Human Reproduction Vol.19, No.6 pp. 1272±1280, 2004
Advance Access publication April 29, 2004
DOI: 10.1093/humrep/deh256
Progesterone withdrawal causes endothelin release from
cultured human uterine microvascular endothelial cells
MaÊns Edlund1,2, Eva Andersson1 and Gabriel Fried1
1
Department of Women and Child Health, Division of Obstetrics and Gynaecology, Karolinska Institutet and Hospital,
S-171 76 Stockholm, Sweden
2
To whom correspondence should be addresssed. E-mail: [email protected]
BACKGROUND: Current theories on the physiology of menstrual bleeding in humans offers an explanation for the
shedding of the endometrium as a result of a breakdown of the extracellular matrix due to an in¯ammatory reaction. The link between the fall in progesterone levels and these events is not clear. Neither has an explanation been
presented for the vasoconstriction in the coiled arteries occurring during menses. We have hypothesized a chain of
events where the fall in progesterone levels induces an upregulation of the thrombin receptor in the small uterine
arteries leading to an increased thrombin response and subsequent endothelin release. METHODS: Endothelial
cells from human umbilical cord (HUVECs) and from human small uterine arteries (UtMVECs) were cultured
under conditions partly resembling the female hormonal cycle with progesterone withdrawal. RESULTS: Following
progesterone increase and subsequent withdrawal, we found an increased production of thrombin receptor and an
increased release of endothelin from UtMVECs compared with HUVECs. CONCLUSION: Endothelin release in
response to progesterone withdrawal in UtMVECs can offer an explanation for the vasoconstriction seen in the
coiled arteries during menses in humans.
Key words: coiled artery/endothelin/menstrual bleeding/progesterone/thrombin receptor
Introduction
The monthly human menstrual bleed is an intricately and
highly regulated event involving the endocrine, cardiovascular
and haemostatic systems. Despite enormous advances in
molecular biology during the last decades, details of the
regulation of menstrual bleeding are still not known. The
shedding of the endometrium follows in the absence of or at
failed implantation. It is preceded by the demise of the corpus
luteum and, as a consequence thereof, a fall in circulating
progesterone. The progesterone withdrawal initiates a
sequence of events that ultimately lead to the shedding of the
endometrium and menstrual bleeding.
Past and present theories regarding the mechanisms of
menstruation involve, to some extent, vasoconstriction in the
coiled arteries (also called spiral arteries) of the endometrium.
In the early works by Markee, this constriction was suggested
to precede menstrual bleeding (Markee, 1940). For >60 years,
scientists have searched with increasingly advanced methods
for a link between progesterone withdrawal and the regulation
of menstrual bleeding. However, the link between progesterone
withdrawal and the subsequent coiled artery vasoconstriction
has been poorly studied and is not at all clear. Many factors
have been suggested to be involved; hormones, enzymes,
in¯ammatory mediators, haemostatic factors and growth
factors. We have focused our interest on the 21 amino acid
polypeptide endothelin (ET). Endothelins consist of a group of
1272
three peptides, ET-1, ET-2 and ET-3, with potent biological
properties. ET-1 is produced mainly by endothelial cells and is
the most potent vasoconstrictor peptide known to date
(Yanagisawa et al., 1988). Other tissues produce ET-2 and
ET-3.
The ETs appear to act primarily by a direct effect on vascular
smooth muscle cells (Gude et al., 1991). Thus, the main site of
action of ET in the endometrium is likely to be the vascular
smooth muscle cells of the small endometrial vessels.
Normally, circulating ET levels, as well as production of the
peptide in isolated blood vessels, are rather low due to the
absence of stimuli and the presence of potent clearance
mechanisms. Important stimulators of ET production are
thrombin, angiotensin, arginine, vasopressin and transforming
growth factor-b, as well as certain cytokines and physicochemical factors such as hypoxia (Luscher et al., 1993).
We hypothesized that uterine microvascular endothelial
cells might release endothelin in response to progesterone
withdrawal, and that ET therefore could be one agent that
contributes to endometrial vessel contraction. The resulting
endometrial ischaemia could then be one of the regulators of
menstrual bleeding.
Menstrual bleeding also intimately involves the coagulation
system. As a candidate for the link between progesterone
withdrawal, ET release and the coagulation system, we have
studied thrombin. Progesterone has been found to regulate
Human Reproduction vol. 19 no. 6 ã European Society of Human Reproduction and Embryology 2004; all rights reserved
Progesterone withdrawal induces endothelin release
synthesis of the G protein-coupled protease-activated receptor1 (PAR-1), the receptor for thrombin, in the human
endometrium (Hague et al., 2002) and in the arterial wall of
the aorta (Herkert et al., 2001). Thrombin is known to
upregulate endothelin-converting enzyme (ECE-1) in human
endothelial cells (Eto et al., 2001). Furthermore, thrombin is
one factor that stimulates ET release, as shown by several
authors (Yanagisawa and Masaki, 1989; Ohlstein and Storer,
1992). One endothelial cell response to stimulation by
thrombin through the thrombin receptor is release of the
contents of the Weibel±Palade bodies. These contain several
locally effective proteins, including von Willebrand factor
(Wagner et al., 1982) and ET (Russell et al., 1998). Thrombin
also induces secretion of another storage particle containing
tissue plasminogen activator (tPA) (Emeis et al., 1997).
Following progesterone withdrawal, there is an increase of
tPA (Lockwood and Schatz, 1996) possibly as a response to an
increased thrombin response.
To test the hypothesis that endothelial cells in the uterine
small arteries might upregulate thrombin receptors and release
ET in response to progesterone withdrawal, we have cultured
uterine microvascular endothelial cells (UtMVEC-Myos) in
the presence of estrogen and progesterone, and analysed ET
levels in the culture supernatants as well as thrombin receptor
occurrence in the cells under conditions simulating progesterone withdrawal. In order to compare the reactions to
progesterone withdrawal in UtMVECs with another endothelial cell type commonly used for studies of endothelial
production of vasoactive substance, we also performed the
same experiment in human umbilical vein endothelial cells
(HUVECs). There are documented differences between the two
endothelial cell types (Lang et al., 2003).
Materials and methods
Cells and cell culture
HUVECs (passage 243) and UtMVEC-Myos (passage three) were
from BioWhittaker (Cambrex, Walkersville, MD). Veri®cation of
endothelial cell purity was determined by immunoreactivity for CD31 and lack of immunoreactivity for human smooth muscle actin. The
culture media used were Endothelial Cell Growth Medium BulletKit-2
(EGM-2) and Microvascular Endothelial Cell Growth Medium-2
(EGM-2 MV), respectively (BioWhittaker, Cambrex, Walkersville,
MD). These media contain endothelial cell growth medium with
human recombinant epidermal growth factor, human ®broblast growth
factor, vascular endothelial growth factor, ascorbic acid, hydrocortisone, human recombinant insulin-like growth factor, heparin,
gentamicin, amphotericin-B and fetal bovine serum (FBS; 2% for
HUVECs, 5% for UtMVECs). Trypsinization was performed with
Reagent Pack containing trypsin/EDTA, trypsin neutralizing solution
and HEPES-buffered saline solution (BioWhittaker). Culture was
performed in 25 cm2 cell culture ¯asks from Corning (New York) and
cultures were seeded with 62 500 (HUVEC) to 140 000 (UtMVECMyo) cells. After 4 days, the cells were con¯uent to ~75% and each
¯ask contained 0.23±1.35 3 106 cells/ml.
Treatment of cells
Each culture was done in three ¯asks simultaneously with the same
batch of medium and with treatments in parallel. The cultures were
grown for 4 days, the medium volume was 5 ml. The experiment was
repeated once for each cell line. The three treatments used were as
follows: (i) control in medium without hormone addition for 4 days;
(ii) treatment with estrogen 500 pmol/l for 4 days and progesterone
50 nmol/l for 4 days; and (iii) treatment with estrogen 500 pmol/l for 4
days and progesterone 50 nmol/l for the ®rst 2 days only (progesterone
withdrawal). After 4 days, the cells were harvested, centrifuged at
500 g for 10 minutes in room temperature to a pellet, counted and
frozen at ±80°C.
Drugs used for treatment were as follows: estrogen in the culture
was added as a 17b-estradiol:HBC (2-hydroxypropyl-b-cyklodextrin)
complex (catalogue no. E-126), and progesterone was added as 4
pregnene-3,20 dione (catalogue no. P-8783) from RBI Sigma
Chemical Co (St Louis, MO). The estrogen preparation is a watersoluble complex, whereas the progesterone was dissolved in
dimethylsulphoxide (DMSO) (<0.0005% in ®nal solution) (Merck,
Darmstadt, Germany). DMSO was also added in a corresponding
amount in sham solution when progesterone was not added.
Biochemical analyses
The medium in each ¯ask was changed each day and kept frozen at
±18°C until analysed for levels of ET, thrombin and added hormones.
Endothelin. ET concentration was measured with Quantiglo
Chemiluminescent Immunoassay from R&D Systems. With antibodies against human ET-1, the sensitivity limit is <0.16 pg/ml, and
cross-reactivity with ET-2 and ET-3 is 27 and 8%, respectively.
Thrombin. The level of thrombin in culture medium was determined
by use of the chromogenic substrate S-2238 (Hemochrom Diagnostica
AB, MoÈlndal, Sweden) (Axelsson et al., 1976). Optical densities were
measured in a spectrophotometer at 405 nm.
Estrogen and progesterone. Estrogen and progesterone concentrations
were analysed using a competitive immunoassay, Immulite, from
Diagnostic Products Corporation (LA, USA).
Immunoblotting
After solubilization in 1% SDS (RBI Sigma), thrombin receptors in
the cell pellets were analysed by immunoblotting using a mouse
monoclonal IgG antibody raised against amino acids 42±55 of
thrombin receptor of human origin (Santa Cruz Biotechnology, Santa
Cruz, CA). The semi-quantitative analysis of the amount of thrombin
receptor was done using a ChemiDoc Gel Documentation System with
the software Quantity One version 4.4.0 from Bio-Rad (Hercules,
CA).
Immunohistochemistry
We analysed cells cultured on Falcon culture slides (Becton Dickinson
Labware, NY). The slides were ®xed using 4% paraformaldehyde and
0.03% picric acid, and kept in phosphate buffer pH 7.4 (PBS) in a
refrigerator until analysed. Monoclonal mouse antibodies speci®c for
endothelial cells were anti-human CD31 (clone JC70A code no.
M 0823 lot 110, Dako, Denmark). As negative control, we used antihuman smooth muscle actin (clone 1A4 code no. M 0851 lot 070,
Dako) and anti-human ®broblast prolyl 4-hydroxylase (clone 5B5
code no. M 0877 lot 029, Dako). For analysis of the thrombin receptor
(PAR), the cells were labelled with mouse monoclonal IgG antibody
raised against amino acids 42±55 of thrombin receptor of human
origin (ATAP2 sc-13503, Santa Cruz Biotechnology, Inc.). For
analysis of the progesterone receptor, we used mouse monoclonal IgG
antibodies raised against the N-terminal region of the A form of the
human progesterone receptor (NCL-PGR-312; Novocastra, Newcastle
1273
M.Edlund, E.Andersson and G.Fried
upon Tyne, UK). These antibodies are speci®c for the human
progesterone receptor of both the A and the B form. Secondary
antibodies used were goat anti-mouse IgG labelled with green
¯uorescent Alexa 488 dye (Molecular Probes, Eugene, OR).
Negative controls of the thrombin and the progesterone receptors for
these endothelial cells were carried out with secondary antibodies
only. The cells were observed and photographed using a ¯uorescence
microscope (Axioskop, Zeiss, GoÈttingen, Germany) with the
appropriate ®lters.
Coiled arteries response to endothelin
In order to verify that the small uterine vessels (the coiled arteries
speci®cally) contract in response to ET exposure, we performed
in vitro studies of cannulated blood vessels. We used a video
dimension analyser, a vessel chamber and perfusion equipment from
Living Systems Instrumentation (Burlington, VT).
Uterine coiled arteries, 100±120 mm in inner diameter, were
obtained from the myometrial±endometrial border of the uterus in
connection with hysterectomy for benign disease in pre-menopausal
women. The local medical ethics committee approved the study. A
segment of the artery (5±10 mm long) was carefully dissected from
surrounding muscle, endometrium and connective tissues and
transferred to the experimental chamber ®lled with oxygenated
Krebs buffer (NaCl 120 mmol/l, KCl 5.9 mmol/l, MgSO4
1.2 mmol/l, NaHCO3 15.4 mmol/l, KH2PO4 1.2 mmol/l, CaCl2
2.5 mmol/l and glucose 0.2%) at 37°C. The proximal end of the artery
was fed onto a glass microcannula (tip diameter 100±120 mm) and tied
with a single strand (20 mm) of braided nylon thread. After the artery
was ¯ushed with Krebs buffer to remove intraluminal blood, the distal
end of the vessel was similarly cannulated and tied. With the use of a
pressure servo and peristaltic pump instrument, the intraluminal
pressure was kept constant at 70 mmHg (Halpern et al., 1984).
The vessel in the experimental chamber was perfused continuously
with Krebs buffer, passing through an external water bath at 37°C. A
conductive glass in the base of the vessel chamber, a thermocouple
probe and a heat controller maintained the temperature at 37°C. A
gassing stone was inserted into the well for superfusate solution pH
regulation.
The arteriograph containing the pressurized vessel was placed on
the stage of an inverted microscope (Nikon TSM-F, Kanagawa, Japan)
with a monochrome video camera attached to a viewing tube. Arterial
dimensions were measured using a video system that provides
automatic continuous read-out measurements of luminal diameter
(Halpern et al., 1984). The output of the video dimension analyser was
sent to an IBM-compatible computer by means of a serial data
acquisition system (DATAQ Instruments, OH, USA) for visualization
of dynamic responses of the diameter.
Data analysis and statistics
All data analyses were run on a Dell latitude laptop computer and the
calculations were performed using GraphPad Prism version 3.03 for
Windows (GraphPad Software, San Diego CA). Statistical analysis of
ET changes was carried out using the Mann±Whitney test and
considered signi®cant at the level of P < 0.05.
Results
Endothelin levels
Basal ET levels were 0.6 to 6.5 pg/ml culture medium from
both cell types before treatments were commenced. During the
4 days of treatment, ET production increased daily, but at
different rates in the three treatments studied and in the two cell
1274
Figure 1. Total ET content in culture medium during the 4 days of
culture. The y-axis shows the ET content (pg/ml) and the x-axis
shows the days of culture. Values given are mean and SEM.
(a) Control without added hormones. (b) Treatment with
estrogen + progesterone for 4 days. (c) Treatment with estrogen for
4 days + progesterone for the ®rst 2 days (progesterone withdrawal).
types. In the control group without treatment, there was an
approximately linear increase in ET content during the 4 days
of culture (Figure 1a), whereas with the two types of hormone
treatment, ET changed differentially (Figure 1b and c). In order
to quantitate and compare changes in ET production in the
cultures, we de®ned the change in ET production over time
(DET) as the difference in ET production between one day and
the previous day, i.e. the ET production at the observed time
point as compared with the ET production 24 h earlier. Using
the DET index, it is clearly seen that there was no difference in
ET production in the control group without added hormones
(Figure 2a) or in the group where progesterone was added for
all 4 days (Figure 2b). However, in the group subjected to
progesterone withdrawal, we found an increased ET production
24 h after progesterone was withdrawn in UtMVECs as
compared with HUVECs, as clearly seen by a signi®cantly
increased DET (Figure 2c). In HUVECs, the opposite was seen,
Progesterone withdrawal induces endothelin release
Thrombin receptors
The presence and the relative occurrence of thrombin receptors
in the cultured endothelial cells were studied by immunohistochemistry and by immunoblotting. We found evidence for
the presence of thrombin receptors in both cell types by
immuno¯uorescence staining (Figure 4). In UtMVECs, there
was a strong perinuclear and granular staining pattern, with
occasional staining also more peripherally in the cytoplasm
(Figure 4a). In HUVECs, staining was weaker, but the
localization was similar to that seen in UtMVECs (Figure 4b).
The negative control gave no staining (not shown).
A semi-quantitative analysis of thrombin receptor content by
immunoblotting extracts of the two different types of cells
showed an increased amount of thrombin receptor in HUVECs
and a decreased amount in UtMVECs after 4 days of exposure
to progesterone. In the cells subjected to progesterone withdrawal, this difference was reduced (Figure 5). After 4 days of
progesterone, the relative density of the thrombin receptor band
in HUVECs was 1.4 (60.7) and in UtMVECs 0.5 (60.2).
Thrombin
We could not detect any production of thrombin in the cell
lines with our treatment regimens. All the thrombin content in
the culture medium originated from the FBS (not shown).
Progesterone receptors
We found, by immunohistochemistry, the presence of progesterone receptors in the nuclei of both UtMVECs (Figure 6a)
and HUVECs (Figure 6b). Cells exhibited a strong nuclear
staining as compared with controls incubated with only
secondary antibody (not shown).
Figure 2. Changes in ET content/24 h (DET) in the culture medium
during the 4 days of culture. The y-axis shows the ET content
(pg/ml) and the x-axis shows the days of culture Values given are
mean and SEM. *Signi®cantly (P < 0.05) greater increase in the last
24 h. (a) Control without added hormones. (b) Treatment with
estrogen + progesterone for 4 days. (c) Treatment with estrogen for
4 days + progesterone for the ®rst 2 days (progesterone withdrawal).
a decreased ET production (Figure 2c). Thus, the only culture
condition where we found an increased rate of production of
ET during the last 24 h of culture was in UtMVECs, where
ET production was increased in the period 24±48 h after
progesterone withdrawal.
Cell origin control
We checked that we had endothelial cells in culture by
immunohistochemistry using endothelial cell-speci®c markers
(Figure 3). Cells from the ®rst set of controls for each cell line
were cultured on a Falcon culture slide and showed positive
staining for the endothelial cell marker CD-31 in UtMVECs
(Figure 3a) and HUVECs (Figure 3b). As a negative control,
we stained for smooth muscle actin, which was absent in both
cell types (shown only for HUVECs, Figure 3c) and with
secondary antibody only (not shown). Cells were also documented by photography through light microscopy (not shown).
Endometrial vessel responses to ET
The analysis of vessel contractility showed that, like other
arterial vessels, the coiled artery contracts as a result of
exposure to ET. After addition of ET (10±7 mol/l), we obtained
a reversible and reproducible contraction, from 300 to 140 mm
external diameter. This contraction was blocked by the speci®c
ET receptor blocker BQ-123 (10±7 mol/l) (Figure 7).
Discussion
Early theories on the physiology of menstruation originate
from the work of Markee (1940) who postulated that the tissue
destruction initiating the shedding of the endometrium resulted
from anoxia. This was concluded from the ®nding that vessels
in autologuos endometrium transplanted to the anterior
chamber of the eye of rhesus monkeys reacted to progesterone
withdrawal with vasoconstriction. Current theories suggest an
in¯ammatory response with cytokine upregulation with effects
on numerous systems including the matrix metalloproteinases
(MMPs) (Henriet et al., 2002). This in turn initiates menstruation by tissue destruction and breakdown of the extracellular
matrix, preceding as well as following vasoconstriction
(Salamonsen et al., 1999; Critchley et al., 2001b; Kelly et al.,
2002).
ET is one of the most potent endogenous vasoconstrictors
known. Previous work has shown that ET may be involved, but
1275
M.Edlund, E.Andersson and G.Fried
Figure 3. Endothelial cell origin as demonstrated by immuno¯uorescence staining of the cultured cells. (a) UtMVECs stained positive with
anti-human CD-31. Note the strong staining peripherally in the cytoplasm in areas where there is cell±cell contact. Cell nuclei show very
weak staining. (b) HUVECs stained positive with anti-human CD-31. Note that there is a similar pattern to that seen in UtMVECs; however,
staining is considerably more intense. (c) HUVECs stained negative with anti-human smooth muscle actin. Note the absence of staining in
the cytoplasm. Apparent staining in the nuclei is non-speci®c. (Magni®cation 3400.)
compelling evidence for the role of ET in regulation of
menstruation is lacking. It has been shown previously that ET
has a crucial role in the regulation of vessel contractility
(Yanagisawa et al., 1988) in several organ systems including
the urogenital tract. ET contracts vascular and myometrial
smooth muscle in the uterus (Fried and Samuelson, 1991) and
has been shown to be more potent in contracting small
branches of the uterine artery than in contracting the main stem
(EkstroÈm et al., 1991). Furthermore, previous studies have
demonstrated cyclical changes in the endometrial content of
ET (Economos et al., 1992; Marsh et al., 1994) and ET
receptors (O'Reilly et al., 1992). Marsh et al. (1997) also
showed a correlation between menorrhagia and reduced
ET levels as well as lack of cyclic variation in the glandular
1276
epithelium. Thus several authors postulate a central role
for ET in the events preceding and regulating menstrual
bleeding. In the present experiment, we show an increase in
ET secretion from UtMVECs in response to progesterone
withdrawal.
The ET that would reach the coiled arteries and act as a
vasoconstrictor is likely to be released from vessels preceding
the coiled arteries in the blood ¯ow, from radial artery
endothelial cells. It is known that endothelial cells have
different properties in different types of vessels and in different
organ systems. Earlier studies have shown changes in the
expression of endothelial function markers after treatment with
the progestins medroxyprogesterone acetate (MPA) and
norethisterone (NET) (Mueck et al., 2002). To our knowledge,
Progesterone withdrawal induces endothelin release
Figure 4. Immuno¯uorescence images showing (a) UtMVECs and (b) HUVECs stained with mouse monoclonal IgG antibody against the
thrombin receptor. Note the granular cytoplasmic staining in the immediate perinuclear region, often asymmetrically localized predominantly
adjacent to one nuclear hemisphere. (Magni®cation 3400.)
Figure 5. Thrombin receptor content in the two cell types quantitated by relative optical density after immunoblot. Values given are mean
and range (n = 4, SD). Control: no hormones added, Prog 4 days: addition of estrogen for 4 days and progesterone for 4 days. Prog 2 days:
addition of estrogen for 4 days and progesterone for the ®rst 2 days.
no investigations exist on the difference in ET production from
uterine microvasculature when exposed to progesterone and
subsequent withdrawal.
The present work supports our theory that there is an
increased level of ET in the smallest uterine arteries as a
response to progesterone withdrawal. The reason for the
relative increase of ET in our experimental set-up may be due
to a number of factors. It may be due to increased synthesis of
the precursor peptide pre-proendothelin, increased processing
of the precursor peptide, decreased degradation of ET, changes
in turnover not related to synthesis or degradation, i.e. transport
and/or clearance, or it may even be due to factors related to cell
culture such as cell density, passage number or changes in cell
adhesion properties. It is known that ET production in
HUVECs is strongly dependent on culture conditions (Ranta
et al., 1997). In the two types of cells that we have studied,
UtMVECs and HUVECs, we saw a diametrically opposed
response when exposed to progesterone withdrawal. The
HUVECs are derived from a vein that has the properties of
an artery, and transports oxygenated blood from the placenta to
the fetus. Thus, physiologically, it would be deleterious if a
drop in progesterone levels should induce increased production
of ET and constriction in the umbilical vein. Growth requirements and growth condition for UtMVECs and HUVECs
differ, therefore culture conditions may be a reason for
differences in ET synthesis pattern between the two cell
types. However, our experimental set-up was such that we
believe that factors related to culture conditions are controlled
1277
M.Edlund, E.Andersson and G.Fried
Figure 6. Immuno¯uorescence staining of (a) UtMVECs and (b) HUVECs showing progesterone receptors in the cell nuclei. Note the
granular staining throughout the cell nuclei and absence of staining in the cytoplasm. (Magni®cation 3400.)
Figure 7. Contraction of a coiled artery from 300 to 140 mm
external diameter as a response to ET exposure. The speci®c
endothelin blocker BQ-123 blocks the contraction.
for, and thus would not be the reason for the increase in ET in
UtMVECs after progesterone withdrawal.
Our protocol with 4 days of hormone treatment partly
resembles the female hormonal cycle during the last 4 days of a
normal menstrual cycle. The longer the treatment period
continues, the higher is the risk of confounding factors such as
differing density of cells. Even if it would be bene®cial to
prolong the treatment period, not only to better resemble the
direct effects of hormone in¯uence but also to achieve a
possible estrogen-induced priming of the progesterone receptors, we considered the disadvantages to outweigh the advantages. We also found that previous work presenting similar
studies of hormone effect on cells in culture report results after
24 h (Mueck et al., 2002) to 3 or 5 days (Casey et al., 1991).
A decreased degradation of ET after progesterone withdrawal is one possibility for the observed results. This would
then imply a link between progesterone and enkephalinase, the
main ET-degrading enzyme [neutral endopeptidase (NEP) EC
3.4.24.11] (Vijayaraghavan et al., 1990). The activity of
enkephalinase, localized predominantly to stromal cells in the
1278
human endometrium (Casey et al., 1991; Imai et al., 1992;
Head et al., 1993), changes through the menstrual cycle, with
an increase between early proliferative and mid secretory
phases. The subsequent fall in enkephalinase activity coincides
with the fall in progesterone concentrations after luteolysis
(Casey et al., 1991). During menstruation, enkephalinase
activity is low, and this is a possible explanation for the
increased ET activity seen during menstruation (Casey et al.,
1991; Imai et al., 1992; Head et al., 1993). The mechanism
regulating enkephalinase activity ¯uctuation during the
menstrual cycle is not known.
We believe that another likely explanation for the increase in
ET may be an increase in synthesis or processing. ET is
processed from its precursor peptide pre-proendothelin by the
enzyme ECE-1 (Eto et al., 2001). One factor that both
upregulates ECE-1 and stimulates ET release is thrombin
(Ohlstein and Storer, 1992). Thrombin is a serine protease and
one of the most important enzymes in the blood clotting
cascade. In addition to its action in haemostasis, thrombin has
been found to act as a cofactor in response to sex steroids
(Henrikson et al., 1990, 1994; Henrikson, 1992). Treatment
with progesterone increased the levels of mRNA for the
proteolytically activatable thrombin receptor (PAR-1) in
vascular smooth muscle cells (Herkert et al., 2001). We did
not ®nd evidence for production of thrombin in our cultured
endothelial cells in any of the hormone treatments. We do,
however, ®nd indications of an increased expression of the
thrombin receptor in HUVECs and, interestingly, a decrease in
UtMVECs. This effect is not seen after progesterone withdrawal, which gives the effect of a relative thrombin receptor
upregulation in UtMVECs after progesterone withdrawal. This
would be in agreement with the observation that progesterone
withdrawal decreases ET in HUVECs and increases ET in
UtMVECs, assuming that thrombin receptors mediate ET
Progesterone withdrawal induces endothelin release
release and increased activity of ECE-1. One part of the
difference in ET release after progesterone withdrawal could
then be explained by the increased expression of thrombin
receptor.
In order to respond directly to progesterone withdrawal, it is
likely that the UtMVECs express progesterone receptors. There
are reports of the absence of progesterone receptors in the
vascular endothelium of the actual coiled arteries (PerrotApplanat et al., 1994; Kohnen et al., 2000; Critchley et al.,
2001a). However, cells in the vascular tree immediately
preceding the spiral arteries must also be considered in this
regard. We found in our cultured UtMVECs consisting of
endothelial cells from small uterine vessels including both
spiral arteries and vessels preceding the spiral arteries, a
positive staining for progesterone receptors (Figure 6), explaining the observed reactivity to changes in progesterone
exposure.
The effects of estrogen and progesterone have also been
reported in other organ systems. For example, Mueck et al.
(2002) found a reduced ET production in human female
coronary endothelial cell cultures when treated with estrogen
alone or with estrogen and a progestin (MPA or NET). It may
be hypothesized that a similar inhibitory effect exists in the
endothelial cells in our study. Thus the special property of the
endothelium from the small uterine arteries is to `rebound' and
increase the production of ET when the progesterone inhibition
is removed. Interestingly, it has also been shown in vivo that
high circulating levels of estrogen in women undergoing
hormone stimulation for IVF were accompanied by a decrease
in circulating levels of ET-1 (Cacciatore et al., 1997).
The events that take place in the endometrium following
progesterone withdrawal, leading to menstruation and endometrial repair, are complex. Several cell types and several
subpopulations of each cell type are involved. In this report, we
present the results of a study on one of these cell types that, at
least in vivo, interacts with adjacent and more distal cells
through processes and paracrine actions. We have, however, in
this isolated cell type and as a result of progesterone
withdrawal, been able to detect an increased production of
the potent vasoconstrictor ET, by many postulated to be pivotal
in the constriction of the spiral arteries during menses. These
®ndings can add another piece to the jigsaw of understanding
menstrual physiology.
References
Axelsson G, Korsan-Bengtsen K and Waldenstrom J (1976) Prothrombin
determination by means of a chromogenic peptide substrate. Thromb
Haemostasis 36,517±524.
Cacciatore B, Simberg N, Tiitinen A and Ylikorkala O (1997) Evidence of
interplay between plasma endothelin-1 and 17 beta-estradiol in regulation of
uterine blood ¯ow and endometrial growth in infertile women. Fertil Steril
67,883±888.
Casey ML, Smith JW, Nagai K, Hersh LB and MacDonald PC (1991)
Progesterone-regulated cyclic modulation of membrane metalloendopeptidase (enkephalinase) in human endometrium. J Biol Chem
266,23041±23047.
Critchley HO, Brenner RM, Henderson TA, Williams K, Nayak NR, Slayden
OD, Millar MR and Saunders PT (2001a) Estrogen receptor beta, but not
estrogen receptor alpha, is present in the vascular endothelium of the human
and nonhuman primate endometrium. J Clin Endocrinol Metab 86,1370±
1378.
Critchley HO, Kelly RW, Brenner RM and Baird DT (2001b) The
endocrinology of menstruationÐa role for the immune system. Clin
Endocrinol 55,701±710.
Economos K, MacDonald PC and Casey ML (1992) Endothelin-1 gene
expression and protein biosynthesis in human endometrium: potential
modulator of endometrial blood ¯ow. J Clin Endocrinol Metab 74,14±19.
Ê kerlund M (1991) Differences in vasomotor responses
EkstroÈm P, Alm P and A
between main stem and smaller branches of the human uterine artery. Acta
Obstet Gynecol Scand 70,429±433.
Emeis JJ, van den Eijnden-Schrauwen Y, van den Hoogen CM, de Priester W,
Westmuckett A and Lupu F (1997) An endothelial storage granule for
tissue-type plasminogen activator. J Cell Biol 139,245±256.
Eto M, Barandier C, Rathgeb L, Kozai T, Joch H, Yang Z and Luscher TF
(2001) Thrombin suppresses endothelial nitric oxide synthase and
upregulates endothelin-converting enzyme-1 expression by distinct
pathways: role of Rho/ROCK and mitogen-activated protein kinase. Circ
Res 89,583±590.
Fried G and Samuelson U (1991) Endothelin and neuropeptide Y are
vasoconstrictors in human uterine blood vessels. Am J Obstet Gynecol
164,1330±1336.
Gude NM, King RG and Brennecke SP (1991) Endothelin: release by and
potent constrictor effect on the fetal vessels of human perfused placental
lobules. Reprod Fertil Dev 3,495±500.
Hague S, Oehler MK, MacKenzie IZ, Bicknell R and Rees MC (2002)
Protease activated receptor-1 is down regulated by levonorgestrel in
endometrial stromal cells. Angiogenesis 5,93±98.
Halpern W, Osol G and Coy GS (1984) Mechanical behavior of pressurized
in vitro prearteriolar vessels determined with a video system. Ann Biomed
Eng 12,463±479.
Head JR, MacDonald PC and Casey ML (1993) Cellular localization of
membrane metalloendopeptidase (enkephalinase) in human endometrium
during the ovarian cycle. J Clin Endocrinol Metab 76,769±776.
Henriet P, Cornet PB, Lemoine P, Galant C, Singer CF, Courtoy PJ, Eeckhout
Y and Marbaix E (2002) Circulating ovarian steroids and endometrial
matrix metalloproteinases (MMPs). Ann NY Acad Sci 955,119±138;
discussion 157±158, 396±406.
Henrikson KP (1992) Thrombin as a hormonally regulated growth factor in
estrogen-responsive tissue. Semin Thromb Hemostasis 18,53±59.
Henrikson KP, Jazin EE, Greenwood JA and Dickerman HW (1990)
Prothrombin levels are increased in the estrogen-treated immature rat
uterus. Endocrinology 126,167±75.
Henrikson KP, Hall ES and Lin Y (1994) Cellular localization of tissue factor
and prothrombin in the estrogen-treated immature rat uterus. Biol Reprod
50,1145±1150.
Herkert O, Kuhl H, Sandow J, Busse R and Schini-Kerth VB (2001) Sex
steroids used in hormonal treatment increase vascular procoagulant activity
by inducing thrombin receptor (PAR-1) expression: role of the
glucocorticoid receptor. Circulation 104,2826±2831.
Imai K, Maeda M, Fujiwara H, Okamoto N, Kariya M, Emi N, Takakura K,
Kanzaki H and Mori T (1992) Human endometrial stromal cells and
decidual cells express cluster of differentiation (CD) 13 antigen/
aminopeptidase N and CD10 antigen/neutral endopeptidase. Biol Reprod
46,328±334.
Kelly RW, King AE and Critchley HO (2002) In¯ammatory mediators and
endometrial functionÐfocus on the perivascular cell. J Reprod Immunol
57,81±93.
Kohnen G, Campbell S, Jeffers MD and Cameron IT (2000) Spatially
regulated differentiation of endometrial vascular smooth muscle cells. Hum
Reprod 15,284±292.
Lang I, Pabst MA, Hiden U, Blaschitz A, Dohr G, Hahn T and Desoye G
(2003) Heterogeneity of microvascular endothelial cells isolated from
human term placenta and macrovascular umbilical vein endothelial cells.
Eur J Cell Biol 82,163±173.
Lockwood CJ and Schatz F (1996) A biological model for the regulation of
peri-implantational hemostasis and menstruation. J Soc Gynecol Invest
3,159±165.
Luscher TF, Seo BG and Buhler FR (1993) Potential role of endothelin in
hypertension. Controversy on endothelin in hypertension. Hypertension
21,752±757.
Markee JE (1940) Menstruation in intraocular endometrial transplants in the
rhesus monkey. Contrib Embryol Carnegie Inst Washington 28,219±308.
Marsh MM, Hampton AL, Riley SC, Findlay JK and Salamonsen LA (1994)
1279
M.Edlund, E.Andersson and G.Fried
Production and characterization of endothelin released by human
endometrial epithelial cells in culture. J Clin Endocrinol Metab 79,1625±
1631.
Marsh MM, Malakooti N, Taylor NH, Findlay JK and Salamonsen LA (1997)
Endothelin and neutral endopeptidase in the endometrium of women with
menorrhagia. Hum Reprod 12,2036±2040.
Mueck AO, Seeger H and Wallwiener D (2002) Medroxyprogesterone acetate
versus norethisterone: effect on estradiol-induced changes of markers for
endothelial function and atherosclerotic plaque characteristics in human
female coronary endothelial cell cultures. Menopause 9,273±281.
Ohlstein EH and Storer BL (1992) Oxyhemoglobin stimulation of endothelin
production in cultured endothelial cells. J Neurosurg 77,274±278.
O'Reilly G, Charnock-Jones DS, Davenport AP, Cameron IT and Smith SK
(1992) Presence of messenger ribonucleic acid for endothelin-1, endothelin2, and endothelin-3 in human endometrium and a change in the ratio of ETA
and ETB receptor subtype across the menstrual cycle. J Clin Endocrinol
Metab 75,1545±1549.
Perrot-Applanat M, Deng M, Fernandez H, Lelaidier C, Meduri G and
Bouchard P (1994) Immunohistochemical localization of estradiol and
progesterone receptors in human uterus throughout pregnancy: expression
in endometrial blood vessels. J Clin Endocrinol Metab 78,216±224.
Ranta V, Orpana A, Mikkola T, Ylikorkala O and Viinikka L (1997) Culturing
1280
of human vascular endothelial cells strongly affects their endothelin-1 and
prostacyclin production. Mol Cell Biochem 177,251±255.
Russell FD, Skepper JN and Davenport AP (1998) Evidence using
immunoelectron microscopy for regulated and constitutive pathways in
the transport and release of endothelin. J Cardiovasc Pharmacol 31,424±
430.
Salamonsen LA, Kovacs GT and Findlay JK (1999) Current concepts of the
mechanisms of menstruation. Baillieres Best Pract Res Clin Obstet Gynecol
13,161±179.
Vijayaraghavan J, Scicli AG, Carretero OA, Slaughter C, Moomaw C and
Hersh LB (1990) The hydrolysis of endothelins by neutral endopeptidase
24.11 (enkephalinase). J Biol Chem 265,14150±14155.
Wagner DD, Olmsted JB and Marder VJ (1982) Immunolocalization of von
Willebrand protein in Weibel±Palade bodies of human endothelial cells. J
Cell Biol 95,355±360.
Yanagisawa M and Masaki T (1989) Molecular biology and biochemistry of
the endothelins. Trends Pharmacol Sci 10,374±378.
Yanagisawa M, Kurihara H, Kimura S, Tomobe Y, Kobayashi M, Mitsui Y,
Yazaki Y, Goto K and Masaki T (1988) A novel potent vasoconstrictor
peptide produced by vascular endothelial cells. Nature 332,411±415.
Submitted on January 5, 2004; accepted on March 18, 2004