Estrogen and Lipopolysaccharide Stimulation of Prostacyclin Production and the Levels of Cyclooxygenase and Nitric Oxide Synthase in Ovine Uterine Arteries¹

^a Department of Animal, Dairy, and Veterinary Sciences, Utah State University, Logan, Utah 84322 ^b Departments of Obstetrics/Gynecology, Perinatal Research Laboratories, and Meat/Animal Science, University of Wisconsin-Madison, Madison, Wisconsin 57311

	ABSTRACT

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION NOTE ADDED IN PRESS REFERENCES

 
Several enzymes play a role in vasodilation, including cyclooxygenase, which converts arachidonic acid into prostaglandins, and nitric oxide synthase, which converts arginine to citrulline and yields nitric oxide. The effects of endogenous and exogenous estrogen and lipopolysaccharide on uterine artery production of prostacyclin, and levels of cyclooxygenase and nitric oxide synthase were examined. Uterine arteries collected from ewes during the follicular (Day -1 to 0, Day 0 = estrus) or luteal (Day 10) phase were treated in vitro with lipopolysaccharide. In addition, ovariectomized ewes were treated in vivo with estradiol-17ß (5 µg/kg; 120 min) or a vehicle control; arteries from the uteri were treated in vitro with lipopolysaccharide. After 24 h of lipopolysaccharide treatment, culture media were collected for measurement of 6-keto-prostaglandin F₁ (the stable metabolite of prostacyclin). These uterine arteries were homogenized, and the level of cyclooxygenase and nitric oxide synthase was determined by Western analysis. Lipopolysaccharide stimulated (p < 0.02) prostacyclin production by uterine arteries from both follicular- and luteal-phase sheep although phase of the estrous cycle did not affect prostacyclin responses (p = 0.56) to lipopolysaccharide. In contrast, uterine arteries from ovariectomized sheep treated with estradiol-17ß produced more prostacyclin (p < 0.001) in response to lipopolysaccharide than did uterine arteries from ovariectomized sheep treated with the vehicle control. There was no effect of phase (follicular or luteal) of the estrous cycle on either cyclooxygenase-1 or -2 gene expression. Lipopolysaccharide increased (p = 0.0002) gene expression of cyclooxygenase-2, but not cyclooxygenase-1, in both follicular- and luteal-phase ewes, which was significantly correlated (r² = 0.91, p = 0.003) with uterine artery production of prostacyclin. Uterine arteries from follicular-phase sheep expressed significantly more nitric oxide synthase-III after lipopolysaccharide exposure than did uterine arteries from luteal-phase ewes (p = 0.03). In contrast, nitric oxide synthase-II was not detected in uterine arteries after lipopolysaccharide exposure. These results suggest that estrogen plays a role in regulating uterine artery responses to lipopolysaccharide.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION NOTE ADDED IN PRESS REFERENCES

 
Lipopolysaccharide (LPS) is the name given to the lipo-carbohydrate molecules that make up the outer membrane of gram negative bacteria [1]. Overproduction of secretory mediators (i.e., tumor necrosis factor [TNF], interleukins, and prostaglandins) from LPS-activated cells can result in endotoxemia, a pathological disease that often leads to cardiovascular failure and death [2]. Nearly all eukaryotic cells respond to LPS; picomolar concentrations are sufficient to stimulate cells of the immune, inflammatory, and vascular systems [3, 4].

The process of mating often permits introduction of LPS-bearing bacteria into the uterus [5]. LPS has been shown to cause fetal death and abortion in animals [6–10]. One mode of action of LPS is the initiation of cellular events that activate phospholipase-A2, which in turn releases arachidonate [11]. Released arachidonic acid is metabolized by cyclooxygenase (COX) and yields prostacyclin and other prostaglandins [12]. Prostacyclin is released by endothelium and acts as a vasodilator of vascular smooth muscle [6, 13]. Inhibition of COX, a rate-limiting enzyme in prostaglandin production, reduces the effects of LPS including pregnancy loss [1, 6]. There are two isoforms of COX, COX-1 and COX-2, which have similar activities but are encoded by different genes and have marked regulatory and functional differences [6]. For example, COX-1 is constitutively expressed in many tissues, including sheep uterine arteries [14], and COX-1 is believed to be involved in the homeostatic maintenance of the body [6]. In contrast, COX-2 is believed to have a "differentiative" role in cells and is highly inducible during inflammation and in response to LPS [6, 15].

Nitric oxide (NO) also acts as a vasodilator when produced by blood vessels [16, 17]. Production of NO is enhanced after LPS exposure, probably through increases in the enzyme nitric oxide synthase-II (also known as inducible NOS or iNOS) [18]. NOS-II, like COX-2 because of its inducible nature, has been implicated in immune response, while NOS-III (also known as endothelial NOS or eNOS), like COX-1 because of its constitutive nature, has been implicated in homeostasis [19]. NOS-III gene expression is elevated in the sheep during periods of high estrogen (follicular phase) compared to those of lower estrogen (luteal phase) (see Note Added in Press) and is elevated during the last trimester of pregnancy (a prolonged period of high estrogen) [20]. NOS-III has been shown to mediate the relaxing effects of estradiol-17ß (E₂) in rat aorta and rabbit femoral arteries [21]. Recently a role for NO in the regulation of COX activity has been proposed. In the rat, LPS was shown to activate both the COX and NOS systems within a similar time frame, and there was an increase in 6-keto-prostaglandin F₁ (6-keto-PGF₁, the stable metabolite of prostacyclin) in the urine and plasma due to the elevation in COX [22]. It was shown that NO inhibitors inhibited production of this prostaglandin, but not through inhibition of COX directly. It has been suggested that NO directly activates the COX enzymes or inhibits autoinactivation of the COX enzymes [23, 24]. Regardless of the mechanism, the potential regulatory interaction between NO and COX is an important observation.

Several studies have shown E₂ to be important in regulating cellular responses to LPS. Sugimoto et al. [25] showed that simultaneous stimulation of mice with E₂ and LPS resulted in a more rapid increase of hepatic spermidine/spermine N¹-acetyltransferase activity compared to LPS alone. Moreover, TNF mRNA expression by mouse macrophages was enhanced in LPS-treated cells exposed to E₂ [26]. Miyagi et al. [27] reported that high E₂ doses resulted in an enhanced production of prostaglandin E₂ (PGE₂) by LPS-stimulated human peripheral monocytes, while low doses of E₂ suppress the production of PGE₂ by these monocytes. Leslie and Dubey [28] observed that human monocytes collected during the luteal phase produce more PGE₂ and prostacyclin in response to LPS than do monocytes collected from men. These authors further showed that when LPS was absent, there were no differences in PGE₂ or prostacyclin production throughout the menstrual cycle in the female (follicular vs. luteal phase), or between females and males [28]. Taken together, these data suggest a role for estrogen in the regulation of cellular response to LPS. It may be that estrogen acts to "prime" cells/tissues for an increased response to LPS. This enhanced response may be beneficial to the host as prostacyclin appears to improve tissue blood flow, reduce metabolic acidosis, and act in a cytoprotective manner after LPS insult [29].

The objective of this study was to determine the effect of endogenous estrogen (as in the follicular phase) or exogenous estrogen on uterine artery response to in vitro LPS exposure by evaluating 6-keto-PGF₁ production by these uterine arteries as well as COX-1, COX-2, NOS-II, and NOS-III levels in tissue homogenates of these uterine arteries. COX-1, COX-2, NOS-II, and NOS-III were examined as they represent both constitutive (COX-1 and NOS-III) and inducible (COX-2 and NOS-II) isoforms of vasodilatory enzymes.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION NOTE ADDED IN PRESS REFERENCES

 
Procedures for animal handling were approved by the University of Wisconsin-Madison Research Animal Care and Use Committees of both the Medical School and the College of Agriculture and Life Science, and follow the National Institutes of Health guide and the Report of the American Veterinary Medical Association Panel on Euthanasia.

Follicular- and Luteal-phase Ewes: Synchronization

Polypay and mixed Western breed ewes (50-60 kg) were observed daily for signs of behavioral estrus using a vasectomized ram. Ewes (n = 8) exhibiting normal estrous cycles (16-18 days) were given 2 injections i.m., 5 mg each, of PGF₂ (Lutalyse; UpJohn, Kalamazoo, MI) 4 h apart. Ewes were then monitored for behavioral estrus at approximately 48 h after the first PGF₂ injection. Ewes exhibiting estrus (Day 0) were randomly paired and assigned into one of two groups: follicular (Day -1 to 0, n = 4) or luteal (Day 10, n = 4). Eight days after the first sign of estrus for each pair (n = 4), the ewe assigned to the follicular group was given PGF₂ as described above, and the ewe assigned to the luteal group was given saline (2 injections i.m., equivalent volume to that of PGF₂, 4 h apart). This synchronization protocol has been reported [30] to result in the appearance of estrus in follicular phase ewes within 48 h after the first injection, while luteal-phase ewes do not show estrus at this time. Follicular-phase ewes were killed 44 h after the first injection of PGF₂ (Day -1 to 0), and the paired luteal-phase ewe was killed on the same day (Day 10). Uterine arteries were collected as described below. Ovarian structures were recorded as follows: small follicles 1-3 mm were counted, and medium (4–5 mm) and large ( 6 mm) follicles were measured. In addition, the size of the corpus luteum was measured, and its appearance (vascular or blanched) was recorded.

Ovariectomized Ewes with or without Exogenous E₂

Polypay and mixed Western breed ewes (n = 12; 50–60 kg) exhibiting normal estrous cycles were ovariectomized (OVX). After at least 10 days, a 19-gauge polyvinyl indwelling catheter was introduced into the jugular vein via a percutaneous needle puncture and advanced to the level of the right ventricle, and all ewes were primed with 1 µg/kg body weight E₂ i.v. for 5 days to increase uterine artery responsiveness to estrogen. Injections were followed immediately with a 5-ml sterile saline flush. The E₂ was dissolved in sterile 95% ethanol and stored at 4°C at a stock concentration of 1 mg/ml. The E₂ priming period was followed by a 2- to 4-day withdrawal period. Previous studies have established that the maintenance of maximal, normal, and consistent E₂-induced increases in uterine blood flow and cardiac output (systemic flows) were achieved with this priming regime [31, 32]. After the withdrawal period, and on the day before or day of the study, ewes were transferred into an open-top "metabolic" cart. An observation control period of not less then 60 min was allowed, after which time ewes were randomly paired, and 6 ewes were given 5 µg/kg body weight E₂ i.v. and another 6 ewes were given the vehicle control (VEH; in saline); injections were followed immediately with a 5-ml sterile saline flush. The E₂ stock solution was diluted in sterile saline to achieve the appropriate dilution for each animal; the vehicle consisted of 95% ethanol in sterile saline, as described for E₂. One hundred-twenty minutes after treatment (E₂ and VEH), tissue was collected from anesthetized ewes. The dose of E₂ and time of tissue collection were specifically chosen from previous dose-response and time-course studies that demonstrated when maximal and steady-state uterine and systemic vasodilation occurs [31, 32].

Procurement of Arteries

Paired ewes (OVX-E₂) and OVX-VEH; follicular + luteal) were subjected to nonsurvival surgery using general anesthesia (Na⁺ Pentobarbital; Nembutal, 20–50 mg/kg) in order to maintain tissue perfusion and oxygenation throughout the time of tissue collection. The uterus and the mesometrium were rapidly excised and placed into sterile PBS (10 mM sodium phosphate, 0.15 M NaCl, pH 7.4) as previously described [20]. Uterine arteries were dissected free of connective tissue, fat, and veins; and the arteries were rinsed free of blood. The exterior, but not the lumen, of the arteries was gently rinsed in PBS containing 1% penicillin-streptomycin to remove debris and blood, cut into 2-cm pieces, and placed into 2 ml sterile RPMI-1640 containing 1% penicillin-streptomycin with 0, 1, or 10 µg/ml LPS (0111:B4, Sigma, St. Louis, MO) for 24 h at 37°C in a shaking water bath. Media and supplements, purchased from Gibco (Gaithersburg, MD). were free of phenol red and endotoxin. Cultures were performed in serum-free conditions. Each treatment was performed in duplicate. After 24 h, supernatants were collected and stored at -20°C, and tissue segments were weighed and then stored at -20°C in homogenizing buffer (PBS, 1 mM PMSF, 5 µg/ml leupeptin, 5 µg/ml aprotinin, and 0.1% ß-mercaptoethanol; all from Sigma).

Analysis of Supernatants for 6-Keto-PGF₁

Culture media from each artery were diluted (1:1000) and then analyzed for 6-keto-PGF₁ by enzyme immunoassay (EIA; Cayman, Ann Arbor, MI) according to the manufacturer's protocol. Samples were analyzed in triplicate, and means were calculated for each treatment; these values were then divided by the weight of the tissue so that final numbers represented nanograms of 6-keto-PGF₁ per milligram of tissue wet weight produced over the 24-h culture period.

Preparation of Tissue for Western Analysis

Uterine artery segments were solubilized by grinding in homogenization buffer (described above) and sonicated as previously described [14, 20]. Connective tissue was removed from vessel preparations by centrifugation, and the protein concentration of samples was determined by a Lowry assay procedure (Bio-Rad, Hercules, CA). Blots were designed so that duplicates of all treatments (0, 1, 10 µg/ml LPS) for an individual animal and preplanned pairs of animals (OVX-E₂ and OVX-VEH; follicular and luteal) were represented on a single blot within each experiment. Equal concentrations of protein (50 µg/ml; except in the case of COX-1 analysis, see below) were resolved on a 7.5% polyacrylamide gel with 0.1% SDS at 100 V for2–3 h at room temperature before transfer onto Immobilon-P membrane at 100 V for 2 h at 4°C using the Mini-Protean II system (Bio-Rad). The membrane was probed as described by Amersham (Arlington Heights, IL) using their Enhanced Chemiluminescence (ECL) kit and exposure to hyperfilm (15 min).

Western Analysis of Tissue

COX-1 The primary antibody used to detect COX-1 was produced by Cayman and was used at a concentration of 1:4000. The secondary antisera was a horseradish peroxidase-conjugated donkey anti-rabbit (Amersham) and was used at a 1:8000 dilution. A positive control for COX-1 (ram seminal vesicle; Cayman), a COX-2 control (purified COX-2 from sheep cotyledon; Cayman), and molecular weight standards were included on each blot. The COX-1 antibody was reported by the supplier, and validated in our laboratory, to recognize only COX-1 and not COX-2. Equal concentrations of protein (50 µg/ml for OVX animals, 50 µg/ml for the first pair of intact ewes, and 30 µg/ml for the remaining 3 pairs of intact ewes) were resolved on gels as described above. Data were converted to a per nanogram of protein densitometric value and then analyzed as described below.

COX-2 The primary antibody used to detect COX-2 was produced by Cayman and was used at a dilution of 1:4000; the secondary antisera was a horseradish peroxidase-conjugated donkey anti-rabbit (Amersham) and was used at a 1:8000 dilution. A positive control, COX-2 (purified COX-2 from sheep cotyledon; Cayman), was included on each blot, and molecular weight standards were included on most blots. The COX-2 antibody was shown to be specific for COX-2 and, as previously reported, does not recognize COX-1 [33].

NOS-II A polyclonal antibody (Affinity Bioreagents, Golden, CO) produced in rabbit was used to detect NOS-II. The antibody was initially screened using stimulated RAW (a mouse cell line shown to produce NOS-II in response to LPS and interferon-) cell lysates to determine the optimal dilution of the primary and secondary antibody.

To determine whether this antibody recognized sheep NOS-II, tissue from a sheep given a systemic dose of LPS at a level that causes endotoxemia (kindly provided by Dan Traeber, PhD, University of Texas Medical Branch at Galveston) was analyzed. Tissue included lung, liver, and kidney. Similar tissue was collected from a control animal (also provided by Dan Traeber). Each tissue was processed as described above, and 50 µg of protein from each tissue and animal (LPS-treated or control) was loaded on duplicate gels and analyzed as described above; a rainbow molecular weight marker (Amersham) was included. One of the duplicate membranes was incubated with primary antibody at a 1:8000 dilution for 2 h at room temperature. Afterward, this membrane was incubated for 1 h in horseradish-peroxidase conjugated donkey anti-rabbit serum (1:8000; Amersham). To further demonstrate that NOS-II staining was specific, the duplicate blot was subjected to the above protocol except that the membrane was incubated in primary antibody buffer only for 2 h during the "primary antibody" step (secondary antibody only protocol).

NOS-III The previously reported protocol [20] to detect and quantify NOS-III in sheep arteries was used. The primary antibody used to detect NOS-III was a monoclonal antibody from Transduction Laboratories (Lexington, KY; 1:750 dilution) and a control, either ovine placental endothelial cell (OPEC) lysates or human umbilical vein endothelial cell (HUVEC) lysates, was included on each blot. The secondary antisera was a horseradish peroxidase-conjugated goat anti-mouse (Amersham), and this was usedat a 1:3000 dilution. This primary antibody was reported by the supplier to specifically recognize NOS-III and not NOS-II.

Densitometric Analysis

The levels of COX-1, COX-2, and NOS-III, were determined by scanning densitometry using an Alpha Imager (Alpha Inotech Corp., San Diego, CA). The validation of this protocol was determined by scanning blots of serial dilutions of the standard for each of COX-1, COX-2, and NOS-III. All of the standard curve blots were linear within the range of protein levels used in this study (data not shown).

Statistical Analysis

Data describing the synchronization protocol were analyzed by ANOVA, and means were separated by Least Significant Difference. All data collected from uterine arteries treated with LPS in vitro were analyzed by ANOVA using a completely randomized design with split plot arrangements of in vitro treatments. Phase (follicular vs. luteal) or treatment (OVX-E₂ vs. OVX-VEH) was the whole plot effect and the dose of LPS (0, 1, 10 µg/ml) the subplot effect. Means were separated using single degree of freedom orthogonal contrasts to test for the effects of LPS addition (0 µg/ml LPS vs. 1 and 10 µg/ml LPS), LPS level (1 µg/ml LPS vs. 10 µg/ml LPS), and the interactions of LPS addition and LPS level with phase (follicular or luteal) or treatment (OVX-E₂ or OVX-VEH). In addition, simple linear regression was used to explore the relationship between the means of COX-1, COX-2, and NOS-III levels in uterine artery homogenates and the means of 6-keto-PGF₁ produced in the culture media from the same uterine arteries collected from follicular- and luteal-phase ewes.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION NOTE ADDED IN PRESS REFERENCES

 
Follicular- and Luteal-phase Ewes: Synchronization Results

The presence and the size of ovarian structures, i.e., estrogen-producing follicles and progesterone-producing corpora lutea, were determined. There was no difference in the number (19.62 ± 2.16, follicular, vs. 17.55 ± 1.68, luteal) of small (1-3 mm) follicles or the size (4.27 ± 0.14 mm, follicular, vs. 4.28 ± 0.18 mm, luteal) of medium (4-5 mm) follicles between follicular- and luteal-phase ewes produced by this PGF₂ synchronization protocol. However, follicular-phase animals had larger (9.50 ± 1.25 mm; p < 0.03) large ( 6 mm) follicles than luteal-phase ewes (6.80 ± 0.52 mm). In addition, luteal-phase ewes had larger (10.30 ± 0.30 mm; p < 0.001), more vascular corpora lutea than follicular-phase ewes (6.54 ± 0.52 mm), which had blanched corpora lutea. None of the follicular phase ewes had corpora hemorrhagica (recently ovulated follicles). These data are consistent with a high-estrogen, low-progesterone state in follicular-phase ewes and a high-progesterone, low-estrogen state in luteal-phase ewes as previously reported [34].

6-Keto-PGF₁ in Uterine Artery Supernatants from Follicular- and Luteal-phase Ewes

There was no effect (p = 0.56) of phase (follicular vs. luteal) on uterine artery 6-keto-PGF₁ response to LPS. Both LPS addition (0 µg/ml LPS < 1 and 10 µg/ml LPS, p = 0.004) and LPS level (1 µg/ml < 10 µg/ml LPS, p = 0.02) resulted in an enhanced production of 6-keto-PGF₁ (Fig. 1). There was a 10-fold greater 6-keto-PGF₁ production by uterine arteries from intact (follicular or luteal) ewes than from OVX ewes regardless of treatment, as can be seen by comparing the y axes of Figures 1 and 2.

View larger version (22K): [in this window] [in a new window]   FIG. 1. Uterine arteries from follicular-phase (Day -1 to 0, Day 0 = estrus, n = 4) and luteal-phase (Day 10, n = 4) ewes were cultured with 0, 1, or 10 µg/ml LPS for 24 h. The amount of 6-keto-PGF1 (ng/mg of uterine artery tissue) in the media was determined by EIA. There were significant effects of LPS addition (p = 0.004) and LPS level (p = 0.02) on uterine artery production of 6-keto-PGF1.

View larger version (20K): [in this window] [in a new window]   FIG. 2. Uterine arteries from ovariectomized animals treated in vivo with vehicle (VEH, n = 6) or E2 (E2ß, n = 6) were cultured with 0, 1, or 10 µg/ml LPS for 24 h. The amount of 6-keto-PGF1 (ng/mg of uterine artery tissue) in the media was determined by EIA. There was a significant treatment effect (p < 0.001) and a significant LPS addition x treatment (E2 > VEH) interaction (p = 0.001) on uterine artery production of 6-keto-PGF1.

6-Keto-PGF₁ Production by Uterine Arteries from OVX Ewes with or without Exogenous E₂

In OVX ewes, there was a significant (p < 0.001) treatment effect (E₂ > VEH) on uterine artery production of 6-keto-PGF₁ in response to LPS (Fig. 2). There also was a significant (p = 0.001) LPS addition x treatment (E₂ > VEH) interaction, but LPS level (1 µg/ml LPS = 10 µg/ml LPS) x treatment (E₂ vs. VEH) was not significant (p = 0.33), suggesting that the response to LPS by uterine arteries from OVX-E₂ ewes had already reached maximum at 1 µg/ml LPS.

COX-1 and -2 Levels in Uterine Artery Homogenates of Follicular- and Luteal-phase Ewes

COX-1 was identified by Western analysis as a single band at an approximate molecular mass of 70 kDa (data not shown). There was no effect (p = 0.51) of phase (follicular vs. luteal) on COX-1 level in uterine arteries (Fig. 3). Moreover, addition of LPS (0 µg/ml LPS vs. 1 and 10 µg/ml LPS; p = 0.40) and LPS level (1 µg/ml vs. 10 µg/ml LPS; p = 0.13) did not influence COX-1 level in uterine arteries. Furthermore, there was no relationship between COX-1 levels in uterine arteries and 6-keto-PGF₁ produced in vitro by the same uterine arteries (r² = 0.32, p = 0.25).

View larger version (26K): [in this window] [in a new window]   FIG. 3. Uterine arteries from follicular-phase (Day -1 to 0, Day 0 = estrus, n = 4) and luteal-phase (Day 10, n = 4) ewes were cultured with 0, 1, or 10 µg/ml LPS for 24 h. The level of COX-1 per sample was determined from Western blots by scanning densitometry and adjusted to the original protein concentration added to the gel (densitometric units/ng protein). The level of COX-1 in uterine artery homogenates was not altered by phase (p = 0.51), LPS addition (p = 0.40), or LPS level (p = 0.13).

COX-2 was identified in all samples at an approximate molecular mass of 72 kDa and was always in doublet form (Fig. 4). There was no effect (p = 0.42) of phase (follicular vs. luteal) on COX-2 level in uterine arteries (Fig. 5). However, LPS addition (0 µg/ml LPS < 1 and 10 µg/ml LPS) resulted in a significant increase (p < 0.001) in COX-2 level in uterine arteries from both phases of the estrous cycle. There was a trend (p = 0.095) for LPS level (1 µg/ml < 10 µg/ml LPS) to influence COX-2 level in uterine arteries from follicular- and luteal-phase ewes. The pattern of COX-2 level within the uterine arteries following exposure with 0, 1, or 10 µg/ml LPS was significantly correlated (r² = 0.91, p = 0.003) with the pattern of 6-keto-PGF₁ produced by the same uterine arteries in vitro.

View larger version (16K): [in this window] [in a new window]   FIG. 4. Representative COX-2 staining of uterine artery homogenates from one follicular-phase and one luteal-phase animal. Segments of uterine arteries from a single animal were cultured in duplicate with 0, 1, or 10 µg/ml LPS for 24 h. COX-2 was identified by Western/ECL analysis at an approximate molecular mass of 72 kDa and in doublet form.

View larger version (24K): [in this window] [in a new window]   FIG. 5. Uterine arteries from follicular-phase (Day -1 to 0, Day 0 = estrus, n = 4) and luteal-phase (Day 10, n = 4) ewes were cultured with 0, 1, or 10 µg/ml LPS for 24 h. The level of COX-2 per sample was determined from Western blots by scanning densitometry. There was no effect (p = 0.42) of phase on the level of COX-2 protein. The level of COX-2 protein in uterine arteries was increased (p < 0.001) by addition of LPS. There was a trend (p = 0.095) for LPS level to influence COX-2 protein level in uterine arteries from follicular- and luteal-phase ewes.

COX-1 and -2 Levels in Uterine Artery Homogenates of OVX Ewes with or without Exogenous E₂

COX-1 was barely detectable in uterine arteries from both VEH- and E₂-treated OVX ewes (data not shown). Densitometry was not performed on these blots because of the very low COX-1 signal. However, there appeared to be no difference in intensity of COX-1 signal in VEH- and E₂-treated ewes regardless of LPS treatment. No COX-2 was detected in the homogenates of uterine arteries collected from OVX ewes (data not shown).

Validation of NOS-II Western Analysis Protocol

A positive band for NOS-II at an approximate molecular mass of 130 kDa was detected in stimulated RAW cell lysates and in kidney and liver (but not lung) tissue from sheep treated in vivo with LPS (data not shown). No NOS-II was detected in the tissue from the control animal. Furthermore, the 130-kDa NOS-II band was not detected when the duplicate membrane was subjected to the secondary-antibody-only protocol. NOS-II was not detected in uterine arteries of follicular- or luteal-phase ewes or in uterine arteries of OVX ewes.

NOS-III Levels in Uterine Artery Homogenates of Follicular- and Luteal-phase Ewes

NOS-III was readily detected in uterine arteries of follicular- and luteal-phase ewes, and there was a significant effect of phase (follicular > luteal; p = 0.03) on the level of NOS-III found in uterine artery homogenates (Fig. 6). While there was no significant overall effect (p = 0.14) of LPS addition (0 µg/ml LPS vs. 1 and 10 µg/ml LPS) on NOS-III level in artery homogenates, the higher dose of LPS resulted in more (10 µg/ml LPS > 1 µg/ml LPS; p = 0.025) NOS-III in artery homogenates. There was a trend (p = 0.08) for an interaction of LPS with phase: NOS-III increased in response to increasing LPS concentrations in follicular-phase but not luteal-phase ewes (Fig. 6). In follicular-phase ewes, NOS-III levels in uterine arteries after exposure to 0, 1, or 10 µg/ml LPS was significantly correlated (r² = 0.99, p = 0.06) with the amount of 6-keto-PGF₁ produced by these arteries in vitro. However, there was a lack of correlation between the level of NOS-III in luteal-phase uterine arteries and the amount of 6-keto-PGF₁ produced by these arteries after LPS exposure in vitro (r² = 0.18, p = 0.72).

View larger version (25K): [in this window] [in a new window]   FIG. 6. Uterine arteries from follicular-phase (Day -1 to 0, Day 0 = estrus, n = 4) and luteal-phase (Day 10, n = 4) ewes were cultured with 0, 1, or 10 µg/ml of LPS for 24 h. The level of NOS-III per sample was determined from Western blots by scanning densitometry. There was an effect (p = 0.03) of phase (follicular > luteal) on NOS-III level in uterine artery homogenates. The level of NOS-III protein in uterine arteries was not significantly increased (p = 0.14) by LPS addition, but was increased (p = 0.025) by LPS level.

NOS-III Levels in Uterine Artery Homogenates of OVX Ewes with or without Exogenous E₂

NOS-III was readily detected in uterine arteries from OVX ewes (Fig. 7). Treatment of OVX ewes with E₂ did not alter (p = 0.58) NOS-III level in uterine arteries compared to control (VEH). Neither the addition of LPS (p = 0.57) nor the level of LPS (p = 0.83) resulted in an elevation in NOS-III in uterine arteries collected from OVX sheep of either group (VEH, E₂).

View larger version (25K): [in this window] [in a new window]   FIG. 7. Uterine arteries from ovariectomized animals treated in vivo with vehicle (VEH, n = 6) or E2 (E2ß, = 6) were cultured with 0, 1, or 10 µg/ml of LPS for 24 h. The level of NOS-III per sample was determined from Western blots by scanning densitometry. There was no effect (p = 0.58) of in vivo treatment (VEH or E2), LPS addition (p = 0.57), or LPS level (p = 0.83) on NOS-III level in uterine arteries from ovariectomized ewes.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION NOTE ADDED IN PRESS REFERENCES

 
Uterine arteries collected from ewes during the follicular- and luteal phase produced more 6-keto-PGF₁ after LPS exposure compared to controls. However, there was no difference in 6-keto-PGF₁ production between follicular- and luteal-phase uterine arteries after LPS exposure. The lack of difference in 6-keto-PGF₁ production between follicular-phase (high estrogen) and luteal-phase (low estrogen) uterine arteries after LPS exposure may be because only a small amount of estrogen is needed to enhance arterial response to LPS. It was noted that although the size of the large follicles was greatly reduced in the luteal-phase ewes compared to follicular-phase ewes, follicular activity was similar for these two phases. This observation suggests that small amounts of estrogen or other follicular products produced throughout the estrous cycle may act to stimulate prostacyclin production by uterine arteries after LPS exposure. Another explanation is that progesterone, which is high during the luteal phase, acts to stimulate an LPS-mediated prostacyclin response by uterine arteries in a fashion similar to that of estrogen.

Estradiol treatment of OVX ewes was necessary for uterine arteries to respond to LPS with increases in 6-keto-PGF₁. This observation is consistent with others that have shown that E₂ enhances cellular responses to LPS [25–28]. There was substantially more (10-fold) 6-keto-PGF₁ produced by uterine arteries from intact ewes (follicular or luteal phase) than from OVX ewes regardless of E₂ treatment, suggesting that the hormonal profile found in intact ewes provides a more favorable environment for prostacyclin production than the environment provided by estrogen alone. That uterine arteries from VEH-treated OVX sheep did not respond to LPS with an increased production of 6-keto-PGF₁, while arteries from OVX-E₂, follicular, and luteal phase did respond to LPS supports the hypothesis that estrogen is important in regulating this uterine artery response to LPS. Estrogen may increase the ability of uterine arteries to produce prostacyclin after LPS exposure by increasing the number of LPS receptors (sCD14) or the amount of LPS binding protein found in the circulation [35, 36]. The presence of sCD14 or LPS binding protein is necessary for endothelial and vascular smooth muscle cells to produce prostacyclin, cytokines, and adhesion molecules in response to LPS [35, 36].

In the uterine arteries from follicular- and luteal-phase ewes, elevations of 6-keto-PGF₁ production in culture media after LPS exposure were found to be highly correlated with elevations in COX-2, but not COX-1, in these arteries. Elevations of COX-2 in response to LPS is consistent with the proposed role of COX-2 in differentiation and inflammation [6, 15] and with the reported associated increase of COX-2 and another prostaglandin, PGF₂, by sheep endometrial explants [37]. The lack of a relationship between 6-keto-PGF₁ in the culture media and COX-1 level in the uterine arteries in the present study is consistent with the proposed role for COX-1 in homeostatic maintenance [6, 15] and is consistent with previously reported observations that no relationship exists between PGF₂ accumulation in the media and COX-1 levels in tissues from sheep endometrial extracts [37]. In addition, treatment of OVX ewes with E₂ did not increase COX-1 in endometrial extracts [37]. It appears that increases in prostacyclin and PGF₂ in the uterine arteries and the endometrium, respectively, is performed by the same enzyme, since COX-2 was associated with increases in both 6-keto-PGF₁ (the present study) and PGF₂ [37].

COX-2 was not detected in the uterine arteries from OVX animals. This observation is consistent with the study of Charpigny et al. [37], who showed that prolonged exposure to progesterone was necessary for COX-2 gene expression in endometrial extracts of OVX ewes. The amount of 6-keto-PGF₁ produced by arteries collected from OVX-E₂ ewes was 10-fold lower than the amount of 6-keto-PGF₁ produced by arteries from intact (follicular or luteal phase) ewes. If COX-2 is responsible for increases in 6-keto-PGF₁ in uterine arteries, perhaps only small amounts of COX-2 were required for the increased production of 6-keto-PGF₁ by uterine arteries from OVX-E₂ compared to OVX-VEH ewes. The small amount of COX-2 necessary for the small increases of 6-keto-PGF₁ in OVX-E₂ uterine arteries may be below the sensitivity of the Western analysis used in this study.

NOS-III, but not NOS-II, was readily detected in uterine arteries from intact (follicular or luteal phase) and ovariectomized (E₂ or VEH) ewes after exposure to 0, 1, or 10 µg/ml LPS. Possibly NOS-III, rather than NOS-II, serves as the immune responsive NOS in sheep, at least in the case of uterine arteries. It appears that NOS-III is the primary NOS used by arteries to regulate increases in blood flow associated with the follicular compared to the luteal phase (see Note Added in Press). The selective (follicular over luteal) increase in NOS-III following LPS could be due to estrogen's enhancement of NOS-III ([38] and Fig. 7, 0 µg/ml LPS). In contrast, progesterone has been shown to inhibit NOS-II production by murine macrophages [39]. Perhaps the progesterone concentrations present during the luteal phase of ewes is inhibitory to NOS-III, thus explaining a lack of NOS-III response by luteal-phase uterine arteries exposed to LPS in this study.

In uterine arteries from follicular-phase ewes, but not luteal-phase ewes, both COX-2 and NOS-III were elevated by LPS, and the level of both of these proteins in the uterine tissue was significantly correlated with the level of 6-keto-PGF₁ in the media. Although it has been proposed that NO, the product of NOS-III, enhances COX activity [23, 24], no additive effect or enhancement of COX-2 activity, as measured by an increase in uterine artery production of 6-keto-PGF₁, was detected for follicular-phase animals. One cannot overlook the possibility that if another end product of COX-2 activity had been measured (such as PGF₂ or PGE₂), differences between follicular- and luteal-phase animals after LPS exposure would have been detected. However, the primary prostanoid produced by ovine uterine arteries is prostacyclin [40].

In uterine arteries from both follicular- and luteal-phase ewes, only uterine arteries from follicular-phase ewes responded to LPS with an increase in NOS-III. E₂ treatment of OVX ewes restored LPS-induced up-regulation of 6-keto-PGF₁ production by uterine arteries. These data support an interactive role between estrogen and LPS in production of vasoregulatory enzymes such as COX-2 and NOS-III. The interaction between estrogen and LPS may be important in insuring enhanced immune protection during the estrous period and an adequate blood supply to the feto-placental unit when the mother is challenged by an immunogen.

	NOTE ADDED IN PRESS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION NOTE ADDED IN PRESS REFERENCES

 
In the time since this manuscript was accepted for publication, the results of another study that substantiates the information presented here have also been accepted for publication. The citation is as follows: Vagnoni KE, Shaw CE, Phernetton TM, Meglin BM, Bird IM, Magness RR. Endothelial vasodilator production by uterine and systemic arteries: III. Effects of the ovarian cycle and estrogen treatment on NO synthase expression. Am J Physiol 1998; (in press).

	ACKNOWLEDGMENTS

 
The authors would like to thank Terrance Phernetton and Doug Blood for technical assistance, and David Vagnoni for statistical analysis.

	FOOTNOTES

 
¹ Research was supported by the United States Department of Agriculture grants #95372032647 and #9735044912 and National Institutes of Health grants HL49210, HD33225, and HL57653, and by the Utah Agricultural Experiment Station, Utah State University, Logan, UT. Approved as paper no. 7045.

² Correspondence. FAX: 435 797 2118; kvagnoni{at}cc.usu.edu

Accepted: June 9, 1998.

Received: April 13, 1998.

	REFERENCES

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION NOTE ADDED IN PRESS REFERENCES

 

1. Burell R. Human responses to bacterial endotoxin. Circ Shock 1994; 43:137–153.[Medline]
2. Kielian TL, Blecha F. CD14 and other recognition molecules for lipopolysaccharide: a review. Immunopharmacology 1995; 29:187–205.[CrossRef][Medline]
3. Wright SD, Kolesnick RN. Does endotoxin stimulate cells by mimicking ceramide? Immunol Today 1995; 16:297–302.[CrossRef][Medline]
4. Ulevitch RJ, Tobias PS. Receptor-dependent cell stimulation by bacterial endotoxin. Annu Rev Immunol 1995; 13:437–457.[CrossRef][Medline]
5. Hunt JS, Chen H-L, Miller L. Tumor necrosis factors: pivotal components of pregnancy? Biol Reprod 1996; 54:554–562.[Abstract]
6. Silver RM, Edwin SS, Trautman MS, Simmons DL, Branch DW, Dudley DJ, Mitchell MD. Bacterial lipopolysaccharide-mediated fetal death: production of a newly recognized form of inducible cyclooxygenase (COX-2) in murine decidua in response to lipopolysaccharide. J Clin Invest 1995; 95:725–731.
7. Schlafer DH, Yuh B, Foley GL, Elssaser TH, Sadowsky D, Nathanielsz PW. Effect of salmonella endotoxin administered to pregnant sheep at 133–142 days gestation on fetal oxygenation, maternal and fetal adrenocorticotropic hormone and cortisol, and maternal plasma tumor necrosis factor concentrations. Biol Reprod 1994; 50:1297–1302.[Abstract]
8. Wrathall AE, Wray C, Bailey J, Wells DE. Experimentally induced bacterial endotoxemia and abortion in pigs. Br Vet J 1978; 134:225–230.[Medline]
9. Fredriksson G, Kindahl H, Stabenfeldt GH. Endotoxin-induced prostaglandin mediated effects on the corpus luteum in the mare. Theriogenology 1986; 25:309–316.
10. Foley GL, Schlafer DH, Elssaser TH, Mitchell M. Endotoxemia in pregnant cow: comparison of maternal and fetal effects utilizing the chronically catheterized fetus. Theriogenology 1993; 39:739–762.
11. Smith WL. Prostaglandin biosynthesis and its compartment in vascular smooth muscle and endothelial cells. Annu Rev Physiol 1986; 48:251–262.[CrossRef][Medline]
12. Bottoms GD, Adams HR. Involvement of prostaglandins and leukotrienes in the pathogenesis of endotoxemia and sepsis. J Am Vet Med Assoc 1992; 12:1842–1848.
13. Magness RR. Endothelium-derived vasoactive substances and uterine blood vessels. In: Chaudhuri G, Ignarro LJ, McNellis D (eds.), National Institutes of Health Symposium, Seminars in Perinatology. Philadelphia: W.B. Saunders Co.; 1991: 15:68–78.
14. Janowiak MA, Magness RR, Habermehl DA, Bird IM. Pregnancy increases ovine uterine artery endothelial cyclooxygenase-1 expression. Endocrinology 1998; 139:765–771.[Abstract/Free Full Text]
15. Habib A, Creminon C, Frobert Y, Grassi J, Pradelles P, Maclouf J. Demonstration of an inducible cyclooxygenase in human endothelial cells using antibodies raised against the carboxyl-terminal region of cyclooxygenase-2. J Biol Chem 1993; 268:23448–23454.[Abstract/Free Full Text]
16. Nathan C. Nitric oxide as a secretory product of mammalian cells. FASEB J 1992; 6:3051–3064.[Abstract]
17. Sladek SM, Magness RR, Conrad KP. Nitric oxide and pregnancy. Am J Physiol 1997; 272:R441–R463.
18. Bandaletova T, Brouet I, Bartsch H, Sugimura T, Esumi H, Ohshima H. Immunohistochemical localization of an inducible form of nitric oxide in various organs of rats treated with Propionibacterium acnes and lipopolysaccharide. APMIS 1993; 101:330–336.[Medline]
19. Salvemini D, Misko TP, Masferrer JL, Seibert K, Currie MG, Needleman P. Nitric oxide activates cyclooxygenase enxymes. Pharmacology 1993; 90:7240–7244.
20. Magness RR, Shaw CE, Phernetton T, Zheng J, Bird IM. Endothelial vasodilator production by uterine and systemic arteries: II. Effects of pregnancy on nitric oxide synthase expression. Am J Physiol 1996; 272:H1730–H1740.
21. Collins P. Vascular aspects of oestrogen. Maturitas 1996; 23:217–226.[CrossRef][Medline]
22. Salvemini D, Misko TP, Masferrer JL, Seibert K, Currie MG, Needleman P. Nitric oxide activated cyclooxygenase enzymes. Proc Natl Acad Sci USA 1993; 90:7240–7244.[Abstract/Free Full Text]
23. Salvemini D, Settle SL, Masferrer JL, Seibert K, Currie MG, Needleman P. Regulation of prostaglandin production by nitric oxide; an in vivo analysis. Br J Pharmacol 1995; 114:1171–1178.[Medline]
24. Salvemini D, Seibert K, Masferrer JL, Settle SL, Misko TP, Currie MG, Needleman P. Nitric oxide and the cyclooxygenase pathway. Adv Prostaglandin Thromboxane Leukotriene Res 1995; 23:491–493.[Medline]
25. Sugimoto H, Matsuzaki Y, Yuasa K, Kaneko M, Yamada S, Hamana K, Matsuzaki S. Estradiol-17ß modifies the induction of spermidine/spermine N¹-acetyltransferase activity in the liver of lipopolysaccharide-treated mice. Liver 1992; 12:147–151.[Medline]
26. Frazier-jessen MR, Kovacs EJ. Estrogen modulation of JE/monocyte chemoattractant protein-1 mRNA expression in murine macrophages. J Immunol 1995; 154:1838–1845.[Abstract]
27. Miyagi M, Morishita M, Iwamoto Y. Effects of sex hormones on production of prostaglandin E₂ by human peripheral blood monocytes. J Periodontal 1993; 64:1075–1078.
28. Lesli CA, Dubey DP. Increased PGE₂ from human monocytes isolated in the luteal phase of the menstrual cycle. Implication for immunity? Prostaglandins 1994; 47:41–54.
29. Pueringer RJ, Hunninghake GW. Lipopolysaccharide stimulates de novo synthesis of PGH synthase in human alveolar macrophages. Am J Physiol 1992; 262:L78–L85.
30. Taraska T, Reynolds LP, Redmer DA. In vitro secretion of angiogenic activity by ovine follicles. In: Hirshfield AN (ed.), Growth Factors and the Ovary. New York: Plenum Press; 1989: 267–272.
31. Magness RR, Rosenfeld CR. Local and systemic estradiol-17ß: effects on uterine and systemic vasodilation. Am J Physiol 1989; 256:E536–E542.
32. Magness RR, Rosenfeld CR. The role of steroid hormones in the control of uterine blood flow. In: Rosenfeld CR (ed.), The Uterine Circulation. Ithaca, NY: Perinatology Press; 1989: 10:239–271.
33. Johnson JKL, Wimsatt J, Buckel SD, Dyer RD, Maddipati KR. Purification and characterization of prostaglandin H synthase-2 from sheep placental cotyledons. Arch Biochem Biophys 1995; 324:26–34.[CrossRef][Medline]
34. Magness RR, Rosenfeld CR, Carr BR. Protein kinase C in uterine and systemic arteries during the ovarian cycle and pregnancy. Am J Physiol 1991; 260:E464–E470.
35. Kielian TL, Blecha F. CD14 and other recognition molecules for lipopolysaccharide: a review. Immunopharmacology 1995; 29:187–205.
36. Haziot A, Rong G-W, Silver J, Goyert SM. Recombinant soluble CD14 mediates the activation of endothelial cells by lipopolysaccharide. J Immunol 1993; 151:1500–1507.[Abstract]
37. Charpigny G, Reinaud P, Tamby J-P, Creeminon C, Martal J, Maclouf J, Guillomont M. Expression of cyclooxygenase-1 and -2 in ovine endometrium during the estrous cycle and early pregnancy. Endocrinology 1997; 138:2163–2171.[Abstract/Free Full Text]
38. Rosenfeld CR, Cox BE, Roy T, Magness RR. Nitric oxide contributes to estrogen-induced vasodilation of the ovine uterine circulation. J Clin Invest 1996; 98:2158–2166.[Medline]
39. Miller L, Alley EW, Murphy WJ, Russell SW, Hunt JS. Progesterone inhibits inducible nitric oxide synthase gene expression and nitric oxide production in murine macrophages. J Leukocyte Biol 1996; 59:442–450.[Abstract]
40. Magness RR, Osei-Boaten K, Mitchell MD, Rosenfeld CR. In vitro prostacyclin production by ovine uterine and systemic arteries: effect of angiotensin II. J Clin Invest 1985; 76:2206–2212.