This Article Abstract Full Text (PDF) All Versions of this Article: 145/1/113    most recent Author Manuscript (PDF) Purchase Article View Shopping Cart Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Copyright Permission Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Chen, D.-b. Articles by Magness, R. R. Search for Related Content PubMed PubMed Citation Articles by Chen, D.-b. Articles by Magness, R. R. Pubmed/NCBI databases Compound via MeSH Substance via MeSH Hazardous Substances DB CALCIUM COMPOUNDS CALCIUM, ELEMENTAL ESTRADIOL NITRIC OXIDE

Membrane Estrogen Receptor-Dependent Extracellular Signal-Regulated Kinase Pathway Mediates Acute Activation of Endothelial Nitric Oxide Synthase by Estrogen in Uterine Artery Endothelial Cells

Department of Reproductive Medicine (D.-B.C.), University of California San Diego, La Jolla, California 92093; Perinatal Research Laboratories, Departments of Obstetrics and Gynecology (I.M.B., J.Z., R.R.M.) and Animal Sciences (R.R.M.), University of Wisconsin-Madison, 7E. Meriter Hospital, Madison, Wisconsin 53715

Address all correspondence and requests for reprints to: Dongbao Chen, Ph.D., Division of Maternal-Fetal Medicine (MC0802), Department of Reproductive Medicine, University of California San Diego, 9500 Gilman Drive, La Jolla, California 92093-0802. E-mail: dochen{at}ucsd.edu.

	Abstract

Top Abstract Introduction Materials and Methods Results Discussion References

 
Rapid uterine vasodilatation after estrogen administration is believed to be mediated by endothelial production of nitric oxide (NO) via endothelial NO synthase (eNOS). However, the mechanism(s) by which estrogen activates eNOS in uterine artery endothelial cells (UAEC) is unknown. In this study, we observed that estradiol-17ß (E2) and E2-BSA rapidly (<2 min) increased total NO_x production in UAEC in vitro. This was associated with rapid eNOS phosphorylation and activation but was unaltered by pretreatment with actinomycin-D. estrogen receptor- protein was detectable in isolated plasma membrane proteins by immunoblotting, and E2-BSA-fluorescein isothiocyanate binding was evident on the plasma membrane of UAEC. E2 did not mobilize intracellular Ca²⁺, but E2 and ionomycin in combination induced greater eNOS phosphorylation than either E2 or ionomycin alone. E2 did not stimulate rapid Akt phosphorylation. E2 stimulated rapid ERK2/1 activation in a time- and dose-dependent manner, with maximal responses observed at 5–10 min with E2 (10 nM to 1 µM) treatment. Acute activation of eNOS and NO_x production by E2 could be inhibited by PD98059 but not by LY294002. When E2-BSA was applied, similar responses in NO_x production, eNOS, and ERK2/1 activation to those of E2 were achieved. In addition, E2 and E2-BSA-induced ERK2/1 activation and ICI 182,780 could inhibit NOx production by E2. Thus, acute activation of eNOS to produce NO in UAEC by estrogen is at least partially through an ERK pathway, possibly via estrogen receptor localized on the plasma membrane. This pathway may provide a novel mechanism for NO-mediated rapid uterine vasodilatation by estrogen.

	Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

 
ESTROGEN IS A POTENT vasoactive hormone that can initiate very rapid vasodilatation in various vascular beds and tissue perfusion throughout the body (1, 2). In ovariectomized animals, after estrogen administration uterine vasodilatation as measured by a rise in uterine blood flow takes place within 30–45 min and increases up to 10-fold within 90–120 min (3, 4). This rapid vasodilatory effect of estrogen in the uterus has shown to be mediated largely by artery endothelial production of the potent vasodilator nitric oxide (NO), which is primarily derived from the conversion of L-arginine to L-citrulline by the Ca²⁺-dependnent endothelial NO synthase (eNOS). This concept is drawn based on the observations that eNOS is the primary NOS isoform expressed in uterine artery endothelium (5, 6, 7) and that infusion of specific NOS inhibitors into the uterine circulation suppresses estrogen-induced rise in uterine (8, 9) and systemic (10) blood flows. However, the molecular mechanism by which estrogen stimulates eNOS to produce NO in uterine artery endothelium is unknown.

It has been well established that steroid hormones elicit their diverse biological functions by binding to their nuclear receptors interacting with DNA to regulate gene transcription (11). This classical model of steroid actions may be important for estrogen-induced uterine vasodilatation because chronic estrogen treatment increases eNOS protein expression in uterine artery endothelium in vivo (5, 6, 7). Increased eNOS protein expression is expected to stimulate NO production. However, this cannot fully explain the rapid timing of estrogen-induced rises in uterine and systemic blood flows. Obviously, mechanism(s) other than gene expression may play an even more important role in stimulating eNOS to produce NO during estrogen-induced vasodilatation because gene expression requires hours to days to occur. Because previous studies have shown that blockade of mRNA synthesis by actinomycin-D had no effects on estrogen-induced rise in uterine blood flow (12, 13), we therefore propose that increased eNOS activity but not de novo synthesis of eNOS protein is required principally for at least the initiation of estrogen-induced uterine vasodilatation. On estrogen stimulation an array of rapid cellular responses, including mobilization of intracellular Ca²⁺ (14) and activation of MAPK (14, 15) and protein kinase B/Akt (16, 17), etc., also can occur within seconds to minutes. These rapid estrogen actions are presumptively mediated by estrogen receptors localized on the plasma membrane and so named as nongenomic actions of estrogen (18, 19). It is therefore possible that estrogen may use some of these rapid signaling pathways alternatively to activate eNOS resulting in NO mediated uterine vasodilatation.

eNOS protein possesses multiple putative phosphorylation sites, which can be phosphorylated by various protein kinases including Akt (20, 21) and ERK2/1 (22). It has been shown that eNOS is a direct substrate for Akt (20, 21). On estrogen stimulation, Akt is rapidly phosphorylated that in turn activates eNOS in human umbilical vein (16) and rat lung vascular (23) endothelial cells. Moreover, estrogen rapidly activates ERK2/1 (24). However, the role of ERK2/1 in regulating eNOS activity is less well understood. It has been recently shown that ERK activation is involved in acute activation of eNOS by estrogen in ovine fetal pulmonary artery (15) and human umbilical vein (25) endothelial cells. In contrast, acute treatment with bradykinin results in ERK-sensitive phosphorylation of eNOS, which leads to decreased eNOS activity in bovine aortic endothelial cells (22). Thus, it seems that the role of ERK2/1 in the regulation of eNOS activity is specific to the endothelial cell types studied. In uterine artery endothelial cells (UAECs), phosphorylation of ERK2/1 induced by a broad range of physiological stimuli strongly and positively correlates to elevated NO production (26), which suggest that in UAECs activated ERK2/1 may be stimulatory for eNOS activity. The present study was undertaken to determine the acute effects of estrogen on eNOS-NO, Ca²⁺, ERK2/1, and Akt pathways and to delineate the role of Ca²⁺, ERK2/1, and Akt in estrogen activation of eNOS in UAECs. The involvement of estrogen receptor (ER) in the acute activation of eNOS by estrogen was also examined. We present data herein to show that estradiol-17ß (E2) and the membrane impermeable E2-BSA rapidly induce ER-dependent activation of eNOS and NO production, phosphorylation, and activation of ERK2/1 with no apparent Akt phosphorylation or measurable intracellular Ca²⁺ mobilization. Activation of ERK pathway may play an important role in acute activation of eNOS by estrogen.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

 
Materials
E2 and E2-BSA [ß-estradiol-6-(O-carboxy-methyl)oxime/BSA; 35 mol E2/mol BSA], BSA-fluorescein isothiocyanate (FITC), E2-BSA-FITC, and actinomycin-D were purchased from Sigma Chemical Co. (St. Louis, MO). ICI-182, 780 was obtained from Tocris (Ballwin, MO). Rabbit anti-Akt polyclonal antibody (pAb), phospho-specific Akt pAb, phospho-specific eNOS pAb, and phospho-specific MAPK pAb were obtained from New England Biolabs (Beverly, MA). Protein-A conjugated agarose beads, recombinant [His]₆-tagged-mitogen-activated protein kinase kinase 1 (MEK1) and [His]₆-tagged-ERK2, and glutathione-S-transferase (GST)-fused Elk-1 were obtained from Upstate Biotechnology, Inc. (Lake Placid, NY). Rabbit polyclonal anti-Raf-1 antibody (pAb) conjugated agarose beads, rabbit anti-ERK2 pAb conjugated agarose beads, and anti-Raf-1 pAb were purchased from Santa Cruz Biotechnology, Inc. (Santa Cruz, CA). Anti-Pander monoclonal antibody (mAb) and anti-eNOS mAb (N30020) were from BD Transduction Laboratories (Lexington, KY). Anti-ER pAb (39021) was from Geneka Biotechnology Inc. (Montreal, Canada), and recombinant human ER protein was from Stressgene (Victoria, British Columbia, Canada). [-³²P]ATP (37 MBq, 10 mCi/ml), [³²P]orthophosphate (370 MBq, 150 mCi/ml), and [³H]-L-arginine (9.25 MBq, 1 mCi/ml) were from DuPont (Boston, MA). Tissue culture plasticware was from Corning (Corning, NY). Fetal calf serum (FCS), medium-199 (M-199), and phosphate-free DMEM were from Life Technologies Inc. (Grand Island, NY). Electrophoresis reagents, precasted SDS-PAGE gels, and Dowex AG 50WX-8 (Na⁺ form) resin were from Bio-Rad Laboratories (Hercules, CA). Enhanced chemiluminescence kits were from Amersham (Arlington Heights, IL). Immobilon-p [polyvinyl difluoride (PVDF)] membrane was from Millipore (Bedford, MA). PD98059 and LY294002 were purchased from Calbiochem (La Jolla, CA).

Cell culture, experimental conditions, and preparation of total cell extracts
UAECs were isolated from pregnant (d 120–130) ewes under nonsurvival surgery by collagenase digestion, cultured in growth media (D-Val MEM with 20% FCS, 100 U/ml penicillin, and 100 µg/ml streptomycin), and propagated as previously described (26). Frozen UAEC aliquots (passage 3) were thawed and plated in T-75 flasks to grow to about 70% confluence in growth media, and subcultured in Phenol-red-free M-199 containing 20% FCS, 0.1% BSA, 25 mM HEPES, 1% antibiotics for experimental use at passages 4–5. Before experiment, cells at about 80% confluence were serum starved in treatment media (Phenol red-free M-199 containing 0.1% BSA, 25 mM HEPES) for 16–20 h. The media were then replaced with treatment media, and the cultures were allowed to equilibrate for 1 h. Agonists and/or antagonists were added for the time period as described in the figure legends. Cell stimulation was terminated by aspiration of the media. After rinsing twice with ice-cold PBS, the cells were lysed with a nondenaturing lysis buffer A (27) on ice with continuous shaking for 30 min. The total cell extracts were collected using a disposable cell scraper, vortexed, and clarified by centrifugation (13,000 x g, 5 min). The protein content of the samples was measured by a Bio-Rad procedure using BSA as the standard. Aliquots of the extracts were used for immunoprecipitation or frozen at -80 C until Western blot analysis could be performed.

eNOS-specific activity assay
eNOS was immunoprecipitated from total cell extracts (>200 µg/sample) as described below, and its activity was measured by the ability of converting [³H]-L-arginine to [³H]citrulline as described (22) and modified as below. Briefly, eNOS immunoprecipitates were washed three times with lysis buffer and twice with NOS assay buffer [50 mM Tris-HCl (pH 7.5), 0.1 mM EDTA, 0.1 mM EGTA, 2 µM leupeptin, 1 µM pepstatin, 1 µM aprotinin, 1 mM phenylmethylsulfonyl fluoride, 10 mM NaF, 1 mM Na₃VO₄], and resuspended in 50 µl of NOS assay buffer. After mixed with 50 µl of assay cocktail containing 1 mM nicotinamide adenine dinucleotide phosphate reduced, 3 µM tetrahydrobiopterin, 100 nM calmodulin, 2.5 mM CaCl₂, and 10 µM L-arginine, and 0.2 µCi [³H]-L-arginine, the reaction mix was incubated at 37 C for 30 min (mixed every other 5 min). The reaction was stopped by the addition of 1 ml of cold stop buffer [20 mM HEPES (pH 5.5), 2 mM of each EDTA and EGTA], chilled on ice for 5 min, and then centrifuged (12,000 x g, 3 min). The supernatant was applied on 1-ml columns of Dowex AG 50WX-8 (Na⁺ form; preequilibrated in stop buffer). [³H]citrulline formed was eluted with 1 ml water and quantified by liquid scintillation counting. The beads were resuspended in Laemmli buffer and analyzed by immunoblotting with an anti-eNOS mAb to monitor equal eNOS loading in the assay. Control reactions were run in parallel with buffer alone or without eNOS immunoprecipitates.

Immunoprecipitation, SDS-PAGE, and immunoblotting
Equal amounts of total cell lysates (>200 µg protein per sample) were incubated with 2 µg of a specific antibody of interest, and the volume was bought up to 1.0 ml by the addition of lysis buffer. This mixture was incubated overnight at 4 C with end-over-end rotation. Protein-A agarose beads (50/50 slurry beads, 50 µl, Upstate Biotechnology Inc.) were added and incubated for 1–2 h at 4 C. The beads (immunoprecipitates) were then captured by centrifugation (13,000 x g, 4 C, 5 min), washed, and resuspended in buffers as described for the measurement of each specific enzyme. The protein samples or immunoprecipitates were heat denatured (95 C, 10 min) in Laemmli buffer, separated on precasted SDS-PAGE, and electrically (100 V, 2 h) transferred to PVDF membranes. Immunoblotting was conducted as described previously (28, 29). The dilution factor for each primary antibody was indicated in the figure legends, whereas the secondary antibodies were diluted at 1:2000 for antimouse or 1:4000 for antirabbit peroxidase conjugated IgGs, respectively. Bound antibodies were visualized using enhanced chemiluminescence reagents, followed by exposure to x-ray films. The immunoreactive signals were analyzed by densitometry using the Gel-Pro Analyzer software (Media Cybernetics, Silver Spring, MD).

Preparation of plasma membrane and nuclear extracts
UAECs in two 100-mm dishes (15 x 10⁶ cells) were collected in cold PBS using a cell scraper and transferred in a 1.5-ml tube. The cells were pelleted and resuspended in 300 µl of a homogenization buffer [20 mM Tris-HCl (pH 7.0), 0.25 M sucrose, 1 mM EDTA, 1 mM Na₃VO₄, 1 µM okadaic acid, 1 mM phenylmethylsulfonyl fluoride, and 10 µg/ml each of aprotinin, leupeptin, and pepstatin] and homogenized by passing through a 25-guage needle 15 times. The resulting homogenate was centrifuged for 10 min at 3,000 x g. The supernatant was collected and kept on ice. The pellet was resuspended in 200 µl of homogenization buffer and centrifuged again for 10 min at 3,000 x g. The two supernatants were pooled and centrifuged for 20 min at 12,000 x g. The supernatants were collected and then pelleted for 1 h at 100,000 x g. The pellets were dissolved in 200 µl of the nondenaturing buffer A (27) and used as plasma membrane extracts. Nuclear extracts were prepared as described previously (30) and modified as below. UAECs in two 100-mm dishes were rinsed once with ice-cold PBS; once with PBS containing 20 mM NaF, 1 mM Na₃VO₄; and once with hypotonic buffer [20 mM HEPES (pH 7.9), 20 mM NaF, 1 mM Na₃VO₄, 1 mM Na₄P₂O₇, 1 mM EDTA, 1 mM EGTA, 1 mM dithiothreitol, 0.5 mM phenylmethylsulfonyl fluoride, 10 µg/ml of each leupeptin, aprotinin and pepstatin-A]. Then hypotonic buffer (1 ml) containing 0.1% Nonidet P-40 was added. The cells were scrapped into Eppendorf tubes, passed through a 21-gauge needle 10 times, mixed, and centrifuged (15,000 x g) for 20 sec. The pelleted nuclei were resuspended in 200 µl of the nondenaturing buffer A (27) and were used as nuclear extracts.

E2-BSA-FITC binding studies
Cells were plated sparsely on gelatin-coated glass coverslips and allowed to grow for 2 d. The cells were washed with cold PBS and fixed with 4% paraformaldehyde. After incubation with 1% BSA in PBS for 1 h, the cells were incubated with 50 µM of either BSA-FITC (control) or E2-BSA-FITC for 1 h in dark at room temperature. The cells were then washed four times (5 min each) with PBS and examined with a TE-300 fluorescence microscope (Nikon, Tokyo, Japan).

Analysis of eNOS phosphorylation
Metabolic labeling with [³²P]orthophosphate was used to measure in situ phosphorylation of eNOS. Serum-starved UAECs in 60-mm dishes were washed and labeled with [³²P]orthophosphate (0.1 mCi/ml) in fresh phosphate- and phenol red-free DMEM for 4 h. The cells were then treated for 5 min with agonists as described in the figure legends. After rapidly rinsing twice with cold PBS, the cells were solubilized with lysis buffer as described above, and aliquots were taken for scintillation counting to normalize the ³²P-counts between samples. eNOS was then immunoprecipitated from equal amounts of cell extracts (in ³²P counts) with 1 µg of anti-eNOS mAb and captured by protein-A conjugated beads. eNOS immunoprecipitates were eluted from the beads with Laemmli buffer, separated on 7.5% SDS-PAGE, and transferred onto Immobilon-P (PVDF) membranes. The membrane was subjected to autoradiography to detect phosphorylated eNOS, followed by immunoblotting with anti-eNOS mAb for monitoring equal loading. For analysis of eNOS serine phosphorylation, eNOS was immunoprecipitated from total cell extracts (>200 µg of protein) prepared from serum- and steroid-starved cells. The eNOS immunoprecipitates were dissolved in Laemmli buffer, boiled, size fractionated by SDS-PAGE, and transferred onto PVDF membranes. Serine phosphorylation of eNOS was analyzed by immunoblotting with a phospho-specific anti-eNOS pAb reacting only with eNOS phosphorylated on serine¹¹⁷⁷.

Fura-2 Ca²⁺ imaging studies (26)
UAECs were plated to low density (10–20% confluence) on 35-mm dishes with glass coverslip windows (Intracellular Imaging, Inc., Cincinnati, OH) the night before use to allow attachment. The next day immediately before use, cells were loaded with fura-2/AM for 40 min. After rinsing three times in prewarmed (37 C) Krebs buffer (with 2 mM CaCl₂), the cells were incubated in Krebs buffer (2 ml final volume). Fura-2 loading was verified by viewing at 380 nm UV excitation on a Nikon Diaphot inverted microscope (InCyt Im2, Intracellular Imaging, Inc.). A single isolated cell was then set in the field of view and recordings commenced, using alternate excitation at 340 and 380 nm at 50-msec intervals and measuring emitted light using a photomultiplier. From the ratio of emission at 510 nm detected at the two excitation wavelengths and by comparison with a standard curve established for the same settings using buffers of known free [Ca²⁺], the [Ca²⁺]_i was then calculated in real time using the InCyt Im2 software online (Intracellular Imaging, Inc.). E2 and ATP were made in an equal volume (1 ml) of buffer to ensure rapid mixing.

Analysis of Raf-1MEK1/ERK signaling cascade
ERK2/1 phosphorylation was analyzed by immunoblotting with the phospho-specific MAPK pAb (28, 29). Raf-1 activation was measured by a two-step immunocomplex kinase assay by the ability of immunoprecipitated Raf-1 to phosphorylate recombinant MEK1 and phosphorylated MEK1 to subsequently phosphorylate ERK2 in a kinase reaction as described previously (28, 29). ERK2 activation was measured by the ability of immunoprecipitated ERK2 to phosphorylate GST-Elk-1 in a kinase reaction as previously described (28). Kinase reactions were stopped by the addition of Laemmli buffer. The reaction mixtures were boiled, separated on SDS-PAGE, and transferred onto PVDF membrane. Phosphorylated substrates (MEK1, ERK2, and Elk-1) were detected by phosphor imaging analysis, and the protein bands were cut out and quantified by liquid scintillation counting. A portion of Raf-1 or ERK2 immunoprecipitates was analyzed by immunoblotting to monitor equal loading in the kinase reaction.

Analysis of Akt phosphorylation
Equal amounts of total cell extracts (20 µg/lane) were separated on 10% SDS-PAGE and transferred onto PVDF membrane. Akt phosphorylation was determined by immunoblotting with phospho-Akt pAb as described above.

Measurement of total nitrite/nitrate (NO_x) production
Serum-starved UAECs in 24-well plates were washed twice with freshly prepared Krebs buffer (125 mM NaCl, 5 mM KCl, 1 mM MgSO₄, 1 mM KH₂PO₄, 6 mM glucose, 25 mM HEPES, and 2 mM CaCl₂). Prewarmed (37 C) buffer was added into each well (500 µl/well) and the cultures were equilibrated for 1 h. For inhibition experiments, inhibitors were added during the last 30 min of the 1-h equilibration period. Cell stimulation was initiated by the addition of E2 or E2-BSA for the time period and concentrations in triplicate as described in the figure legends. Buffers were sampled for the measurement of total NO_x production and the cells in each well were solubilized in lysis buffer for protein determination. NO_x production in 100 µl sample was measured by electrochemical detection using a NO analyzer (NOA 280, Silvers Instruments, Boulder, CO) based on a reaction of conversion of nitrite/nitrate to NO_x in the media. NO_x production was calculated by a standard curve generated with sodium nitrate as the standard and normalized by the protein content of corresponding wells. For some experiments, total NO_x production was measured by the StressXpress total NO_x detection kit (Stressgene). Quality controls include Kreb’s buffer alone and buffers incubated with the experiments in the same culture plate without cells.

Statistical analysis
Data are presented as means ± SEM and analyzed by one-way ANOVA (SigmaStat, Jandel Scientific, San Rafael, CA). When an F test was significant (P < 0.05), treatment responses were compared with their corresponding controls by Bonferroni’s multiple comparisons.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

 
Estrogen rapidly stimulates eNOS activity and NO_x production
On estrogen stimulation, endothelial cells produce NO rapidly in a variety of endothelial cell types due to the activation of eNOS (31). To evaluate whether acute treatment with estrogen also leads to rapid activation of eNOS in UAECs, serum- and steroid-starved UAECs were challenged with 10 nM E2 for up to 60 min. The specific activity of eNOS is measured by the ability of immunoprecipitated eNOS to convert [³H]-L-arginine to [³H]-L-citrulline. After treatment with 10 nM E2, eNOS activity was increased by 34% within 2 min, maximized (47%) at 5–10 min, and reached plateau thereafter and maintained up to 60 min (Fig. 1A). Similar time-course studies were performed to examine the effects of E2 and E2-BSA on NO_x accumulation in UAECs. As shown in Fig. 1B, treatment with 10 nM E2 also rapidly stimulated NO_x accumulation by UAEC within 2 min, maximized (2.5- to 4-fold) at 5–10 min, and reached plateau thereafter and maintained up to 60 min. Interestingly, treatment with the membrane-impermeable E2 analog E2-BSA (10 nM) induced a similar time-course response in NO_x accumulation to that of E2 by UAECs. In addition, pretreatment with a transcription inhibitor actinomycin-D (20 µM) did not significantly alter E2- and E2-BSA-induced NO_x production (Fig. 1C).

View larger version (22K): [in this window] [in a new window]   FIG. 1. Estrogen rapidly activates eNOS and stimulates NOx production in uterine artery endothelial cells. A, Subconfluent UAECs in 100-mm dishes were serum and steroid starved for 16 h and then treated with or without E2 (10 nM) for up to 60 min. Total cell extracts were prepared and eNOS was immunoprecipitated from equal amounts of proteins (>200 µg/sample). The activity of eNOS was measured by the ability of immunoprecipitated eNOS to convert [3H]arginine to [3H]citrulline as described in Materials and Methods. B, Serum- and steroid-starved cells in 24-well plates were equilibrated with fresh Kreb’s buffer for 4 h and then challenged with 10 nM of E2 (solid circle) or E2ß-BSA (open circle) for up to 60 min. Media were sampled at various time points for detection of total NOx contents by using a NO analyzer (NOA 280, Silvers Instruments) based on a reaction of conversion of nitrite/nitrate to NOx in the media. NOx production was calculated by a standard curve generated with sodium nitrate as the standard and normalized by the protein content of corresponding wells. C, Serum- and steroid-starved cells in 24-well plates were equilibrated with fresh Kreb’s buffer for 4 h, pretreated with 20 µM actinomycin-D for 1 h, and then challenged with 10 nM of E2 (solid circle) or E2ß-BSA for 10 min. Media were sampled for detection of total NOx contents by the StressXpress total NOx detection kit, normalized by the protein content of corresponding wells. The data were converted to fold over control and presented as mean ± SEM from three independent experiments using different cell preparations. *, P < 0.05 (vs. control).

View larger version (45K): [in this window] [in a new window]   FIG. 2. Evidence for membrane ERs in UAECs. A, Plasma membrane and nuclear extracts of UAECs were prepared as described in Materials and Methods. Proteins (50 µg/lane) were separated on 10% SDS-PAGE and transferred onto a PVDF membrane. ER protein was analyzed with a specific rabbit antihuman ER antibody (1:500, Geneka Biotechnology Inc.) following the manufacturer’s instructions. Recombinant human ER protein (50 ng) was run in parallel for positive control. B, Subconfluent UAECs grown on gelatin-coated glass coverslips for 2 d, washed with cold PBS, and fixed with 4% paraformaldehyde. After incubation with 1% BSA in PBS for 1 h, the cells were incubated with 50 µM of either BSA-FITC (control) or E2-BSA-FITC for 1 h in dark at room temperature. The cells were then washed four times (5 min each) with PBS and examined with a with a Nikon TE-300 fluorescence microscope. Images shown represent one of at least three similar independent experiments.

View larger version (42K): [in this window] [in a new window]   FIG. 3. E2 rapidly induces eNOS phosphorylation in uterine artery endothelial cells. Serum- and steroid-starved cells in 60-mm dishes were labeled with [32P]orthophosphate (0.1 mCi/ml) in phosphate-free DMEM for 4 h and then treated with vehicle, E2ß (10 nM), ionomycin (1 µM), and their combination for 10 min. Total cell lysates were prepared and eNOS was immunoprecipitated as described in Materials and Methods. The immunoprecipitates were analyzed with 10% SDS-PAGE, transferred onto a PVDF membrane, followed by autoradiography. A, A representative autoradiogram represents the phosphorylation of eNOS (upper panel). The membrane was immunoblotted with anti-eNOS mAb for monitoring equal amounts of eNOS were immunoprecipitated from each sample (lower panel). B, Data represent mean ± SEM of [32P]-incorporation into eNOS from three independent experiments. *, P < 0.05 (vs. control).

View larger version (31K): [in this window] [in a new window]   FIG. 4. E2 rapidly induces serine phosphorylation of eNOS in uterine artery endothelial cells. Serum- and steroid-starved cells in 100-mm dishes were treated with E2 (10 nM) for up to 60 min and then used for preparation of total cell lysates. Equal amounts of protein (>200 µg/sample) were used for immunoprecipitation of eNOS as described in Materials and Methods. The same amounts of mouse IgG were run in parallel for monitoring antibody specificity. The immunoprecipitates were size fractionated with 7.5% SDS-PAGE and transferred to a PVDF membrane. A, Serine phosphorylation of eNOS was analyzed with a phospho-eNOS pAb (1:1000) that reacts only with phosphorylated eNOS on serine1177 (upper panel). A portion of immunoprecipitates was analyzed by immunoblotting with anti-eNOS mAb (1:750) for ensuring equal loading (lower panel). B, Data represent mean ± SEM from three independent experiments using different cell preparations. *, P < 0.05 (vs. control).

View larger version (36K): [in this window] [in a new window]   FIG. 5. Effects of estrogen and ATP on intracellular Ca2+ mobilization in uterine artery endothelial cells. Serum- and steroid-starved cells at low density on glass dishes were loaded with Fura-2/AM and then challenged with 10 µM E2 or 30 µM ATP as described in Materials and Methods. Recordings of changes in the Fura-2 fluorescence ratio of a two- or three-cell group are shown.

View larger version (27K): [in this window] [in a new window]   FIG. 6. Temporal and concentration-dependent effects of E2 on phosphorylation of ERK2/1 in uterine artery endothelial cells. Serum- and steroid-starved cells in six-well plates were treated with increasing concentrations (0.001–10 µM) of E2 for 10 min (A) or with 10 nM E2 for various time points as indicated in B. Total cell lysates (20 µg/lane) were separated on 10% SDS-PAGE and transferred onto a PVDF membrane. Phosphorylation of ERK2/1 was analyzed by immunoblotting with the Phosphoplus MAPK pAb (1:1000) recognizing only phosphorylated forms of ERK2/1. Data from three independent dose-response and time-course experiments are summarized in their corresponding panels. *, P < 0.05; **, P < 0.01 (vs. control).

View larger version (20K): [in this window] [in a new window]   FIG. 7. E2 rapidly activates Raf-1 and ERK2 in uterine artery endothelial cells. Serum- and steroid-starved cells in 100-mm dishes were treated with or without E2 (10 nM) for 10 min. Total cell lysates were prepared for measuring Raf-1 and ERK2 activity. A, Raf-1 was immunoprecipitated from 200 µg proteins, and its activity was measured by [32P]ATP incorporation into GST-tagged-MEK1 (1 µg) and its subsequent phosphorylation of GST-tagged-ERK2 (0.1 µg) by the two-step immunocomplex kinase assay as described in Materials and Methods. An autoradiogram from one of three independent experiments is shown in the upper panel. Lane 1, Raf-1 immunoprecipitate does not phosphorylate ERK2 in the absence of MEK1. Lane 2, MEK1 phosphorylates ERK2 in the absence of cellular protein. Lanes 3 and 4, Raf-1 initiated phosphorylation of MEK1 and K52R in Raf-1 immunocomplexes isolated from cells treated with control media (lane 3) and E2 (lane 4). [32P]-incorporation into substrates from three independent experiments expressed as means ± SEM is summarized in the lower panel. In each immunocomplex assay, the levels of 32P-ATP incorporation into MEK1 and K52R were obtained after subtraction of their corresponding levels of phosphorylation in the absence of Raf immunocomplexes. Autophosphorylation of MEK1 was 128 ± 19 cpm and phosphorylation of K52R in the presence of MEK1 was 215 ± 35 cpm. B, ERK2 immunoprecipitated from 200 µg protein per sample was incubated with GST-Elk-1 (0.5 µg) and [-32P]ATP for 20 min at 30 C. The reaction mixtures were then analyzed as described in Materials and Methods. An autoradiogram from one of three independent experiments is shown in the upper panel. [32P]-incorporation into GST-Elk-1 from three independent experiments expressed as means ± SEM is summarized in the lower panel. *, P < 0.05.

View larger version (40K): [in this window] [in a new window]   FIG. 8. Effects of ICI 182,780 and PD98059 on phosphorylation of ERK2/1 by estrogen in uterine artery endothelial cells. Serum- and steroid-starved cells in six-well plates were pretreated with or without 2 µM ICI 182, 780 for 60 min (A) or with or without increasing concentrations (1, 20, 50 µM) of PD98059 for 60 min (B), followed by E2 (10 nM) for 10 min. Total cell extracts (20 µg/lane) were separated on 10% SDS-PAGE and transferred onto a PVDF membrane. Phosphorylation of ERK2/1 was analyzed by immunoblotting with a Phosphoplus MAPK antibody (1:1000) that recognizes only phosphorylated forms of ERK2/1. Data represent means ± SEM from three independent experiments in A and one of at least three similar independent experiments is shown in B.

View larger version (32K): [in this window] [in a new window]   FIG. 9. Effects of PD98059 on estrogen-induced eNOS activation in uterine artery endothelial cells. Serum- and steroid-starved cells in 100-mm dishes were pretreated with or without PD98059 (20 µM) for 60 min, followed by treatment with or without E2 (10 nM) or 10 nM E2-BSA for 10 min. Total cell extracts were prepared and eNOS was immunoprecipitated from equal amounts of proteins (>200 µg/sample) and used for measuring specific eNOS activity or eNOS phosphorylation. A, The specific eNOS activity is measured by the ability of immunoprecipitated eNOS to convert [3H]arginine to [3H]citrulline as described in Fig. 1A. The results are converted to fold over control, and the data are pooled from three independent experiments. B, Serine phosphorylation of eNOS is measured by immunoblotting with a specific anti-eNOS pAb recognizing only eNOS phosphorylated on serine1177. Data represent means ± SEM of the ratio of densitometric densities between phosphorylated eNOS and eNOS from three independent experiments. *, P < 0.05.

View larger version (36K): [in this window] [in a new window]   FIG. 10. IGF-1, but not E2, rapidly activates the PI-3-kinase-Akt pathway in uterine artery endothelial cells. Serum- and steroid-starved cells in six-well plates were pretreated without the specific PI-3-kinase inhibitor LY294004 (20 µM) for 60 min and then treated with E2 (10 nM) or IGF-1 (50 ng/ml) for 10 min (A). The cells were pretreated with increasing concentrations of LY294002 (0, 0.02, 0.2, 2, 20 µM) for 60 min and then treated with IGF-1 (50 ng/ml) for 10 min (B). Total cell lysates (20 µg/lane) were separated on 10% SDS-PAGE and transferred onto a PVDF membrane. Phosphorylation of Akt was analyzed by immunoblotting with a phospho-specific Akt antibody (1:1000) (upper image panels in A and B). The membranes were stripped and reprobed with an anti-Akt pAb (1:1000) to ensure equal loading of the samples (lower image panels in A and B). Data represent means ± SEM from three independent experiments in A, and one of three similar independent experiments is shown in B.

View larger version (29K): [in this window] [in a new window]   FIG. 11. Effects of ICI 182,780, LY294002, and PD98059 on E2-induced NOx production in uterine artery endothelial cells. Serum- and steroid-starved cells in 24-well plates were equilibrated with fresh Kreb’s buffer for 4 h. The buffer was replaced with fresh buffer, and the cells were pretreated ICI 182,780 (2 µM), PD98059 (20 µM), or LY294004 (20 µM) for 1 h. The cells were then challenged with or without E2 (10 nM) for 60 min. Media were sampled for the measurement of total NOx contents as described in Fig. 1B. NOx production was calculated by a standard curve generated with sodium nitrate as the standard and normalized by the protein content of corresponding wells and converted to percentage of control. Data shown are mean ± SEM from three independent experiments using three different cell preparations. *, P < 0.05.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

When many of us are thinking of estrogen actions, an image of nuclear responses first pops up in our mind because estrogen receptors (ER and ERß) belong to the superfamily of ligand-activated transcription factors working in the nucleus to regulate gene expression (11). However, during the last few years, rapid nongenomic pathways of estrogen actions, which are outside the nucleus and independent of gene transcription, have been proposed based on data from extensive studies using many endothelial and nonendothelial cell systems (18, 19, 31). Binding studies by using either radioactive labeled estrogens or fluorescent-labeled estrogen-BSA conjugates as the ligand have shown that E2 binding sites are present on the plasma membrane in various cell types including endothelial cells (39, 40, 41, 42, 43, 44, 45, 46, 47, 48). The nongenomic cellular responses of estrogen are presumptively mediated by these plasma membrane receptors (18, 19, 31, 41). Recent studies show that membrane ER and ERß are derived from the same transcripts as their nuclear compartments (31, 45). In addition, a 46-kDa truncated ER isoform is detectable on the plasma membrane of endothelial cells (44). In this report, we have detected relative lower levels of classic about 67-kDa ER protein in isolated plasma membrane than nuclear extracts in UAECs by immunoblotting with a specific antihuman ER antibody. Similar to previous reports using nonendothelial (45, 46, 47) and endothelial cell types (16, 48), our binding studies, by using E2-BSA-FITC as a ligand, also show specific membrane estrogen binding sites in UAECs.

To further support the presence of functional membrane ERs on UAECs, we performed series experiments to compare the effects of the membrane impermeable E2 analog E2-BSA with these of E2. E2-BSA conjugate has been widely used as a selective membrane ER ligand by many investigators to study various nongenomic actions of estrogen in a variety of cell types including human EA.hy926 endothelial cell line (16) and bovine aortic artery endothelial cells (48) as well as human arterial endothelial cells (49). In all of these studies, except one showing differential cellular effects of E2 and E2-BSA (50), it has shown that E2 and E2-BSA possess similar acute cellular responses in all cell types studied (16, 42, 45, 46, 47, 48, 49, 50). Our data show that E2-BSA is capable of mimicking E2-induced eNOS activation and NO_x production. Moreover, both E2 and E2-BSA-induced NO production in UAECs is unaltered in the presence of actinomycin-D. These findings in UAECs are consistent with previous reports showing that in pulmonary artery endothelial cells the effects of estrogen on rapid NO production is independent on nuclear transcription activities (15). In addition, E2-BSA induced rapid activation of eNOS-NO pathway in UAECs can be attenuated by pretreatment with ICI 182,789. These data show that the ER responsible for acute activation of the eNOS-NO pathway by estrogen in UAECs is most likely localized on the plasma membrane.

On estrogen stimulation, various rapid cellular responses including increased cAMP production (51), inositol phosphate turnover (52), Ca²⁺-mobilization (14, 46, 47), and activation of MAPK (15, 22, 24, 25) and Akt (16, 21, 23) can occur within in seconds to minutes. Mobilization of Ca²⁺ and activation of MAPK and PI-3-kinase/Akt are of specific interest because these signaling molecules are directly involved in the regulation of eNOS activity (15, 22, 23, 24, 25). eNOS is a Ca²⁺-dependent enzyme and binding to Ca²⁺-activated calmodulin is essential for its activity (32). However, a measurable rise in intracellular Ca²⁺ may not be required for eNOS activation by estrogen (53) or ceramide (54) in bovine aortic endothelial cells. Our present data show that in UAECs estrogen appears not to be able to induce a measurable rise in intracellular Ca²⁺ levels. However, our data also show that in UAECs the combination of Ca²⁺ mobilizing agent ionomycin and estrogen induced a much greater response in eNOS phosphorylation than that of either estrogen or ionomycin alone. These results suggest that the presence of Ca²⁺ is required for the full activation of eNOS in UAECs. A synergistic effect of estrogen and ionomycin on eNOS phosphorylation may be explained by the caveolae localization of eNOS in endothelial cells. Caveolae are 60–100 nm -shaped plasma membrane invaginations that are abundantly present in endothelial cells (55). In resting endothelial cells, eNOS is sequestered in the caveolae by directly binding to caveolin-1, the principal residual protein of caveolae (56). It has been recently shown that estrogen can induce a rise in Ca²⁺ levels in the microenvironment of caveolae in endothelial cells (57). Given the fact that eNOS is a Ca²⁺-dependent enzyme, it is therefore possible that in some types of endothelial cells although estrogen activation of eNOS does not accompany a measurable intracellular Ca²⁺ mobilization, estrogen may be able to increase caveolae Ca²⁺ levels for the full activation of eNOS. However, this theory has not yet been tested in UAEC.

Alternatively, other posttranslational mechanisms (31) such as phosphorylation by protein kinases Akt and MAPK, fatty acid modifications including palmitoylation and myristoylation, and interactions with caveolin-1 and heat shock protein-90 may be involved in acute activation of eNOS by estrogen in these endothelial cell types. In this study we focused on phosphorylation-dependent mechanism for eNOS activation by estrogen. Other investigators have recently shown a critical role of Akt in eNOS activation by estrogen in other endothelial cell types (16, 23, 31). Surprisingly, in UAEC Akt-dependent phosphorylation seems not to play a leading role in eNOS activation by estrogen because estrogen alone does not induce Akt phosphorylation and blocking Akt activation by pretreatment with LY294004 does not inhibit estrogen-induced NO_x production. Our data show that IGF-1 induces a burst Akt phosphorylation in UAECs. Although the physiological role of IGF-1-induced activation of Akt needs to be determined in UAECs, this result indicates that Akt does present in UAECs and can be activated by IGF-1. However, it is unclear at the moment why estrogen does not activate Akt in UAECs like it does in many other endothelial cells. Regardless, UAECs may possess a unique molecular mechanism for controlling eNOS activation and NO production than other endothelial cells.

Recent studies have also shown that activation of ERK2/1 may be stimulatory (15, 23, 25) or inhibitory (22) for eNOS activity, depending on the endothelial cell types studied. Our current data have shown that both E2 and E2-BSA rapidly induce phosphorylation and activation of ERK2/1 and that blocking ERK activation by PD98059 effectively inhibits estrogen-stimulated eNOS activity and NO_x production in UAECs. This leads us to conclude that activation of ERK2/1 plays an important role in eNOS activation by estrogen in UAECs. To further support a stimulatory effect of ERK2/1 on eNOS activation in UAECs, we recently reported that a strong positive correlation between phosphorylation of ERK2/1 and NO_x production exists in UAECs on stimulation with a number of other physiological stimuli of uterine vasodilatation (26). Our present data are consistent with the majority of studies showing a stimulatory role of ERK2/1 in the regulation of eNOS activity such as in pulmonary artery endothelial cells by estrogen (15) and human umbilical vein endothelial cells by adenosine (25) as well as bovine aortic endothelial cells by hydrogen peroxide (58), despite an inhibitory role of ERK2/1 in eNOS activation in bovine aortic endothelial cells by bradykinin (22). Our data show that E2-induced serine phosphorylation of eNOS could be blocked by pretreatment with a pharmacological ERK2/1 inhibitor PD98059. These results are consistent with the observation that in pulmonary artery endothelial cells, inhibition of ERK2/1 by PD98059 also results in an increase in eNOS activity (15), which suggest that serine phosphorylation of eNOS may be a mechanism by which ERK2/1 play an important role in stimulating eNOS activation by estrogen in UAECs and pulmonary artery endothelial cells. Although the role of ERK2/1 in eNOS phosphorylation by bradykinin in bovine aortic endothelial cells is opposite to that in UAECs and pulmonary artery endothelial cells, phosphorylation is also shown to be a mechanism whereby ERK2/1 inhibit eNOS activation in this endothelial cell type (22). Despite these studies showing a direct/indirect relationship between ERK2/1 and eNOS activation, the exact mechanism by which ERK2/1 regulate eNOS activation is currently still unresolved.

The MAPK cascades are evolutionarily conserved signal transduction modules that participate into a wide variety of biological responses. In vertebrates, multiple isoforms of MAPK have been identified and categorized into at least four subfamilies, i.e. ERKs, p38^mapk, and the Jun N-terminal kinases or stress-activated protein kinases as well as BMK/ERK5. The classical ERKs, ERK2/1, positioned downstream of Raf-1 and MEK1 and together comprise an orderly signaling cascade, can be activated by diverse extracellular stimuli including estrogen (59). In this report, we also demonstrated that estrogen activates Raf-1, the direct upstream kinase for MEK-1-dependent ERK2/1 activation in UAECs. Our results agree with others showing that estrogen activates Raf-1 in other cell types (23, 60). Because E2-BSA imitates E2-induced phosphorylation of ERK2/1 in UAECs, it is possible that estrogen activation of Raf-1-ERK2/1 signaling cascade is membrane ER mediated. However, a gap still exits between membrane ER and Raf-1 activation by estrogen in UAECs. Recent studies may have provided some hints toward this gap. For example, membrane ER and ERß are associated with G proteins that in turn initiate downstream signaling events (42, 61), tyrosine phosphorylation of Src (24), and trans-activation of epidermal growth factor receptor (62). These events may be responsible for Raf-1 activation by estrogen. Our more recent data also demonstrate that at least ERK2/1, p38^mapk, Jun N-terminal kinase 1, and ERK5 are expressed in ovine fetal placental artery endothelial cells and all of which can be activated on stimulation with hydrogen peroxide (63). In other cell types, it has been shown that estrogen differentially activates each of these MAPK modules that play distinct roles in regulating cell functions (42). However, whether MAPK family members other than ERK2/1 are also expressed and what role(s) they play in UAECs is waiting for further investigation.

In ovariectomized ewes, our recent data have shown that infusion of ICI 182,780 locally into the uterine circulation is able to inhibit, to a great extent, the rise of uterine blood flow induced by a bolus iv estrogen injection (64). These data clearly suggest that estrogen-induced uterine vasodilatation is an ER-mediated process in vivo. Although recent data suggest that vascular endothelium expresses both ER and ERß and that ERß possibly plays a more important cardiovascular protective role than ER (65), the role of different ER isoforms in estrogen actions in artery endothelium is still unclear. In particular, the ER isoform(s) responsible for the acute activation of eNOS by estrogen in uterine artery endothelium is an important and intriguing question to be answered. It was recently shown that ER is present in endothelial caveolae (31, 48) and that direct interaction with caveolin-1 targeting ER into caveolae mediates rapid estrogen responses (43). A so-called steroid receptor fast-action complex has been recently described in the caveolae of ovine fetal pulmonary artery endothelial cells. Estrogen binding to ER on the SRFC in endothelial caveolae leads to G activation. This results in downstream activation of MAPK and Akt signaling, stimulation of heat shock protein 90 binding to eNOS, and perturbation of the local calcium environment, leading to eNOS phosphorylation and calmodulin-mediated eNOS activation (31). More recently in vivo physiological studies using selective ER and ERß ligands have shown that the former ones are with much greater potency in stimulating uterine blood flow in ovariectomized ewes (66). According to these data, therefore, it is possible that the ER responsible for estrogen-induced uterine vasodilatation might be ER. Regardless, in this report our in vitro data have shown pretreatment with ICI 182,780 can inhibit acute activation of eNOS to produce NO and rapid activation of the Raf-1-MEK1-ERK2/1 signaling cascade.

These results, combined with the presence of ER protein in isolated plasma membrane and specific E-BSA-FITC binding sites on UAECs, highlight the importance of ER in estrogen actions in uterine artery endothelium in vitro. Furthermore, several lines of available evidence suggest that estrogen initiation of vasodilatation may not require de novo synthesis of eNOS but rather increased eNOS activity. First, in vascular endothelium eNOS is a constitutively expressed enzyme with a 20-h half-life (67). Second, in certain organs such as the mammary gland estrogen-induced vasodilatation is not accompanied by an increased endothelial eNOS expression (6). Third, acute treatment with estrogen stimulates very rapid NO production in endothelial cells without altering eNOS expression (Refs. 15 and 53 and the present study). Lastly, blockade of mRNA synthesis by actinomycin-D does not have any effects on estrogen-induced uterine blood flow (12, 13). Therefore, our present study showing acute activation of eNOS to produce NO by estrogen is mediated largely by a membrane ER-dependent ERK pathway in UAECs appears to describe a critical molecular mechanism responsible for at least the initiation of estrogen-induced uterine vasodilatation.

	Footnotes

 
This work was supported in part by NIH Grants HL70562 (to D.B.C.), HL64601 (to I.M.B.), HL64703 (to J.Z.), HL49210, HD33255, HL57653, and HD38843 (to R.R.M.) and a pilot grant from the University of Wisconsin-Madison National Institute of Environmental Health Sciences Center (to D.B.C.).

Abbreviations: E2, Estradiol-17ß; eNOS, endothelial NO synthase; ER, estrogen receptor; FCS, fetal calf serum; FITC, fluorescein isothiocyanate; GST, glutathione-S-transferase; M-199, medium-199; mAb, monoclonal antibody; MEK1, mitogen-activated protein kinase kinase 1; NO, nitric oxide; NOx, nitrite/nitrate; pAb, polyclonal antibody; PI-3-kinase, phosphoinositol-3-kinase; PVDF, polyvinyl difluoride; UAEC, uterine artery endothelial cell.

Received May 1, 2003.

Accepted for publication September 15, 2003.

	References

Top Abstract Introduction Materials and Methods Results Discussion References

 

1. MaRkee JE 1932 Rhythmic vascular uterine changes. Am J Physiol 256:E690–E698
2. Magness RR, Rosenfeld CR 1989 The role of steroid hormones in the control of uterine blood flow. In: Rosenfeld CR, ed. The uterine circulation. Chap 10. Ithaca, NY: Perinatology Press; 239–271
3. Greiss Jr FC, Anderson SG 1970 Effect of ovarian hormones on the uterine vascular bed. Am J Obstet Gynecol 107:829–836[Medline]
4. Killam AP, Rosenfeld CR, Battaglia FC, Makowski EL, Meschia G 1973 Effect of estrogens on the uterine blood flow of oophorectomized ewes. Am J Obstet Gynecol 115:1045–1052[Medline]
5. Magness RR, Shaw CE, Phernetton TM, Zheng J, Bird IM 1997 Endothelial vasodilator production by uterine and systemic arteries II: pregnancy effects on NO synthase expression. Am J Physiol 272:H1730–H1740
6. Vagnoni KE, Shaw CE, Phernetton TM, Meglin BM, Bird IM, Magness RR 1998 Endothelial vasodilator production by uterine and systemic arteries. III. Ovarian and estrogen effects on NO synthase. Am J Physiol 275:H1845–H1855
7. Zervou S, Klentzeris LD, Old RW 1999 Nitric oxide synthase expression and steroid regulation in the uterus of women with menorrhagia. Mol Hum Reprod 5:1048–1054[Abstract/Free Full Text]
8. Van Buren GA, Yang DS, Clark KE 1992 Estrogen-induced uterine vasodilatation is antagonized by L-nitroarginine methyl ester, an inhibitor of nitric oxide synthesis. Am J Obstet Gynecol 167:828–833[Medline]
9. Rosenfeld CR, Cox BE, Roy T, Magness RR 1996 Nitric oxide contributes to estrogen-induced vasodilation of the ovine uterine circulation. J Clin Invest 98:2158–2166