| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLES |
Department of Biochemistry (Y.F., P.L., M.N.-H.), Biophysics and Molecular Biology, Department of Zoology and Genetics (J.T.F., R.T.H.), and Department of Animal Sciences (S.P.F.), Iowa State University, Ames, Iowa 50011
Address all correspondence and requests for reprints to: Marit Nilsen-Hamilton, Department of Biochemistry, Biophysics and Molecular Biology, Iowa State University, 3206 Molecular Biology Building, Ames, Iowa 50011. E-mail: marit{at}iastate.edu
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
During pregnancy, many growth factors produced by the uterus and the
fetus are believed to regulate growth and differentiation of fetal and
maternal tissues. The expression of growth factors such as insulin-like
growth factor, epidermal growth factor, heparin-binding epidermal
growth factor, transforming growth factor type 
, and
platelet-derived growth factor are regulated by estrogens (4, 5, 6, 7, 8, 9, 10, 11, 12, 13).
These growth factors are believed to be estromedins, mediating the
uterine proliferative response to estrogen. Estrogen-mediated uterine
proliferation occurs during each estrous cycle and can be easily
demonstrated in the ovariectomized mouse (14, 15). During pregnancy,
the uterus undergoes additional growth that is related to the presence
of the fetus. A potential source of this growth signal are the
MRP/PLFs, which are secreted by the placenta in midgestation (1, 16).
At the peak, the amount of MRP/PLFs produced during pregancy is
directly proportional to the number of fetuses (16). Thus, the MRP/PLFs
could provide a signal for uterine growth that is proportional to the
number of fetuses present.
Five mrp/plf messenger RNA (mRNA) sequences have been identified, one of which is an alternatively spliced form of one of the four identified gene products (17–20; Fassett JT, Nilsen-Hamilton M, Hamilton RT, manuscript submitted, GenBank accession no. AF 128884). The four mrp/plf genes are plf1, plf2, mrp3, and mrp4 (18–20; Fassett JT, Nilsen-Hamilton M, Hamilton RT, manuscript submitted, GenBank accession no. AF 128884). The sequences of the four proteins expressed from these genes vary in just a few positions. These few differences in amino acid sequence don’t enable the different proteins to be distinguished by antibodies, whereas the different mRNAs can be distinguished with specific oligonucleotide probes and by diagnostic RT-PCR, as used for this report.
The aim of the work described here was to identify the fetal-maternal interactions that might involve the actions of MRP/PLF family members. Here, we demonstrate that three mrp/plf genes (plf1,mrp3, and mrp4) are expressed in the placenta, with the most abundant mRNA being mrp3. During gestation, the mrp4 gene is expressed one day later than the mrp3 and plf1 genes. Whereas MRP3 and/or PLF1 leave the placenta and enter the bloodstream and the amniotic fluid, MRP4 is only found in the placenta. We also provide evidence that, in the second half of gestation, one or more signals from the uterus feed back to arrest production of MRP/PLF. This feedback results in a peak of MRP/PLF production in midgestation, which coincides with a large increase in uterine mass and DNA content in mid- to late gestation. The time period of expression of the mrp/plf genes relative to uterine growth is consistent with the hypothesis that one or more of these proteins may regulate uterine growth in vivo during pregnancy. This observation is consistent with our previous observation that PLF1 stimulates uterine cell proliferation (1).
From these and previous results, we propose that part of the maternal-fetal interaction in reproduction involves the production of MRP/PLFs by the placenta. These proteins stimulate uterine growth and placental vascularization. Our data also suggests that estrogens are not responsible for uterine growth during pregnancy. Cessation of uterine growth is correlated with a decrease in MRP/PLF concentrations in the placenta in late gestation. The decrease in MRP/PLFs is due to a maternal to fetal signal from the uterus, which feeds back to control MRP/PLF expression and to terminate the growth signal.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
Minced placental tissue culture and metabolic labeling with
35S-methionine
Freshly dissected placentae from pregnant CF1 mice between days
8 and 16 of gestation were rinsed in ice-cold HBSS and minced into
pieces of approximately 1 mm3. For primary placental tissue
culture, the minced tissues were washed with DMEM with 25
mM glucose. They were then incubated in DMEM with 2% FCS,
20 mM HEPES, 50 U/ml each of penicillin and streptomycin
for 16 h at 37 C in a humidified atmosphere containing 10%
CO2. After the incubation, the minced placentae were washed
by suspending them in a Tris-buffered, balanced salts solution (140
mM NaCl, 5 mM KCl, 0.68 mM
CaCl2, 0.49 mM MgCl2, 0.37
mM Na2HPO4, 0.25 mM
Tris x HCl, pH 7.4) and collected by centrifugation. The minced
and washed placentae were resuspended in DMEM (containing 1/10 the
normal methionine concentration), 100 µCi/ml
35S-methionine (NEN Life Science Products,
Boston, MA) or 35S-Trans label (ICN Radiochemicals, Irvine, CA), and 0.2% FCS. The placental
suspensions were dispensed into the wells of multiwell tissue culture
dishes (200 µl/well) and incubated at 37 C in an atmosphere of 10%
CO2 for 4 to 4.5 h with gentle rocking. The medium and
minced tissues were then removed and spun at 2,000 x g
for 20 min to separate the remaining tissue and the medium. The
supernatants were removed and used for immunoprecipitation. Duplicate
5–10 µl aliquots of the medium were removed to determine
acid-insoluble cpm by precipitation with 10% trichloroacetic acid.
These values were used to normalize the results of cpm incorporated
into secreted MRP/PLFs so as to calculate the specific changes in
secreted MRP/PLF.
Antisera
The antiserum 4269B was raised against MRP/PLFs purified from
3T3 cells. The antiserum, which was used for most RIAs and Western
blots, detects the four known MRP/PLFs, PLF1, PLF2, MRP3, and MRP4.
This detection was ascertained by analysis with the individual proteins
expressed from cloned complementary DNAs (cDNAs) in insect or mammalian
cells. The antisera, 89Rb11 and 89Rb13, were raised against MRP/PLFs
purified from the conditioned medium of BN/L cells. The anti-rPLF1
serum was a gift from David Denhardt and was raised against PLF1
expressed in Escherichia coli.
Endoglycosidase H and neuraminidase treatment
Samples of medium from 35S-methionine-labeled minced
placentae were incubated in 10 mM NaPi, 100 mM
NaOAc, pH 5.5 for 20 h at 37 C with or without 0.2 U/ml
endoglycosidase H (E.C. 3.2.1.96, Roche Molecular Biochemicals, Indianapolis, IN). Alternatively, samples were
treated in 0.9 mM CaCl2 10 mM NaPi,
100 mM NaOAc, pH 5.3 with or without 0.34 U/ml
neuraminidase (E.C.3.2.1.18) from Vibrio cholerae
(Sigma, St. Louis, MO). The reactions were stopped by
diluting the mixtures by 2-fold with twice concentrated buffer pH 7.4.
The MRP/PLF was then immunoprecipitated. The proteins in the
immunoprecipitates were resolved by SDS PAGE as described previously
(22).
Immunoprecipitation
MRP/PLF was immunoprecipitated from the conditioned medium of
metabolically labeled cells as described previously (23).
Immunoprecipitates were resolved by SDS-PAGE (22). The resulting gels
were impregnated with PPO, dried, and exposed to film.
RIA for MRP/PLFs
MRP/PLF standards or samples were incubated with anti-MRP/PLF
antibody (1:5000) in 1.5 ml microcentrifuge tubes in buffer containing
10 mM sodium chloride, 10 mM EDTA, 1%
crystalline BSA, 10 mM NaPi, pH 7.5, for a period of 6
h at 4 C. Then 125I-MRP/PLF (approximately 15,000 cpm/tube)
was added, and the mixture (250 µl) was incubated for another 16
h at 4 C. The antigen-antibody complex was separated from the free
125I-MRP/PLF by incubating the solution with 1% (wt/vol)
pansorbin (24, 25) for 1 h at room temperature, followed by
centrifugation for 10 min at room temperature. The supernatant was then
carefully removed, and the radioactivity of the pellet and the
supernatant were measured in a GammaTrac 1191 radiospectrometer. The
assay sensitivity, calculated as the signal that is 2 SDs
away from the signal for the zero dose, was 2.8 ng/ml (0.7 ng/tube)
MRP/PLF. The cpm in the pellet from reactions containing 1,000 ng/ml
MRP/PLF was considered to be the nonspecific binding, and it was
subtracted from the cpm of all other pellets to obtain the specifically
bound 125I-MRP/PLF. The cpm in pellets from reactions
containing no added nonradioactive MRP/PLF was considered to be the
maximum binding. In a typical experiment 30% of
125I-MRP/PLF was bound in the absence of added MRP/PLF
present in the reaction mixture (1,000 ng/ml). When normal rabbit serum
was used instead of anti-MRP/PLF antiserum at 1:5,000 dilution, 4.5%
of 125I-MRP/PLF was bound. The linear range of the assay is
1–100 ng/ml of MRP/PLF.
The within-assay coefficient of variation in the assay was estimated by assaying a pooled sample of maternal plasma three times in duplicates in the same assay, and was 2.2 ± 0.22 µg/ml (mean ± SD). The between-assay coefficient of variation estimated by assaying the same sample in three different assays was 2.6 ± 0.74 µg/ml.
Tissue extracts
Placentae were dissected from CF1 mice and homogenized in 0.5%
Nonidet P40, 1% Trasyol, 10 mM phenylmethylsulphonyl
fluoride, 20 mM HEPES, pH 7.4 (200 mg tissue per 100 µl
buffer) at room temperature for 20 sec using a Tekmar tissumizer. The
homogenate was placed on ice and then centrifuged at 133 x
g for 15 min at 4 C. The supernatant was removed and spun at
13,000 x g at 4 C for 20 min. An aliquot of the
supernatant was removed for protein analysis by the Bradford method
(26) using reagents from Pierce Chemical Co. (Coomassie
Plus Protein Assay Reagent, Pierce Chemical Co., Rockford,
IL) and the remainder was frozen for later Western blot analysis.
Uteri were dissected from CF1 mice. The tissue was weighed and then homogenized at 4 C using a Tekmar tissumizer at a ratio of 1 g tissue per 5 ml of 0.3 M sucrose, 1 mM phenylmethylsulfonyl fluoride. The homogenate was spun three times sequentially at 3,000 x g for 20 min, 13,000 x g for 20 min, and then at 100,000 x g for 90 min. The resulting supernatant was removed, dispensed in aliquots, and stored at -70 C.
Western blot analysis
MRP/PLFs in blood sera, amniotic fluids, and extracts of
placentae were detected by Western blot analysis. Samples were first
resolved by SDS-PAGE (7.5–15% linear polyacrylamide gradient gels) at
12.5 mAmp/gel for 5 h at room temperature (22). The proteins were
transferred to a nitrocellulose membrane using a Hoefer TE
50 apparatus at 90 V for 4 h with water cooling. The membrane was
stained with 0.2% Ponceau S in 3% trichloroacetic acid to ascertain
uniform transfer of proteins to the membrane. After destaining with
water, the membrane was incubated for 12 h at 4 C in blocking
solution (5% (wt/vol) nonfat milk in 140 mM sodium
chloride, 5 mM potassium chloride, 0.4 mM
sodium phosphate, 0.02% (wt/vol) sodium azide, 25 mM Tris,
pH 7.4). The blocking solution was then decanted and the membrane was
soaked in a fresh blocking solution containing anti-MRP antiserum to a
final dilution of 1/200. The membrane was washed three times with 0.14
M NaCl, 0.4 mM NaPi, 5 mM KCl, and
25 mM Tris x HCl, pH 7.45 for 10 min each. The second
wash buffer also contained 0.05% (vol/vol) NP-40. Finally, the
membrane was incubated in a fresh blocking solution containing
1–5 x 105 cpm/ml of 125 I-protein A.
After an additional three washes, as described above, the membrane was
dried and the radioactivity detected by autoradiography.
Quantitation of radioisotope incorporated into individual protein
bands
Fluororadiograms of Western blots were analyzed quantitatively
from density scans obtained with an LKB Zeinah
scanning densitometer (LKB Produkter AB, Bromma, Sweden).
The areas under the scanned curves were determined by planimetry or
with the integrator built into the densitometer. Alternatively, the
amount of 125I radiolabel was determined in individual
protein bands by use of a phosphorimager. The linearity of each assay
was verified by measuring the output from increasing dilutions of
plasma containing MRP/PLF.
Quantitation of tissue DNA content
Uteri were dissected from CF1 female mice at various stages of
gestation; each sample included both horns. The total wet weight was
recorded and the tissue was frozen. For analysis, sections were
randomly cut from the frozen tissue, weighed, and then analyzed for
DNA/g wet weight according to Burton (27). The total DNA of each uterus
was determined by multiplying the DNA/g by the total uterine
weight.
RNA and diagnostic RT-PCR
Frozen tissue was pulverized in liquid nitrogen using a mortar
and pestle, and RNA was isolated using Tri-Reagent (Life Technologies, Inc., Rockville, MD) according to the
manufacturer’s instructions. For RT, 250 ng of total RNA was incubated
in 5 µl reaction mixture containing 75 mM KCl, 3
mM MgCl2, 10 mM DTT, 50
mM Tris-HCl (pH 8.3 at RT) 112 ng poly dT primers, and 200
µM dNTPs, with or without 40 U Superscript II reverse
transcriptase (Life Technologies, Inc.). The reaction
mixture was incubated at 42 C for 50 min, 50 C for 10 min, then 75 C
for 15 min. For PCR, the following primers were used:
mrp/plf downstream exon I; 5'CTCTGCAGAGATGCTCCCTTC3' (+59 to
+79), mrp/plf upstream exon V; 5'CATGATATTTCAGAAGCAGAGCAC3'
(+778 to +756). These primers are complementary to the identical
sequences in all known mrp/plf cDNAs. Briefly, 0.5 µl cDNA
was amplified in a 25 µl reaction containing 20 pmoles of each
primer, 200 µM dNTPs, 1.5 mM
MgCl2, 50 mM KCl, 10 mM Tris-HCl
(pH 8.3) and 1 U Taq Polymerase. The thermal cycling
conditions were as follows: 95 C for 2 min to denature, then 40 cycles
at 95 C (30s), 66 C (1 min), and 72 C (1.5 min) to amplify with a final
3 min at 72 C for extension. The PCR products were resolved by
electrophoresis through a 1.5% agarose gel and analyzed by ethidium
bromide staining.
The diagnostic RT-PCR assay for the mrp/plfs is based on
differences in the cDNA sequences of these closely related gene
products which result in variations in restriction sites within the
cDNAs. To distinguish the mrp/plf mRNAs using restriction
digestion, one fifth of the reaction mixture from the RT-PCR reaction
was subjected to one round of PCR using the same primers as for the
original amplification but with 62.5 µM dNTPs, 0.9
mM MgCl2, and 1 µCi 32P-CTP in
a final 25 µl volume. The sample was heated to 98 C for 4 min, 64 C
for 1 min, then 72 C for 20 min. The radiolabeled cDNA was precipitated
with 70% ethanol in the presence of 10 µg yeast transfer RNA as
carrier. Digestion was carried out using 1 U each of BsoFI (New England Biolabs, Inc., Beverly, MA) and BstXI at 55 C
overnight, or 5 U of NdeI at 37 C including 100 µg/ml BSA.
The radiolabeled and digested products were resolved by electrophoresis
through a 2% agarose gel or an 8% polyacrylamide gel. The gels were
dried, and the amount of radioactivity associated with each band was
determined using a phosphorimager with ImageQuant software
(Molecular Dynamics, Inc., Sunnyvale, CA). Positive
controls for identification of specific mrp/plf cDNAs were
cloned cDNAs or mRNA isolated from COS cells that had been transiently
transfected with specific cDNAs. The ability of the procedure to
determine the relative concentrations of the four mRNAs was tested by
making mixes of the standard cDNAs in various ratios, amplifying and
then determining the ratio of the cDNAs in the final products. The
results showed that this method faithfully reproduced the starting
ratios of the mrp/plf cDNAs in the original mixture.
Estradiol-17ß analysis
Nonpregnant mice (n = 2) and mice of known gestational ages
(n = 18) were killed by cervical dislocation, and a sample of
systemic blood was collected by heart puncture immediately after the
mice were killed. In addition, the gravid uterus was removed, and
samples of both placental and uterine tissue associated with each
conceptus were collected, weighed, then immediately frozen under liquid
nitrogen. Tissue samples were homogenized in 1 ml deionized water using
a tissue TissueTearor homogenizer. Blood plasma and tissue homogenates
were frozen at -20 C until assayed for estradiol-17ß by RIA. Before
RIA, aliquots of all placental homogenates, and aliquots of all uterine
homogenates from a pregnant female were pooled to form a single
placental tissue pool and a single uterine tissue pool. An aliquot of
each tissue pool (placental and uterine) was then extracted and assayed
for estradiol-17ß (28) as previously described.
Estradiol-17ß in systemic plasma was also quantitated by RIA as previously described (29). All tissue or plasma samples were evaluated together in the same assays. Intraassay coefficients of variation for estradiol-17ß using tissue or plasma pools were 3.3% and 7.4%, respectively. Values were normalized to mg wet weight of tissue.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
).
|
). The estimated Mr’s of
these two forms of MRP/PLF as determined by SDS-PAGE were 38 ± 1
(n = 6) and 27 ± 1 (n = 5). Only the higher molecular
weight form was found in the bloodstream and in the amniotic fluid.
Both forms were found in the placentae, and they were both secreted by
primary minced placental cultures. Three different polyclonal
antibodies raised against MRP/PLF identified the same two forms of
MRP/PLF (Fig. 3). Each antiserum had been
raised against a different preparation of MRP/PLF; two of the protein
preparations were isolated from eukaryotic cells, one from BN/L cells
and the other from tumor-promoter treated Swiss 3T3 cells. The third
preparation was a recombinant PLF1 protein from transformed
Escherichia coli. The four antisera showed the same two
bands of MRP/PLF proteins at 38 and 27 kDa, although the relative
intensities of the two bands varied for each antiserum, particularly
for the antiserum against the recombinant bacterial protein compared
with that for the purified protein produced in mammalian cells. These
differences probably reflect the impact of glycosylation on the
recognition by the antibodies of protein epitopes as the bacterial
protein was not glycosylated whereas the other proteins used as
antigens were the highly glycosylated 38-kDa forms.
|
|
). The sum of both MRP/PLFs results in
the pattern of placental expression shown in Fig. 1
.
|
, B and C;
21). Although plf2 was cloned from the placenta (19), we
were unable to detect its mRNA in placental RNA samples of CF1 or
Balb/c mice (Fig. 4C). The gestational profile of the 38-kDa protein
matches the profiles of expression of mrp3 and
plf1 mRNAs, and the profile of the 27-kDa protein matches
the profile of mrp4 mRNA (Fig. 4). In each case, the mRNA
level increased the day before the increase in protein level was
observed. The molecular sizes of the identified protein products, 38
kDa for PLF1 and MRP3 and 27 kDa for MRP4, are consistent with the
expected sizes of these proteins as determined by their expression from
cDNAs in cultured cells (21).
Carbohydrate components of the MRP/PLFs
Because carbohydrate modifications can influence protein stability
and access to particular tissues and cellular compartments, we
examined the carbohydrate components of the two MRP/PLF protein forms.
According to the cDNA sequences, all three mrp/plf genes
encode polypeptides of the same size. The differences in their apparent
molecular weights are expected to be a function of their different
N-glycosylation patterns (Fassett JT, Nilsen-Hamilton M, Hamilton RT,
manuscript submitted, GenBank accession no. AF 128884). To establish
that the two MRP/PLF protein bands observed in the placenta correspond
to the same-sized polypeptide chains (as expected from the cDNA
sequences), minced placental cultures were labeled with
35S-methionine in the presence and absence of the inhibitor
of glycosylation, tunicamycin. Medium was collected from the labeled
cells and the MRP/PLF immunoprecipitated (Fig. 5). In the presence of tunicamycin, both
protein bands collapsed to a single apparent molecular weight at
21K as expected from the cDNA sequences.
|
). The molecular weight shift of the 27K form after
neuraminidase treatment was very small and was only observed with the
secreted protein. In comparison, the molecular weight of cathepsin L,
another protein secreted by placentae and immunoprecipated from the
same samples, was not shifted by treatment with neuraminidase (Fig. 5).
Thus, all the MRP/PLFs produced by the placenta are sialylated, but the
38K protein is more heavily sialylated than the 27K protein.
Different intracellular processing rates for MRP4 vs. PLF1 and
MRP3
The completion of glycosylation occurs after protein synthesis and
during the time that the protein is trafficking to the surface membrane
for secretion. Thus, a different glycosylation profile might be
accompanied by a different time period before release. The lag-times
for secretion of the two forms of MRP/PLF were compared in minced
placental cultures by pulse-labeling the cultures with
35S-methionine for 15 min and then measuring the secreted
proteins as a function of time over the chase period which was
inititated by replacing the radiolabeling medium with growth medium
lacking radioisotope. There was a significantly longer delay before
secretion of the 27-kDa protein (MRP4) than the 38-kDa protein (PLF1
and MRP3) (Fig. 6). From three different
sets of time courses, the average time after the beginning of labeling
before the 27-kDa MRP/PLF began to be secreted (intercept of the
extrapolated line with the abscissa) was 41 ± 4 min; for the
38-kDa protein it was 25 ± 8 min. The 27-kDa protein, which
appears in the medium later than the 38-kDa proteins, was not likely to
be a product of the 38-kDa proteins because incubation of the medium
containing a mixture of the two proteins with the cells did not result
in conversion of the 38-kDa protein to the 27-kDa protein (data not
shown). Thus, surprisingly, the 27-kDa protein (MRP4) with the least
amount of glycosylation took the longest to traffic to the surface
membrane.
|
, top panel). The rate of increase in uterine DNA content
measured between days 9 and 14 was 0.2 µg DNA/uterus/day, which
corresponds to an increase of 10–20% of the uterine mass per day.
This profile of uterine increase is consistent with the hypothesis that
local MRP/PLFs promote uterine growth.
|
, bottom panel) showed that the gestational profiles of
estradiol-17ß in these tissues and in the blood stream did not
correlate with the period of uterine proliferation during
gestation.
Effect of steroids on MRP/PLF production by the placenta
We examined the question of whether the MRP/PLFs were induced in
minced placental cultures after treatment for 24 h with
estradiol-17ß in the concentration range of 10-6 to
10-9 M. The MRP/PLF proteins were detected by
Western blot analysis. There was no increase in the amounts of the
38-kDa or 27-kDa proteins found in the medium or in the cells after
estrogen treatment (data not shown).
Regulation of placental expression of MRP/PLF in minced placental
cultures
The percent of trophoblasts that contain MRP/PLF and the MRP/PLF
content of the placenta drops precipitously after day 13 (Fig 1; 31).
To determine whether the signal(s) for this decrease in MRP/PLF
expression come from the placenta or from another source, we removed
placentae from animals at different days of gestation and cultured them
in minced tissue culture for 16–18 h. The cultures were then labeled
with 35S-methionine and the secreted, metabolically labeled
MRP/PLFs were immunoprecipitated from the medium. The increase in
MRP/PLFs from day 8–10 in vivo was reflected in these
in vitro cultures by the increase in production of the
27-kDa and 38-kDa proteins by placental minces of
increasing gestational ages from days 7–10.
However, the parallel between the in vivo and in
vitro results broke down after gestational day 13. Whereas the
amount of the 38-kDa MRP/PLF in the placenta in vivo
decreased after day 13 (shaded area in Fig. 8), the rate of synthesis of MRP/PLF by
the placental minces was about the same in all minced placental
cultures through day 17 (Fig. 8). By contrast, the level of expression
of the 27-kDa MRP4 protein decreased after day 13 in the minced tissue
cultures in the same way as was observed in the placenta in
vivo (Fig. 8, right panel). Thus, it seems that an
external signal or combination of signals is responsible for the
decrease in 38-kDa MRP/PLF expression in the latter half of gestation,
whereas the signal(s) necessary for regulating expression of MRP4 are
present in the isolated placental tissue.
|
). However, strips of the uterus taken
from regions of placental-uterine contact suppressed MRP/PLF
production. The effect of the uterine cocultures was 2-fold. First,
they suppressed the secretion of all 35S-methionine
proteins (Fig. 9, left panel). Second, the uterine strips
specifically suppressed the secretion of MRP/PLFs by late gestational
placentae (Fig. 9, right panel; values normalized to the
total secreted 35S-methionine-labeled proteins). Day 16
placentae responded to uterine coculture with a much larger specific
decrease in MRP/PLF production than did day 10 placentae. Late
gestational uterine strips were more effective in inhibiting MRP/PLF
production than were strips from midgestational uteri (Fig. 10). However, there was also inhibition
by midgestational uteri. Conditioned medium from cultured uterine
strips did not inhibit 38-kDa MRP/PLF production by the minced
placental cultures. But, uterine extracts inhibited MRP expression with
late extracts from gestational uteri being more effective (Fig. 11). It is known that the decrease in
MRP/PLF protein in vivo accompanies a decrease in
mrp/plf mRNA in the placenta (16). Thus, it appears that the
uterus produces one or more compounds that inhibit expression of the
mrp/plfs by the placenta.
|
|
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The expression of the mrp4 gene is regulated differently from plf1 and mrp3. The latter genes are expressed in the placenta with different gestational profiles and reach their peak expression earlier than mrp4. Differences in mechanisms regulating the expression of mrp4 are also evident in that the gestational profile of MRP4 expression is maintained in the isolated placentae, whereas the negative regulation on 38-kDa MRP/PLF (MRP3 and PLF1) expression is lost after the placenta is separated from the uterus. Another difference between the proteins is in their trafficking from the cells. Despite its being less glycosylated than PLF1 and MRP3, the time for MRP4 processing and secretion by the cells is longer than for the 38-kDa MRP/PLFs. Either MRP4 takes a different route through the cell’s secretory network or it is produced by a different subset of trophoblastic giant cells than produce PLF1 and MRP3.
The placental content of MRP/PLFs remains high from days 10 through 13, after which it drops off rapidly. This profile is consistent with the results of earlier studies in which the MRP/PLF content of the placenta was measured in terms of the number of giant cells that contain the protein as measured by immunohistochemistry (31). The increase in MRP/PLF production parallels the growth of the placenta and is likely due to increased numbers of trophoblastic giant cells which express mrp/plfs.
MRP/PLFs stimulate DNA synthesis in cultured primary uterine cells (1).
Here we show that the period of MRP/PLF production in vivo
is correlated with the period of maximum growth of the uterus. The
MRP/PLF begins to be observed in the blood on day 9. The
highest-circulating levels of MRP/PLF are found on day 10 (Fig. 1). The
placental levels of MRP/PLF remain high until day 13, after which the
MRP/PLF content of the placenta decreases and the rate of uterine
growth slows the next day (32, Fig. 8).
Estrogen stimulates uterine growth in immature females and ovariectomized adult females (4, 5, 6, 7, 8, 9, 10, 11, 12, 13). Thus, although the MRP/PLF levels correlated in time with the growth period of the uterus, it was expected that uterine growth might be regulated by estrogens. Consequently, we determined the profiles of estradiol-17ß in the uterus, placentae, and blood. The results showed no correlation of the uterine growth period with the profile of estradiol in any of these tissues. Our results for the circulating levels of estradiol are consistent with those previously published (33).
Estrogen, derived from the ovaries, is required for implantation in the mouse (34). The very high concentration of uterine and placental estradiol-17ß content on day 6 that has decreased by day 8 is probably relevant to this requirement. The mechanism by which estrogen acts to permit implantation probably involves the local induction of polypeptide growth factors in the uterus (4, 5, 6, 7, 8, 9, 10, 11, 12, 13). However, this same mechanism does not seem responsible for uterine growth during pregnancy as tissue and circulating estradiol levels are low during the period of uterine growth and do not rise in the pregnant mouse until after uterine growth has ceased. Thus, the MRP/PLFs rather than estrogen are the more likely immediate regulators of uterine growth in midgestation.
The large increase in the estradiol content of the placenta and uterus around the time of implantation is followed 3 days later by a rapid increase in mrp/plf expression in the placenta. Thus, although this is a signficant time span between the two events, it is possible that the earlier increase in placental estradiol-17ß content stimulates the mid-gestational increase in MRP/PLF production by the placenta. However, this mechanism is unlikely, as we found that the production of MRP/PLF is not stimulated by estradiol-17ß in minced placental cultures.
The close correlation between the presence of MRP/PLFs in the placenta, the growth of the uterus, and the ability of MRP/PLFs to stimulate uterine cell proliferation suggests that one or more MRP/PLF regulates proliferative growth of the uterus during gestation. MRP/PLF production is proportional to the amount of fetal placental tissue, and our data suggests that it is independent of circulating estrogen levels. Thus, the MRP/PLFs would provide a signal for uterine growth in proportion to the number of fetuses present and link uterine growth to fetal mass (16).
The decrease in mrp/plf expression in late gestation seems to be largely due to feedback regulation from the uterus. Our studies of MRP/PLF production by placental minces showed that uterine extracts and strips of uterine tissue can inhibit the placental production of the MRP/PLFs in late gestation. As the strips of tissue are not in contact with the majority of the cells in the minced placental tissue, the uterine factor is presumed to be a soluble, secreted factor. Attempts to demonstrate that this factor was in the growth medium from cultured uterine cells were unsuccessful. This latter result could suggest that the factor is not secreted by uterine cells or that it is very unstable. However, not all uterine cells survive in primary culture, and dispersed cultured cells do not express all genes that are expressed in the intact tissue. Thus, we conclude, in the absence of evidence to the contrary, that the uterine inhbitory factor is probably secreted by the intact uterus. The effect of the uterine factor is specific for MRP/PLFs over other secreted proteins. In the absence of this uterine influence, the in vitro rate of production of MRP/PLF was not lower in placentae of increasing age. The uterus contains receptors for MRP/PLFs (1). Thus, it seems reasonable to propose that uterine inhibition of mrp/plf expression is part of a feedback control loop which prevents further uterine growth into late gestation.
In summary, we have demonstrated different gestational profiles of three mrp/plf genes and their protein products. One of these proteins, MRP4, does not seem to leave the placenta and is a good candidate for the proposed local placental angiogenic activity of this protein family. We also describe a newly observed uterine-placental communication in the mouse in which the uterus feeds back to inhibit mrp/plf expression. The period of maximal uterine growth is coincident with the presence of MRP/PLF. This, and our previous observation that PLF1 stimulates primary uterine cell growth, suggests that PLF1 and perhaps also MRP3 and MRP4 stimulate murine uterine growth in midgestation. The lack of correlation of uterine growth and estradiol-17ß suggests that during pregnancy, estrogens are not responsible for uterine growth. As well, estradiol-17ß does not seem responsible for the increase in mrp/plf expression by the placenta. In the second half of gestation, the uterus produces an inhibitory agent that suppresses MRP/PLF production by the trophoblasts with a resulting decrease in MRP/PLF secretion. The decrease in MRP/PLF levels is correlated with a decrease in uterine growth rate. Thus, we propose that the MRP/PLFs are the growth factors that mediate mid-gestational uterine growth. Because they are produced in proportion to fetal mass, the MRP/PLFs provide a signal to the uterus for proportional growth to accommodate the number of fetuses that it contains. In turn, we propose that the uterus feeds back to shut off MRP/PLF production in late gestation to halt further uterine growth.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
2 Current address: Coron Biomedics, 755 Stanford Circle, Rochester
Hills, Michigan 48309.
3 Current address: Muta University, P.O. Box 7, Alkarak,
Jordan.
Received February 26, 1999.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
and ß-subunits in the mouse
uterus and vagina: potential mediator of estrogen action. Endocrinology 136:2325–2340[Abstract]
is a potential mediator of estrogen action in the mouse
uterus. Endocrinology 131:1657–1664[Abstract]
This article has been cited by other articles:
 |
|
 |
|
  K. A. Longo, J. A. Kennell, M. J. Ochocinska, S. E. Ross, W. S. Wright, and O. A. MacDougald Wnt Signaling Protects 3T3-L1 Preadipocytes from Apoptosis through Induction of Insulin-like Growth Factors J. Biol. Chem., October 4, 2002; 277(41): 38239 - 38244. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. J. Toft, S. B. Rosenberg, G. Bergers, O. Volpert, and D. I. H. Linzer Reactivation of proliferin gene expression is associated with increased angiogenesis in a cell culture model of fibrosarcoma tumor progression PNAS, October 16, 2001; (2001) 231364798. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. T. Fassett and M. Nilsen-Hamilton Mrp3, a Mitogen-Regulated Protein/Proliferin Gene Expressed in Wound Healing and in Hair Follicles Endocrinology, May 1, 2001; 142(5): 2129 - 2137. [Abstract] [Full Text]
|
|
|
|
|
|
J. T. FASSETT, R. T. HAMILTON, and M. NILSEN-HAMILTON Mrp4, A New Mitogen-Regulated Protein/Proliferin Gene; Unique in this Gene Family for its Expression in the Adult Mouse Tail and Ear Endocrinology, May 1, 2000; 141(5): 1863 - 1871. [Abstract] [Full Text] [PDF]
|
|
, 
represent two
different sets of MRP/PLF standards. Bottom panel,
Samples of plasma and placental extracts were analyzed by RIA to
determine levels of MRP/PLFs. The data set used for Western blot
analysis is a subset of the same samples as assayed by RIA. The
placental samples used in the RIA analysis were different from those
used for the Western blot analysis. The number of values contributing
to the averages for the plasma are: 3 (days 16, 17), 5 (days 8, 9, 15),
and 6 (days 10–14). Placental values are each the average of two
estimates. The inset shows a typical standard curve.
•, plasma; 
, 38-kDa MRP/PLFs. B, Diagnostic RT-PCR was
performed on total RNA samples isolated from placentae at various
stages of development as described in Materials and
Methods. The results of one of three independent data sets is
shown. To obtain the values shown, the proportion of each
mrp/plf was multiplied by the total amount of
mrp/plf mRNA determined from Northern blot analysis of RNA
isolated from each stage of gestation (
), mrp3
(
) and placenta (
