Vol. 80, No.

1
Journal of Clinical Endocrinology and Metabolism Printed in U.S.A.
Copyright 0 1995 by The Endocrine Society

Reduction of Extracellular Matrix Protein Expression in

Human Amnion Epithelial Cells by Glucocorticoids:
A Potential Role in Preterm Rupture of the Fetal
Membranes*
SETH GULLER, LINDA KONG, ROBERT WOZNIAK, AND CHARLES J. LOCKWOOD

Downloaded from https://academic.oup.com/jcem/article/80/7/2244/2655608 by guest on 19 April 2021

Departments of Obstetrics/Gynecology and Reproductive Science (S.G., R. W., L.K., C.J.L.) and
Biochemistry (S.G.), Mount Sinai Medical Center, New York, New York 10029

ABSTRACT in amnion cells did not result from a general reduction in protein
Low levels of expression of extracellular matrix (ECM) proteins in synthesis. Cortisol and DEX reduced FN expression in amnion cells,
chorioamniotic membranes is a characteristic of prematurely rup- with half-maximal effective concentrations of approximately 60 and
tured membranes, a condition associated with 40% of preterm deliv- 8 nmol/L, respectively. In immunoprecipitation studies, DEX treat-
eries. In light of the rise in levels of glucocorticoids (GC) in amniotic ment reduced FN and collagen III synthesis to 20% of control levels,
fluid associated with preterm labor, in the present study we examined suggesting that GC may coordinately reduce the synthesis of major
the effects of GC on the expression of major ECM proteins in cultures ECM proteins in amnion cells. Similarly, DEX treatment reduced the
of amnion epithelial cells recovered after digestion of human term levels of FN messenger ribonucleic acids in amnion cells to approx-
amnions. Amnion cells were maintained with and without 10m7 mol/L imately 15% ofcontrol levels. DEX treatment also promoted a marked
dexamethasone (DEX), and levels of the ECM protein fibronectin (FN) reduction in FN expression in amnion cells cultured in serum-free
were determined by enzyme-linked immunosorbent assay. DEX treat- medium to lo-50% of control levels. Our results indicate that GC
ment reduced FN expression in amnion epithelial cells to 15-30% of negatively regulate ECM protein expression in amnion epithelial
control levels and reduced FN expression in placental cells to 30-50% cells, suggesting a potential role in the genesis of altered fetal mem-
of control levels. Conversely, DEX treatment weakly stimulated FN brane ECM protein expression associated with prematurely ruptured
expression in chorion cell cultures. DEX treatment did not affect the membranes. (J Clin Endocrinol Metab 80: 2244-2250, 1995)
total level of amnion cell protein, indicating that the effects of DEX

P RETERM LABOR in humans is known to be associated

with marked changes in the composition and/or dis-
tribution of extracellular matrix (ECM) proteins synthesized
by cellsin the fetal membranes (l-4). A general thinning and
reduction in tensile strength of amniochorionic membranes
have been specifically implicated in premature rupture of the
fetal membranes (PROM), a condition associatedwith 40% of
premature deliveries (1,2). The collagen content of the am-
nion decreases,whereas collagenolytic properties of the am-
nion increase in patients with PROM who subsequently de-
velop preterm labor (l-3). Oncofetal fibronectin (onfFN), an
extremely abundant ECM protein synthesized by the pla-
centa and fetal membranes, has been localized to regions of
uterine-placental and decidual-fetal membrane contact (4,5).
The appearance of amniochorionic-derived onfFN in vaginal
and cervical secretions of women has been used to predict
preterm and postterm labor (4,6). BecauseECM proteins are
critical regulators of cell adherence (7), changes in the
expression and/or distribution of ECM proteins in fetal
membranes are likely to play a key role in the reduction of
fetal membrane integrity associated with PROM.
Received February 3, 1995. Accepted March
Address all correspondence and requests for reprints
Guller, Department of Obstetrics/Gynecology
Sciences, New York University School of Medicine,
New York, New York 10016.
* This work was supported in part by a Mount
Seed Grant (to S.G.).
Labor in humans, whether occurring before or at term, is
associated with a dramatic increase in the concentration of
glucocorticoids (GC) in amniotic fluid and maternal and fetal
sera (8-10). The amnion epithelium is in direct contact with
amniotic fluid (11) and, as a consequence,will be exposed to
high levels of GC during preterm labor. The focus of the
present study was to employ cultures of amnion epithelial
cells’ to elucidate the potential effects of GC on ECM protein
expression in the amnion associatedwith preterm labor and
PROM.
Protocols describing the isolation of amnion epithelial cells
by trypsin treatment of human term amnions (12, 13) have
been adopted by a number of research groups (14-17). Using
these procedures, amnion epithelial cells are separated from
the underlying basal lamina as an essentially homogeneous
population of cells (12,131.Prostaglandin (PG) production by
amnion epithelial cells was increased by treatment with epi-
dermal growth factor and transforming growth factor-p
(TGFP) (18). Treatment of amnion epithelial cells with GC
increased the synthesis of CRH and modulated the output of
prostaglandin E, (PGE,) (14, 16, 17). In the present investi-
gation we examined the effects of GC treatment on the ex-
10, 1995. pression of FN and collagen III in human amnion epithelial
to: Dr. Seth cells. These two large ECM proteins have been documented
and Reproductive
550 First Avenue,
’ The terms amnion epithelial cells, amniocytes, and amnion cells are
Sinai Medical Center used interchangeably and denote the homogeneous population of cells
obtained from treating whole amnion tissue with trypsin.

2244
 GLUCOCORTICOID CONTROL OF AMNION MATRIX PROTEINS 2245

by biochemical and immunohistochemical procedures to be with a cell viability between 70-90%, as determined by trypan blue
major ECM proteins synthesized by amnion epithelial cells exclusion.
Cultures of placental and chorion cells were obtained using methods
(13). Our results indicate that GC treatment markedly and described by Jones et al. (14). Essentially, all procedures were followed
coordinately suppresses the synthesis of ECM proteins by as described above for amnion cells, except that the first 15-min incu-
amnion epithelial cells, suggesting a mechanism by which bation was omitted from the protocol. Using these procedures, we ob-
GC may play a role in the aberrant expression of ECM tained approximately 10 X lo6 chorion cells and 50 X lo6 placental
cells/l0 g tissue. Based on these protocols, we obtained an essentially
proteins associated with PROM.
homogeneous population of amnion epithelial cells, as described by other
groups (12-18). The placental preparation consisted of cytotrophoblasts and
other cell types, whereas fibroblasts were the predominant cell type in
Materials and Methods chorion cell cultures by day 7.
For the experiments, a&ion, chorion, and placental cells were plated
Materials in BM supplemented with 10% charcoal-stripped calf serum (SCS me-
dium) at a density of 0.4 X lo6 cells/well of a 24-well dish for ELISA
Tissue culture media, plasticware, bovine sera, and reagents used in studies. Amnion cells were plated at a density of 2.4 X lo6 cells/well of

Downloaded from https://academic.oup.com/jcem/article/80/7/2244/2655608 by guest on 19 April 2021

the isolation of cells from human term amnion, placenta, and chorion a 6-well dish for immunoprecipitation and Northern blotting proce-
were from previously described sources (19, 20). The medium supple- dures. Cells were maintained at 37 C in a humidified atmosphere of 5%
ment ITS+ (a mixture containing insulin, transferrin, and selenium) was CO,-95% air, and fresh culture medium was added every 2-3 days.
obtained from Collaborative Research-Becton Dickinson (Bedford, MA). Alternatively for serum-free culture, amnion cells were plated at a den-
Dexamethasone (DEX) and murine monoclonal antibody (mAb) to fi- sity of 0.4 X lo6 cells/well of a 24-well dish in SCS medium. Fresh media
bronectin were purchased from Sigma Chemical Co. (St. Louis, MO). were added every 48 h until confluency was reached on day 6 or 7.
Collagen Ill mAb was obtained from Chemicon (Temecula, CA). The Culture media were then aspirated, cells were washed twice with basal
ClRdE PREP kit used in the purification of plasmid DNA was obtained medium, and serum-free medium [BM supplemented with ITS+, 5
from Bio 101 (La Iolla, CA). I3 SlProtein Labeline Mix and I~-32Pldeoxv- pmol/L FeSO,, 50 kmol/L ZnSO,, 1 nmol/L CuSO,, 50 pg/mL ascorbic
CTP were obtained from New England Nucl& Corp. (Boston, MA). acid, and trace elements (Gibco-BRL, Gaithersburg, MD)], with and
RNazol B was purchased from CINNA/BlOTECX Laboratories (Hous- without lo-’ mol/L DEX, was added. This serum-free medium was
ton, TX). Complementary DNAs (cDNAs) to human FN (21) and human used in human endometrial stromal cell studies (23).
glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (22) were ob-
tained from the American Type Culture Collection (Rockville, MD). onfFN immunoassay. Levels of onfFN in culture medium were quanti-
Other reagents used in immunoprecipitation and Northern blotting tated by immunoassay (Adeza Biomedical) using FDC-6 mAb, as we
procedures were purchased from previously described sources (19,20). described previously (19, 20). The FDC-6 mAb detects FN molecules
The enzyme-linked immunosorbent assay (ELISA) kit using FlX-6 mAb bearing the oncofetal epitope (24). For experiments, media were added
for quantitating levels of onfFN was a gift from Adeza Biomedical (Sunny- to 96-well plates coated with FDC-6 mAb. Incubations with goat anti-
vale, CA). human plasma FN Ab and phenophthalein monophosphate were car-
ried out as previously described (19,20). Optical densities, determined
for triplicate wells, were converted to levels of onfFN based on com-
Methods parison with a curve generated from onfFN standards ranging in con-
centration from O-1000 ng/mL. lntra- and interassay errors were less
Cell culture. Term amnions (n = 15), placentas (n = 3), and chorions
than lo%, as previously reported (19). Medium levels of onfFN were
(n = 3) were obtained from women (aged 24-39 yr) undergoing non-
normalized to levels of cell protein quantitated by the Micro BCA protein
laboring cesarean section at 37-42 weeks gestation. Tissues were ob-
kit from Pierce Chemical Co. (Rockford, IL) (20).
tained from uncomplicated pregnancies with normally grown singleton
fetus. None of the patients’ pregnancies were complicated by preterm Immunoprecipitation of labeled FNand collagen III. Rates of synthesis of FN
labor or preterm rupture of membranes, abruptio placentae, preeclamp- and collagen Ill in amnion cells were examined in immunoprecipitation
sia, diabetes, or other pathological conditions. For the experiments, studies, as we previously described (20). For experiments, cells main-
placentas and fetal membranes were transported to the laboratory im- tained in SCS medium with or without 10e7 mol/L DEX were labeled
mediately after delivery. The fetal membranes with adherent decidua for 4 h with [35S]Protein Labeling Mix in methionine-free SCS medium
vera were separated from the placental disc and washed in sterile saline. in the presence or absence of 10m7 mol/L DEX. Culture media were
The amnion was peeled away from the chorion-decidua, and the decidua collected, and approximately lo6 trichloroacetic acid precipitable cpm
was scraped from the chorion and discarded. Villous tissue dissected labeled culture medium/sample were precleared using BSA-Sepharose.
from the placental disc was also thoroughly washed with saline, and Samples were then incubated for 2 h at 4 C with a 1:50 dilution of anti-FN
approximately 10 g amnion, placental, or chorion tissue were used in the or anti-collagen III mAb and protein G-Sepharose (20). Immune com-
procedures described below. plexes were washed, and sodium dodecyl sulfate-polyacrylamide gel
Isolation of amnion epithelial cells was carried out based on proce- electrophoresis was carried out as previously described (20).
dures described by Rice Okita et al. (12). For the experiments, amnion
tissue was minced in sterile saline, weighed, and placed in a flask Northern blotting. Ribonucleic acid (RNA) was extracted from amnion
containing 150 mL of a digestion solution consisting of basal medium epithelial cells using the RNazol B procedure, as we reported previously
(BM; a 1:l mixture of phenol red-free Ham’s F-12-Dulbecco’s Modified (19, 20). Before electrophoresis, samples were incubated with 40 U de-
Eagle’s Medium) supplemented with trypsin (5 mg/mL) and deoxyri- oxyribonuclease and extracted with phenol-chloroform-isoamyl alcohol
bonuclease (0.1 mg/mL). The mixture was agitated (110 rpm) at 37 C for (25:24:1), and RNA was precipitated using 2 vol ethyl alcohol. Approx-
15 min in a Dubnoff shaking incubator, the contents were poured imately 10 pg total RNA (based on A,,,) per sample were separated on
through four layers of gauze, and the blood-free tissue was retained for a 1% (wt/vol) agarose gel containing 2.2 mol/L formaldehyde (25). RNA
further treatment. The tissue was then agitated (125 rpm) for 30 min in was transferred to nylon membranes, and levels of FN and GAI’DH
200 mL digestion solution at 37 C, and the supernatant was recovered messenger RNAs (mRNAs) were determined using 32P-labeled cDNAs
after passing the mixture through a 0.16-Frn sieve. The undigested tissue generated by random primer synthesis (25). Plasmids containing cDNAs
fragments were recovered and placed in 200 mL fresh digestion solution to human FN (21) and GAPDH (22) were isolated, as previously de-
and incubated at 37 C for an additional 30 min. The supernatants from scribed (20). Autoradiography and densitometry of immunoprecipita-
the second and third incubations were combined and centrifuged (500 tion gels and Northern blots were carried out as reported previously
X g; 5 min), and the cell pellets were resuspended in a 1:l mixture of (19, 20).
Ham’s F-12-Waymouth’s medium supplemented with 10% fetal bovine
serum. The cells were then resuspended and washed twice in fetal Statistics. Results are expressed as the mean 2 SE. Data were analyzed
bovine serum medium before cell counting. Using this procedure, we by Student’s t test using the Statworks Cricket software package from
obtained 20-60 X 10h amnion epithelial cells from 7-15 g amnion tissue, Heyden and Son (London, UK). The values presented in this manuscript
2246 GULLER ET AL. JCE & M . 1995
Vol80. No 7

depict representative data obtained in experiments carried out at least 30-50% of control levels. Conversely, DEX treatment had a
three times. small stimulatory effect on the level of FN expression in
chorion cell cultures between days 8-13.
Results
Suppression of FN expression by GC in amnion cells
Effect of DEX treatment on FN levels in amnion, placental,
and chorion cells In the four independent experiments using cultures of
amnion epithelial cells presented in Table 1, it is important
Primary cell cultures of human amnion, placenta, and to note that although DEX significantly suppressed the ex-
chorion obtained after dispersion of fresh tissue with trypsin pression of FN in amnion cells, it did not significantly affect
were maintained for 13 days in SCS medium with and with- the levels of total cellular protein per culture well. This in-
out lop7 mol/L DEX. Based on an ELISA that detects an dicated that the DEX-mediated suppressionof FN expression
oncofetal epitope in FN, we observed that between days 2-13 in amniocytes did not result from a general reduction in

Downloaded from https://academic.oup.com/jcem/article/80/7/2244/2655608 by guest on 19 April 2021

of culture, DEX treatment reduced the levels of FN in the protein synthesis.
medium of amnion cells to 15-50% of control levels (Fig. 1). As DEX is not a physiological GC, we also examined the
DEX treatment reduced FN expression in placental cells to effects of cortisol on FN expression in amnion cells. We noted
that cortisol and DEX mediated a dose-dependent inhibition
AMNION of FN expression to 20% of control levels, with half-maximal
160,
effective concentrations (EC,,) of approximately 50 and 5
U Control
+ DEX nmol/L, respectively (Fig. 2). In three independent experi-
ments the EC,, values for the cortisol and DEX-dependent
inhibition of FN expression were 60.0 2 9.4 and 8.3 2 5.6
nmol/L, respectively. The finding that DEX was 7- to 8-fold
more potent than cortisol in the reduction of FN expression
in amnion cells may reflect a reduction in the biological
potency of cortisol due to the preferential binding of cortisol,
and not DEX, to corticosteroid-binding globulin present in
0 ‘I 6 12 16
our serum-containing culture medium.
Time (days)

PLACENTA
Effects of DEX treatment on synthesis of ECM proteins by
amnion epithelial cells
Etro’ Amnion cells were maintained for 5 days in SCS medium
in the absenceand presence of 10m7mol/L DEX. After bio-
synthetic labeling, FN and collagen III were immunoprecipi-
tated from culture medium using FN- and collagen III-
specific mAb. We observed that DEX treatment reduced the
/’
20 -;i&-gy
,I-’ TABLE 1. The effects of DEX treatment on levels of onfFN and
o-
0 4 6 12 16 protein in amnion epithelial cells
Time (days)
Amnion
Treatment Protein (fig) onfFN (ng/pg protein)
preparation

CHORION 1 Control 131.9 -+ 6.0 183.9 z 6.1

DEX 141.6 t 3.3 37.1 + 0.2

i2F0’ 2 Control
DEX
168.9
180.6
k 6.6
k 10.4
136.9
21.6
-+ 4.8
k 0.8
3 Control 76.7 2 8.5 244.0 2 29.0
DEX 85.0 -t 8.2 44.9 5 2.4
4 Control 92.8 + 7.6 226.6 2 26.2
DEX 114.0 2 10.3 29.3 2 3.3
Amnion epithelial cells were maintained in SCS medium in the
absence (Control) or presence of 10m7 mol/L DEX for 13-14 days.
0 4 6 12 16 Levels of total cellular protein were determined after lysis of cells on
Time (days) the last day of each experiment. Levels of onfFN in culture medium
were determined by ELISA and represent the maximal levels of
FIG. 1. Modulation of onfFN expression in amnion, placental, and onfFN production achieved on days 7-11 of culture. All values are
chorion cell cultures by DEX. Amnion, placental, and chorion cells presented as the mean k SEM of determinations carried out in trip-
were maintained for 13 days in SCS medium with or without 10m7 licate wells of a 24-well plate. For pooled data, the levels of protein in
mol/T.. DEX, and levels of onfFN in culture medium were determined the four independent experiments are: control, 118.1 ? 15.5; and
at the indicated time by ELISA. Levels of onfFN in triplicate culture DEX, 130.6 k 17.7 (based on Student’s t test, P = 0.676 vs. control).
wells are expressed as the mean k SE. SEs that fell within the symbol Pooled values for onfFN are: control, 197.9 2 20.7; and DEX, 33.2
are not presented. k 5.0 (P < 0.01 vs. control).
 GLUCOCORTICOID CONTROL OF AMNION MATRIX PROTEINS 2247

300
F.- _f_ DEX
Q
FIG. 2. Dose-dependent effects of GC + Cortisol
treatment on onfFN expression in am- p Control
nion epithelial cells. Amion cells were 200 -
maintained for 9 days in SCS medium
without GC (Control) or with the indi- z
cated concentration of cortisol or DEX. 3
Levels of onfFN in culture medium were 5
determined by immunoassay and are 100 -
expressed as the mean 2 SE of determi-
nations carried out in triplicate. SEs z
were within l-20% of mean values, but -E
were not presented for reasons of 0
0. , . , . , . , . , . ,

Downloaded from https://academic.oup.com/jcem/article/80/7/2244/2655608 by guest on 19 April 2021

clarity.
0 0.1 1 10 100 1000

Concentration of Steroid (nmol/L)

levels of FN and collagen III synthesis to approximately 20%
of control levels (Fig. 3). In three independent experiments,
DEX treatment suppressed the synthesis of FN and collagen
III in amnion cells to 31 ? 5.8% (P < 0.001, DEX VS.control)
and 37 k 13.3% (P < 0.01, DEX VS. control) of the control
value, respectively. This suggeststhat GC may coordinately
reduce the synthesis of major ECM proteins in amnion
epithelial cells.
Suppression of FN mRNA levels in amnion cells by DEX
Amnion epithelial cells were cultured for 5 days in SCS
medium with and without 10m7mol/L DEX, and levels of
FN mRNA were quantitated after Northern blotting and
normalization to levels of GAPDH mRNA. Similar to the
results obtained in the ELISA and immunoprecipitation
studies presented above, DEX treatment reduced levels of FN
mRNA in amnion cells to approximately 10% of control
levels (Fig. 4). Statistical analysis of Northern blotting data,
obtained in three independent experiments, indicated that
DEX treatment reduced the expression of FN mRNA to 14.7
+ 4.6% of the control value (P < 0.001, DEX us. control).
Effect of DEX on amniocyte FN expressionin serum-free
medium
To establish defined culture conditions in which to dissect
the hormonal regulation of ECM protein expression in the
amnion, amnion epithelial cells were maintained for 9 days
postconfluence in serum-free medium in the presence or

FN COL 111
+ +

FN

28s

18s
- -
GAPDH

FIG. 3. Effects of DEX treatment on synthesis of FN and collagen III

in amnion cells. Cells incubated for 7 days in medium without (-1 and
with (+) 10e7 mol/L DEX were then metabolically labeled with
[35S1Protein Labeling Mix for 4 h in methionine-free SCS medium.
Immunoprecipitation of labeled ECM proteins was carried out using
mAb specific for FN and collagen III. The migration of labeled FN and
collagen III is indicated by an arrow. Lines reflect the migration
positions of mol wt standards (180K, a,-macroglobulin; 116K, 6-
galactosidase). Quantitation was carried out after autoradiography
and densitometry of immunoprecipitation gels.
FIG. 4. Suppression of levels of FN mRNA in amnion cells by DEX
treatment. Amnion epithelial cells were maintained for 5 days in SCS
without (-) and with (+) 10-r mol/L DEX. Ten micrograms of total
RNA were loaded per lane before electrophoresis and Northern blot-
ting. Levels of FN and GAPDH mRNA were detected concomitantly
on the same blot using 32P-labeled cDNA probes to FN and GAPDH.
Levels of FN mRNA were normalized to levels of GAPDH mRNA after
autoradiography and densitometry.
2248
300
200
100
0
0 2
Time
FIG. 5. Effects of DEX treatment on onfFN expression
postconfluence in serum-free medium in the presence
were determined by immunoassay, and the results are expressed
absenceof 10m7mol/L DEX. Under serum-free conditions,
DEX treatment reduced the levels of FN expression in am-
nion cells to lo-50% of control levels between days 2-9 of
culture.
Our results indicate that GC are major negative regulators
of ECM protein synthesis in human amnion epithelial cells.
Discussion
The amnion epithelial cell layer comprises the innermost
of the five layers of the amnion and, thus, is in direct contact
with amniotic fluid (11,13). Amnion cells are anchored to a
thin basal lamina by keratin intermediate filaments in a
hemidesmosome structure similar to that in the skin and
cornea (13). In immunohistochemical studies, the basal lam-
ina was found to be rich in FN, laminin, and collagen III. It
was previously demonstrated in vitro that amnion epithelial
cells, obtained as a homogeneous population after trypsin
treatment of whole amnion, synthesized extremely large
quantities of FN and collagen III (13).
The studies presented in this manuscript demonstrated
that DEX treatment down-regulated FN expression in both
amnion and placental cultures. Unlike the amnion, trypsin
digestion of human term placenta generates a mixed popu-
lation of cells consisting of cytotrophoblasts and other cell
types (26). We previously demonstrated that DEX treatment
suppressesFN expression in homogeneous cultures of cy-
totrophoblasts isolated from human term placentas (19,201.
Drawing from our present study, it is likely that suppression
of FN expression in mixed placental cell cultures is due in
large part to the GC responsivenessof cytotrophoblasts. It is
interesting to note that the amnion epithelial cell is derived
from a cytotrophoblast precursor during the seventh or
eighth day of development (27). Thus, the GC-mediated sup-
pression of ECM protein synthesis in amnion cells and cy-
totrophoblasts may be a function of their common cellular
origin. The slight increase in FN expression in chorion cell
cultures promoted by DEX in the present study most likely
reflects effects on fibroblastic cells that rapidly proliferate
GULLER ET AL. JCE & M . 1995
Vol80 . No 7
Control
DEX
Downloaded from https://academic.oup.com/jcem/article/80/7/2244/2655608 by guest on 19 April 2021
4 6 a lb
(days)
in amnion cells in serum-free medium. Amnion cells were maintained for 9 days
or absence of 10m7 moVLDEX. At the indicatedtime, levelsof onfFN in culture medium
as the mean 2 SE of determinations carried out in triplicate culture wells.
and outgrow other cell types (e.g. extravillous cytotropho-
blasts) within the culture.
In previous studies, GC were found to decrease the pro-
duction of PGE, in subconfluent cultures of human amnion
epithelial cells (15), whereas GC stimulated the synthesis of
PGE, in postconfluent amnion cells (16, 17). These results
suggest that the cellular growth state may influence GC-
mediated actions in amniocytes. In our experiments, we ob-
served that CC suppressed FN expression between days 2
and 14 of culture. As confluency was reached on days 6-7,
this indicated that the effects of GC on FN expression were
mediated at both sub- and postconfluency. Our serum-free
experiments were conducted on confluent cultures because
the presence of serum in our culture medium increased the
plating efficiency and shortened the time required by cells to
reach confluency (not shown). The use of serum-free me-
dium in future experiments will facilitate the identification
of cofactors or mediators involved in the GC-mediated
reduction of FN expression in amnion cells.
The data presented here indicate that GC suppress the
synthesis of FN and collagen III protein in amniocytes aswell
as reduce the expression of FN mRNA in these cells. It has
been reported that the actions of GC on PGE, levels was
mediated through the GC receptor (GR) (17). It is likely in the
present study that the GC-dependent suppression of FN
expression in amnion cells was mediated by the GR. To the
best of our knowledge, there is no information available
concerning the cellular mechanism of negative regulation of
FN expression by GC. Inhibition of FN expression in amnion
cells by GC could be modulated transcriptionally after bind-
ing of GC to the GR, asdescribed for the effects of GC on PRL
(28) and proliferin genes(29). Alternatively, it is possible that
the GC-mediated suppression of steady state levels of
FN mRNA in amnion cells resulted from a reduction in the
stability of FN mRNA. Interestingly, a recent report
demonstrated that the oncogene HA-ras down-regulated FN
expression in a human osteosarcoma cell line through
nuclear posttranscriptional events, possibly involving al-
tered processing or stability of FN RNA in the nucleus (30).
 GLUCOCORTICOID CONTROL OF AMNION MATRIX PROTEINS 2249

In humans, there is a marked increase in the concentration 3. Vadillo-Ortega F, Gonzalez-Avila G, Karchmer S, Meraz Cruz N,
of GC and CRH in amniotic fluid as well as maternal and fetal Ayala-Ruiz A, Lama MS. 1990 Collagen metabolism in premature
rupture of amniotic membranes. Obstet Gynecol. 75:84-88.
plasma near parturition, and the concentrations of these com- 4. Lockwood CJ, Senyei AE, Dische MR, et al. 1991 Fetal fibronectin
pounds drop precipitously postpartum (8-10). A number of in cervical and vaginal secretions as a predictor of preterm delivery.
reports also indicate that parturition, whether occurring be- N Engl J Med. 325:669-674.
fore or at term, is characterized by changes in the composi- 5. Feinberg RF, Kliman HJ, Lockwood CJ. 1991 Is oncofetal fibronec-
tin a trophoblast glue for human implantation? Am J Pathol. 138:
tion and/or distribution of ECM proteins in fetal membranes
537-543.
(l-4). In addition to the general thinning of chorioamniotic 6. Lockwood CJ, Moscarelli RD, Wein R, Lynch L, Lapinski RH,
membranes and decrease in tensile strength that occur in Ghidini A. 1994 Low concentrations of vaginal fetal fibronectin as
patients with PROM who subsequently deliver preterm (l), a predictor of deliveries occurring after 41 weeks. Am J Obstet
specific alterations in collagen and FN expression in fetal Gynecol. 171:1-4.
7. Schwarzbauer JE. 1991 Fibronectins: from gene to protein. Curr
membranes are also reported to be associated with PROM Opin Cell Biol. 3:786-791.
and preterm labor in humans (Z-4). THe collagen content of 8. Dorr HG, Heller A, Versmold HT, et al. 1989 Longitudinal study

Downloaded from https://academic.oup.com/jcem/article/80/7/2244/2655608 by guest on 19 April 2021

amnions from patients with PROM is extremely low, of progestins, mineralocorticoids, and glucocorticoids throughout
whereas collagenolytic activity of amnions was high relative human pregnancy. J Clin Endocrinol Metab. 68:863-868.
to that in controls (2, 3). Lockwood et al. (4, 6) documented 9. Warren WB, Patrick SL, Goland RS. 1992 Elevated maternal plasma
corticotropin-releasing hormone levels in pregnancies complicated
that the presence of amniochorionic-derived oncofetal FN in by preterm labor. Am J Obstet Gynecol. 166:1198-1207.
vaginal and cervical secretions could be used to predict labor 10. Wolfe CDA, Pate1 SF’, Linton EA, et al. 1988 1988 Plasma CRH in
pre- and postterm. abnormal pregnancy. Br J Obstet Gynaecol 95:1003-1006.
Based on the data presented in this manuscript, we suggest 11. Boume GL, Lacy D. 1960 Ultra-structure of human amnion and its
possible relation to the circulation of amniotic fluid. Nature. 186:
that the matrix-suppressive actions of GC in the amnion may
952-954.
be associated with PROM. With respect to this hypothesis, it 12. Rice Okita J, Sagawa N, Casey ML, Snyder JM. 1983 A comparison
is important to consider other factors that could oppose or of human amnion tissue and amnion cells in primary culture by
mitigate GC action at the cellular level. TGFP, a known morphological and biochemical criteria. In Vitro. 19:117-126.
positive regulator of ECM protein expression (7), present in 13. Alitalo K, Kurkinen M, Vaheri A. 1980 Extracellular matrix com-
ponents synthesized by human amniotic epithelial cells in culture.
high concentrations in amniotic fluid (31,32), could offset the Cell. 19:1053-1062.
actions of GC in the amnion. Recent data suggest that TGFP 14. Jones SA, Brooks AN, Challis JRG. 1989 Steroids modulate corti-
stimulates ECM protein synthesis in the amnion (data not cotropin-releasing hormone production in human fetal membranes
shown) and placenta (33). It is also critical to consider the and placenta. J Clin Endocrinol Metab. 68:825-830.
15. Gibb W, Lavoie JC. 1990 Effects of glucocorticoids on prostaglandin
actions of 11 /3-hydroxysteroid dehydrogenase (11 PHSD) and
formation by the human amnion. Can J Physiol Pharmacol. 68:
corticosteroid-binding globulin in the regulation of GC ac- 671-676.
tion. 1lPHSD oxidizes cortisol and corticosterone to their 16. Mitchell MD, Lytton FD, Varticovski L. 1988 Paradoxical stimu-
ll-keto derivatives, which are receptor inactive and thus do lation of both lipocortin and prostaglandin production in human
not promote GC effects (33). Unlike the placenta and chorion, amnion cells by dexamethasone. Biochem Biophys Res Commun.
151:137-141.
the amnion has been found to be devoid of 11PHSD activity 17. Potestio FA, Zakar T, Olson DM. 1988 Glucocorticoids stimulate
(35,36). Thus, during human pregnancy, cortisol present in prostaglandin synthesis in human amnion cells by a receptor-
amniotic fluid could exert dramatic effects on ECM protein mediated mechanism. J Clin Endocrinol Metab. 671205-1210.
production by the amnion epithelium. Corticosteroid-bind- 18. Kniss DA, Zimmerman PD, Fertel RH, Iams JD. 1993 Trans-
forming growth factor-p potentiates epidermal growth factor-
ing globulin is known to bind cortisol and other physiolog-
induced prostaglandin E, production in amnion cells. Prostag-
ical GC and reduce their biological potency (37). If the level landins. 45:27-33.
of CBG in amniotic fluid is regulated across gestation, this 19. Guller S, Lacroix N, Krikun G, et al. 1993 Steroid regulation
will also impact on effects of GC on ECM synthesis by of oncofetal fibronectin expression in human cytotrophoblasts.
amnion epithelial cells. J Steroid Biochem Mol Biol. 46:1-10.
20. Guller S, Wozniak R, Krikun G, Bumham JM, Kaplan P, Lock-
In summary, our data suggest that GC may play an im-
wood CJ. 1993 Glucocorticoid suppression of human placental fi-
portant role in the genesis of aberrant patterns of ECM bronectin expression: implications in uterine-placental adherence.
protein expression associated with PROM. Endocrinology. 133:1139-1146.
21. Komblihtt AR, Vibe-Pederson K, Barelle FE. 1984 Human fi-
bronectin cell specific alternative mRNA splicing generates
Acknowledgments polypeptide chains differing in the number of internal repeats.
Nucleic Acids Res. 12:5853-5868.
We would like to thank Dr. En-Yu Wang and Rebeca Caze for their 22. Tso JY, Sun X-H, Kao T-H, Reece S, Wu R. 1985 Isolation and
expert technical assistance. We also acknowledge the efforts of characterization of rat and human glyceraldehyde-3-phosphate de-
Rosemary Wein for her help in the procurement of placentas. hydrogenase cDNAs: genomic complexity and molecular evolution
of the gene. Nucleic Acids Res. 13:2485-2502.
23. Irwin JC, Kirk D, King RJB, Quigley - _ NM, Gwatkin RBL. 1989
References Hormonal regulation ofhuman endometrial stromal cells in culture:
an in vitro model for decidualization. Fertil Steril. 52:761-768.
1. Artal R, Sokol RJ, Newman M, Burstein AH, Stojkov J. 1976 24. Matsuura H, Takio K, Titani K, et al. 1988 The oncofetal structure
The mechanical properties of prematurely and nonprematurely of human fibronectin defined by monoclonal antibody FDC-6. J Biol
ruptured membranes. Am J Obstet Gynecol. 153:125-655. Chem. 263:3314-3322.
2. Kanayama N, Terao T, Kawashima Y, Horiuchi K, Fujimoto D. 25. Sambrok J, Fritsch EF, Maniatis T. 1989 Molecular cloning-a
1985 Collagen types in normal and prematurely ruptured amniotic laboratory manual, 2nd ed. Cold Spring Harbor: Cold Spring Harbor
membranes. Am J Obstet Gynecol. 153:899-903. Laboratory; 7.37-7.45.
2250 GULLER ET AL. JCE & M . 1995
Vol80. No 7

26. Kliman HJ, Nestler JE, Sermasi E, Sanger JM, Straus III JF. 1986 32. Lang AK, Searle RF. 1994 The immunomodulatory activity of
Purification, characterization, and in vitro differentiation of cyto- human amniotic fluid can be correlated with transforming growth
trophoblasts from human term placentae. Endocrinology. 118: factor beta-l (TGF-beta 1) activity. Clin Exp Immunol. 97:
1567-1582. 158-163.
27. Pritchard JA, MacDonald PA, Gant NF. 1985 The placenta and fetal 33. Feinberg RF, Kliman HJ, Wang C-L. 1994 Transforming growth
membranes. In: Williams’ obstetrics, 17th ed. Norwalk: Appleton- factor-8 stimulates trophoblast oncofetal fibronectin synthesis in
Century Crofts; 97-117. vitro: implications for trophoblast implantation in vim. J Clin
28. Sakai DD, Helmus S, Carlstedt-Duke J, Gustafsson J-A, Rottman Endocrinol Metab. 78:1241-1248.
FM, Yamamoto KR. 1988 Hormone-mediated repression: a negative 34. Funder JW. 1993 Mineralocorticoids, glucocorticoids, receptors and
glucocorticoid response element from the bovine prolactin gene. response elements. Science. 259:1132-1133.
Genes Dev. 2:1144-1154.
35. Pearson Murphy BE. 1977 Chorionic membrane as an extra-adrenal
29. Diamond ML, Miner JN, Yoshinaga SK, Yamamoto KR. 1990 Tran-
source of foetal cortisol in human amniotic fluid. Nature. 266:
scription factor interactions: selectors of positive or negative regu-
179-181.
lation from a single DNA element. Science. 249:1266-1272.
30. Chandler LA, Ehretsmann CP, Bourgeois S. 1994 A novel mechanism 36. Lopez Bemal A, Flint APF, Anderson ABM, Tumbull A. 1980
of Ha-rus oncogene action: regulation of fibronectin mRNA levels by a lip-hydroxysteroid dehydrogenase activity (E.C.1.1.1.146) in hu-

Downloaded from https://academic.oup.com/jcem/article/80/7/2244/2655608 by guest on 19 April 2021

nuclear posttranscriptional event. Mol Cell Biol. 14: 3085-3093. man placenta and decidua. J Steroid Biochem. 13:1081-1087.
31. Roberts AB, Spom MB. 1992 Differential expression of TGF-beta 37. Sandberg AA, Slaunwhite Jr WR. 1971 Physical state of adrenal
isoforms in embryogenesis suggests specific roles in developing and cortical hormones in plasma. In: Christy NP, ed. The human adrenal
adult tissues. Mol Reprod Dev. 32:91-98. cortex. New York: Harper and Row; 69-86.