration (linear)
effect from ANOVA, bFGF caused a log2
concentration-dependent linear increase in the area of clearing of the
54K Mr band. Multiple comparisons revealed that
the areas of clearing for CM from cells cultured in the presence of 50
and 100 ng/ml bFGF were significantly larger (P <
0.01) than that from cells cultured in the absence of bFGF. TGFß did
not have a significant effect on the area of clearing (data not
shown).
|
). For the upper band, bFGF caused a
log2 concentration-dependent linear increase in
the area of clearing, as indicated by a significant (P
< 0.05) concentration (linear) effect on ANOVA. Multiple comparisons
revealed that concentrations of 12.5, 25, 50, and 100 ng/ml bFGF all
resulted in significantly larger (P < 0.01) areas of
clearing than that with no bFGF. ANOVA using orthogonal polynomials did
not reveal a significant relationship for the effect of EGF on casein
proteolysis. However, when differences between blocks were removed by
ANOVA within blocks followed by multiple comparisons, the areas of
clearing from cells cultured in the presence of 10, 20, 40, or 80 mg/ml
were all significantly (P < 0.01) greater than that
from cells cultured in the absence of EGF. TGFß had no significant
effect on the area of clearing (data not shown).
|
, A and B). The densitometric analyses
are shown in Fig. 4, C and D. ANOVA revealed that EGF (40 ng/ml)
significantly increased the mRNA levels for collagenase-3 (32-fold;
P < 0.01), stromelysin-1 (22-fold; P
< 0.05), and gelatinase B (29-fold; P < 0.01), but
had no effect on the mRNA levels of gelatinase A. ANOVA also revealed
that bFGF (50 ng/ml) significantly increased the mRNA levels for
collagenase-3 (5.3-fold; P < 0.01) and stromelysin-1
(11-fold; P < 0.01), but had no significant effect on
the level of either gelatinase A or B. In addition, TGFß (1 ng/ml)
had no significant effect on any of the four MMPs examined.
|
, A and B). The densitometric analyses
are shown in Fig. 5, C and D. Neither EGF, bFGF, nor TGFß had a
significant effect on transcript levels for any of the TIMPs.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The family of matrix metalloproteinases is able to degrade the numerous ECM substrates. Gelatinases act primarily on denatured collagen, collagen types IV and V, and fibronectin; the interstitial collagenases degrade fibrillar collagen types I, II, and III; and the stromelysins act on laminin, fibronectin, and proteoglycans (11). The present experiments demonstrate that endometrial stromal cells undergoing decidualization in vitro produce several MMPs, suggesting that these cells are capable degrading the ECM during decidualization in vivo.
Activity for the gelatinases was detected using gelatin as a substrate for PAGE, and areas of clearing were observed at the proper sizes for gelatinase A and B (11). When the incubation time of the gelatin-containing gels was doubled, we observed enhanced clearing of the 66–92K Mr bands, presumably because of the additional time during which the enzymes were activated. This longer time also resulted in the detection of an additional band at 54K Mr, the size of collagenase-3 (11). This MMP has less cleaving activity for gelatin than the gelatinases, which is probably why a longer incubation time was necessary to detect its activity. Using casein as a substrate, we detected activity for stromelysin-1 (MMP-3) occurring near its known size of 58K Mr (11). We also detected the presence of the mRNA for gelatinase A, gelatinase B, collagenase-3, and stromelysin-1, suggesting these MMPs are indeed produced by stromal cells in vitro.
Levels of collagenase-3, stromelysin-1, and gelatinase B increased when stromal cells were cultured with EGF and bFGF, but not when TGFß was used. The observation that total protein content of the cells did not change with growth factor treatment suggests that cell numbers were not altered. It, therefore, seems likely that the increased enzyme activities were due to greater amounts of the proteolytic enzymes through either increased production by individual cells or decreased degradation of the MMPs. The mRNA levels for collagenase-3, stromelysin-1, and gelatinase B also increased after EGF and bFGF treatment, suggesting that regulation occurs either through increased transcription or increased mRNA stability. The apparent discrepancy in which bFGF treatment increased activity for gelatinase B, but not the mRNA levels, may have arisen because the band that was analyzed in the substrate-gel electrophoresis may not have been gelatinase B, or because regulation of gelatinase B may occur at the level of translation. Another discrepancy observed was that after EGF and bFGF treatment the increases in RNA levels were larger than the increases in proteolytic activity. This may be because the substrate-gel electrophoresis is not as precise a technique as the Northern blot analyses, because additional regulation of proteolytic activity may be occurring posttranscriptionally and posttranslationally, or because the increased mRNA levels had not yet been translated into a change in protein levels.
The observation that EGF and bFGF increase the production of identical MMPs suggests a similar regulatory system. Indeed, both EGF and bFGF use tyrosine kinase receptors and affect transcription through several pathways, including mitogen-activated protein kinase, which acts via Ets transcription factors, and c-Fos and c-Jun, which form the activating protein-1 transcription factor (37, 38). Mice with a targeted mutation of the Ets2 transcription factor show embryonic lethality due to decreased production of gelatinase B by trophoblast cells and failure of these cells to migrate and invade (39). Upon embryonic rescue, adult mutants have decreased levels of stromelysin-1, gelatinase B, and collagenase-3 in skin and lung tissue. In addition, bFGF was unable to induce the expression of these proteinases. In human breast cancer cells, EGF increases the production of MMPs via the Ets transcription factor (40), providing further evidence that increased MMP levels can result from the mitogen-activated protein kinase/Ets signal transduction pathway. TGFß, on the other hand, uses a serine-threonine receptor and mediates its effects via distinct signal transduction systems (16), which could explain why it had no effect on MMP production in our studies.
Gelatinase A, TIMP-2, and TIMP-3 do not contain Ets-binding sites within their promoter regions (41), which could explain why EGF and bFGF did not affect their levels, although TIMP-1 does. The reasons its levels were unaffected by EGF and bFGF may be due to an additional transcription factor regulating the MMPs or another factor inhibiting TIMP-1 production. We have previously shown that PGE2 increases the mRNA levels for gelatinase A within stromal cells in vitro, but not gelatinase B (18), which, together with the present findings, suggest differential regulation of the two gelatinases.
Many growth factors are elevated within the uterus during implantation
in rodents. EGF (3) and other members of its family, including TGF
(42) and heparin-binding EGF-like growth factor (43), are all present
at elevated levels within implantation sites early during pregnancy
along with their receptors (2). Because all three factors share the
same receptor (44), the increased production of MMPs we observed after
EGF treatment may also be elicited by TGF and heparin-binding EGF.
bFGF is also present within the rat uterus during pregnancy, when
levels rise after initiation of embryo invasion (4), and later, around
day 10 of pregnancy, when angiogenesis and remodeling of the vascular
matrix occur (5). Our observation that bFGF increases the production of
MMPs implicates this factor in mediating the remodeling associated with
angiogenesis in addition to decidualization.
In several physiological systems, TGFß maintains the stability of the ECM by increasing the production of ECM components, decreasing the production of proteolytic enzymes, and increasing the production of protease inhibitors (16). The mRNA for TGFß is present within the uterus during both early and late decidualization (7), but it was interesting that this factor had no effect on MMP or TIMP levels within rat stromal cells. Additionally, TGFß is able to stimulate the stromal cells to undergo apoptosis (20). However, protein and RNA levels within stromal cells were unaffected after treatment with TGFß, suggesting there was no apoptotic effect of this factor.
Despite their effects on the MMPs, EGF and bFGF had no effect on the mRNA levels for the TIMPs, which was interesting because mRNA levels for the TIMPs increase within the uterus during decidualization in vivo (18, 45). The TIMPs inhibit the activity of the MMPs, are important regulators of ECM remodeling, and limit the extent of trophoblast invasion. It was therefore surprising that TGFß, which increases mRNA levels for TIMP-1 and -3 in human decidual cells (46), did not affect the levels of these transcripts within rat decidual cells. We have previously shown that PGE2 increases mRNA levels for TIMP-1 within endometrial stromal cells in vitro (18), suggesting that eicosonoids and not growth factors regulate this inhibitor.
MMPs are involved in remodeling the ECM of the human endometrium during the menstrual cycle when the collagenases, gelatinases, and stromelysins are elevated during the proliferative phase and menstruation, but are less abundant during the secretory phase (47). Estrogen and progesterone are probably involved in regulating the MMPs during these phases. Interestingly, EGF and its receptor are abundant during the proliferative phase, which suggests that this factor could regulate MMP levels during this period (48). However, EGF levels are not elevated during menstruation. BFGF is present at high levels throughout the duration of the cycle, whereas TGFß levels are low during the proliferative phase, but increase during the secretory phase, suggesting that these two growth factors are not involved in regulating MMP levels in the human endometrium.
Tissue remodeling is also regulated by serine proteases, specifically the two forms of plasminogen activators (PAs), urokinase-type PA (uPA) and tissue-type PA (tPA) (49). The PAs cleave plasminogen to plasmin, which degrades ECM components such as fibronectin and laminin, and are activators of MMPs; their natural inhibitors are the plasminogen activator inhibitors. Transcripts for uPA, tPA, and plasminogen activator inhibitor-1 have all been detected within the uterus during decidualization both in vivo and in vitro (50, 51). PGE2 and EGF increase levels for uPA, but not tPA (51, 52), providing additional evidence that EGF increases the proteolytic activity of stromal cells in vitro.
The present findings demonstrate that endometrial stromal cells produce several MMPs during decidualization in vitro, and that the levels of collagenase-3, stromelysin-1, and gelatinase B are increased when the cells are treated with EGF and bFGF. Neither of these growth factors has an effect on gelatinase A or TIMP-1, -2, and -3. These observations as well as the previous observation that EGF increases plasminogen activator levels (45) suggest that specific growth factors may be responsible for increasing the proteolytic activity that occurs during decidualization in vivo.
|
Acknowledgments |
|---|
Received September 10, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
precursors in the mouse
uterus during the periimplantation period and after steroid hormone
treatments. Biol Reprod 50:481–491[Abstract]
This article has been cited by other articles:
 |
|
 |
|
  C.-S. Chou, H. Zhu, C. D. MacCalman, and P. C. K. Leung Regulatory Effects of Gonadotropin-Releasing Hormone (GnRH) I and GnRH II on the Levels of Matrix Metalloproteinase (MMP)-2, MMP-9, and Tissue Inhibitor of Metalloproteinases-1 in Primary Cultures of Human Extravillous Cytotrophoblasts J. Clin. Endocrinol. Metab., October 1, 2003; 88(10): 4781 - 4790. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C.-S. Chou, C.-J. Tai, C. D. MacCalman, and P. C. K. Leung Dose-Dependent Effects of Gonadotropin Releasing Hormone on Matrix Metalloproteinase (MMP)-2, and MMP-9 and Tissue Specific Inhibitor of Metalloproteinases-1 Messenger Ribonucleic Acid Levels in Human Decidual Stromal Cells in Vitro J. Clin. Endocrinol. Metab., February 1, 2003; 88(2): 680 - 688. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Binder, Th. Hagemann, B. Husen, M. Schulz, and A. Einspanier Relaxin enhances in-vitro invasiveness of breast cancer cell lines by up-regulation of matrix metalloproteases Mol. Hum. Reprod., September 1, 2002; 8(9): 789 - 796. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. B. Nothnick Disruption of the Tissue Inhibitor of Metalloproteinase-1 Gene in Reproductive-Age Female Mice Is Associated with Estrous Cycle Stage-Specific Increases in Stromelysin Messenger RNA Expression and Activity Biol Reprod, December 1, 2001; 65(6): 1780 - 1788. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Jokimaa, S. Oksjoki, H. Kujari, E. Vuorio, and L. Anttila Expression patterns of cathepsins B, H, K, L and S in the human endometrium Mol. Hum. Reprod., January 1, 2001; 7(1): 73 - 78. [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 |
