INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

UTERINE BLOOD FLOW (UBF) is elevated during the follicular phase of the ovarian cycle and pregnancy, which are both high estrogen states (9-11, 14, 17, 23, 37). Unlike the follicular phase, which shows dramatic elevations in the estrogen-to-progesterone ratio (14, 23, 37), pregnancy also is characterized by elevations in progesterone (17, 23). Thus alterations in UBF observed during the ovarian cycle are temporally and directly associated with the endogenous ratio of estrogen to progesterone in the systemic blood (9, 17, 23, 37) and locally in the uterine lymph (18, 19) and uterine tissues (31, 40). During menopause, there is a loss in the endogenous secretion of ovarian hormones and a corresponding reduction in endometrial and myometrial endothelial nitric oxide synthase (eNOS) expression (16).

Nitric oxide (NO) metabolite (NO_x) levels in the systemic circulation and/or urine have been reported to be elevated during pregnancy in the rat (6), sheep (44, 45), and the human (4, 30). In women, however, this latter observation has recently been questioned in patients with dietary control before sampling (7). With regard to plasma levels of NO_x, although elevations have been reported in pregnant vs. nonpregnant sheep (44, 45), comparisons throughout the third trimester when plasma cGMP (the primary second messenger of NO) levels are elevated (25, 33) have not been made. It is also unknown if the number of fetuses present in the uterus affect the levels of NO_x in plasma.

The potent local vasodilator NO is produced primarily by the endothelium of both uterine and systemic arteries (24, 25, 36). NO production is increased in uterine vessels but not omental or renal vessels during pregnancy (24, 29, 41-43). We and others (3, 25, 42, 43) have recently demonstrated in uterine arteries that both message and/or protein expression of eNOS, the rate-limiting enzyme for the production of NO, also exhibit increases in their levels during pregnancy compared with nonpregnant ewes. In addition, we noted that follicular-phase uterine artery (UA) eNOS levels were greater than luteal-phase protein levels (38). Moreover, infusion of 17-estradiol into ovariectomized ewes increases eNOS protein expression and eNOS activity in uterine but not systemic (omental and/or renal) arteries (34a, 35, 38, 39). It is unknown, however, if either the follicular-phase increase in vascular eNOS protein expression is paralleled by elevations in mRNA or if the magnitude of eNOS mRNA changes is greater in pregnant vs. follicular-phase sheep. Moreover, in the surgical menopause model of ovariectomy (20-22, 34a, 38), it is possible that both the NOS protein and mRNA expression levels will be decreased compared with intact cycling animals.

We therefore hypothesized that increases in NO production by uterine arteries during high estrogen and UBF states such as pregnancy or the follicular phase of the ovarian cycle are due in turn to elevations in eNOS expression at the level of mRNA and protein and conversely that there will be a decrease in eNOS during surgical menopause. The specific objectives tested in the current study were to determine 1) the effects of ovariectomy, the ovarian cycle (follicular vs. luteal phases), and pregnancy on eNOS mRNA vs. protein expression in UA endothelium; 2) if any changes in eNOS protein expression in UA endothelium during these distinct physiological studies are specific to this vascular bed or are they also seen with another systemic artery, i.e., omental arteries; and 3) the corresponding effect of ovariectomy, the ovarian cycle, and pregnancy on circulating NO_x levels during the third trimester and further whether the number of fetuses has any effect on NO_x levels in circulating plasma from late-gestation sheep.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Tissue collection procedure. Uterine and omental (systemic) arteries (3-5 mm diameter) were obtained from mixed Western breed nonpregnant sheep (n = 8) and pregnant ewes at 119-130 days gestation (n = 8). The ewes were euthanized with intravenous pentobarbital sodium (50-70 mg/kg) according to the recommendations of the "Report of the American Veterinary Medicine Association Panel on Euthanasia" (1). The procedures for euthanasia 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 Sciences. The uterine and omental arteries were rinsed and dissected free of extraneous tissue in PBS and were then placed directly in lysis buffer (50 mM Tris, 0.15 M NaCl, and 10 mM EDTA, pH 7.4, plus the addition of 0.1% Tween 20, 0.1% -mercaptoethanol, 0.1 M phenylmethylsulfonylamide, 5 µg/ml leupeptin, and 5 µg/ml aprotinin; all from Sigma Chemical, St. Louis, MO). The arteries were then further dissected to remove fat and connective tissue at 25°C in petri dishes with PBS. The PBS was removed and replaced from the dish three times during the dissection. The methods for the preparation of the endothelium-denuded [vascular smooth muscle (VSM)] uterine or omental segments, or direct isolation of endothelial protein for SDS-PAGE electrophoresis and Western immunoblotting, were previously described (3, 25, 38).

Synchronization of follicular and luteal-phase ewes. Ewes (50-60 kg) were observed daily for signs of behavioral estrus using a vasectomized ram. Ewes exhibiting normal estrous cycles (16-18 days) were given two intramuscular injections of 5 mg PGF₂ (Lutalyse; Upjohn, Kalamazoo, MI) 4 h apart to induce luteolysis, and the ewes were then monitored for behavioral estrus beginning 48 h after the first PGF₂ injection. At estrus (day 0), ewes were randomly paired and assigned to the following two groups: follicular (day 1 to 0, n = 6) or luteal (day 10, n = 6). On day 8 postestrus for each pair, ewes assigned to the follicular group were given PGF₂ as described above and the ewes assigned to the luteal group were given two intramuscular saline injections, in an equivalent volume as PGF₂, 4 h apart. This synchronization protocol results in ewes showing estrus within ~44-56 h after the first PGF₂ injection (12, 14, 23, 38). We previously have reported the estrogen and progesterone levels using this model (23). Moreover, luteal blood flow and progesterone fell precipitously to basal levels within 6 h, and estrogen levels rose progressively between 18 and 48 h after injection of PGF₂ (23, 37 and R. R. Magness, M. C. Wiltbank, and T. M. Phernetton, unpublished observation). Follicular-phase ewes were euthanized 44 h after the first injection of PGF₂, and the paired luteal-phase ewe was euthanized on the same day, also at ~44-46 h after the first injection of saline.

Ovarian structures. The presence and size of the ovarian structures, i.e., progesterone-producing corpus lutea and estrogen-producing follicles, were determined as previously described (12, 23, 38). Luteal-phase ewes had significantly more vascularized corpora lutea than the follicular-phase ewes (P < 0.01). The follicular-phase corpora lutea were blanched and very avascular. The follicular-phase ewes also demonstrated more large (>6 mm) follicles, indicating that the follicular-phase sheep were in a high estrogen condition while the luteal-phase sheep were in a high progesterone state, as previously reported (9, 12, 14, 23, 26, 38).

Ovariectomized ewes. Ewes (50-60 kg; n = 6) exhibiting normal estrous cycles underwent ovariectomy via midventral laparotomy under general anesthesia as previously described (20-22, 34a, 38). Sheep were euthanized (1) on days 10-11 after ovariectomy (8, 12, 26). Uterine and omental artery endothelial-isolated protein and VSM samples from these ovariectomized ewes were collected and prepared as described above.

Isolation of endothelial cells and cellular RNA. UA endothelial cells were isolated by collagenase dispersion as previously described (3, 12, 13) from each of the six ovariectomized, follicular, and luteal sheep and eight pregnant sheep. These freshly isolated endothelial cells were solubilized in RNazol B (1 ml), and 150 µl chloroform were added to promote phase separation followed by centrifugation (12,000 g, 20 min). The upper aqueous phase was then removed, extracted two times with phenol-chloroform-isoamyl alcohol using heavy-grade phase lock gel (5-Prime,3-Prime, Boulder, CO), and finally mixed with 110% by volume of isopropanol. RNA was precipitated by standing at 20°C for 1 h before recovery by centrifugation (12,000 g, 30 min) and washing of the pellet in 75% ethanol. RNA was then solubilized in molecular biology-grade water (5-Prime,3-Prime) and quantified by spectrophotometry before storage at 70°C.

eNOS RT-PCR assay. eNOS mRNA levels were quantified by coupled RT-PCR amplification in single-tube assays using avian myeloblastosis virus reverse transcriptase and Taq polymerase with 0.1 µg total RNA essentially as described previously (3). All data in each group were normalized to the mean glyceraldehyde-3-phosphate dehydrogenase content of each sample within each subgroup, determined by the same RT-PCR procedure as previously described (3). The copy number of transcripts was calculated from a standard curve of 10⁴ to 10¹⁰ copies of eNOS cDNA plasmid run in each assay.

Western immunoblot analysis. The protocol for SDS-PAGE on 7.5% gels and subsequent Western Immunoblot analysis have been described previously (3, 8, 12, 13, 25, 26). Briefly, endothelial-isolated protein and VSM samples from uterine and omental arteries were solubilized in lysis buffer by homogenization and/or sonication. Solubilized protein was quantified using a modified Lowry assay procedure (Bio-Rad, Hercules, CA). Proteins (20 µg/lane) were then separated by size on 7.5% polyacrylamide gels (100 volts, 2.5 h, MiniProtean II; Bio-Rad) before transfer to an Immobilon P membrane (100 volts, 2 h). The Immobilon P membrane was then probed for eNOS using the enhanced chemiluminescence reagent detection system, as described by Amersham, and exposed to hyperfilm (1 min). The eNOS specific monoclonal antiserum (Transduction Laboratories) was used at a dilution of 1:750, and second antibody (sheep anti-mouse Fab₂/horseradish peroxidase conjugate; Amersham, Arlington Heights, IL) was used at 1:2,000 dilution. Levels of eNOS were then quantified by scanning densitometry (Bio-Rad 670 scanning densitometer) and expressed relative to mean nonpregnant luteal absorbance.

NO_x measurements. Plasma samples were obtained from the jugular venous catheters in ovariectomized (n = 8), follicular-phase (n = 7), luteal-phase (n = 6), and pregnant (n = 33) sheep. The samples obtained from pregnant sheep were at 110 (n = 4), 120 (n = 12), and 130 (n = 17) days gestation to span the time of the third trimester when the UA endothelial samples were obtained as described above. In addition, simultaneous venous and arterial plasma samples were obtained from eight additional chronically instrumented sheep (120-130 days). NO levels were measured as total NO₂-NO₃ (NO_x) using a Sievers model 280 NO analyzer as we have described previously (3, 42, 46).

Statistical analysis. Data were analyzed by Student's t-test or by ANOVA followed by Duncan's post hoc test as appropriate. Data are presented as means ± SE.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Effect of ovariectomy, the ovarian cycle, and pregnancy on eNOS expression. Figure 1 shows the levels of eNOS protein in the three nonpregnant groups compared with pregnancy in uterine and omental artery endothelial-isolated protein. As we have previously reported (25) and as was recently confirmed (42, 43), there was a much stronger eNOS protein signal (4.1- to 6.9-fold) observed in UA but not omental artery endothelial-isolated proteins from pregnant vs. nonpregnant luteal sheep. We also confirm previous reports (25, 38) that the expression of eNOS was quite low in VSM (data not shown). Therefore, in the present study, we report the protein expression data only for the endothelial isolated protein preparations. Compared with ovary-intact, luteal-phase sheep, ovariectomy did not alter eNOS protein in either UA or omental artery endothelial-isolated protein (P > 0.05). Compared with luteal-phase animals, we observed that both UA and omental artery protein appeared to increase only slightly during the follicular phase; however, this did not reach statistical significance by ANOVA. The levels of eNOS protein were significantly different (P = 0.05) in uterine and omental artery when data from the follicular-phase group were compared with the combined values for both of the other two nonpregnant groups, which were similar. In Fig. 2 we show the effects of ovariectomy, the ovarian cycle, and pregnancy on UA eNOS mRNA expression. Data are expressed as copy number determined from the RT-PCR reaction and are based on known copies of eNOS cDNA per microgram total RNA. The UA level of mRNA was decreased in ovariectomized ewes while the follicular phase increased eNOS mRNA in UA endothelium relative to the luteal phase. The most dramatic increase in eNOS mRNA, however, was seen during pregnancy (2.4- to 6.9-fold).

View larger version (18K): [in this window] [in a new window]   Fig. 1.   Effects of ovariectomy (OVEX), the ovarian cycle [follicular (FOL) vs. luteal (LUT) phase], and pregnancy (PREG) on endothelial nitric oxide synthase (eNOS) protein expression as detected by Western Immunoblot analysis in uterine and omental artery endothelial-isolated protein (Endo; 10 µg/lane). Comparison of arbitrary absorbance units (AU) of eNOS expression in uterine artery Endo (filled bars) and omental artery Endo (open bars) from OVEX, FOL, LUT, and PREG sheep. Values are means ± SE; n = 6 nonpregnant ewes/treatment group and n = 8 for pregnant sheep. *P < 0.01, pregnant > all nonpregnant groups. Note that FOL > LUT = OVEX combined (P = 0.05).

View larger version (12K): [in this window] [in a new window]   Fig. 2.   Effect of OVEX, the ovarian cycle (FOL vs. LUT phase), and PREG on eNOS mRNA expression as detected by RT-PCR followed by Southern blot analysis in uterine artery endothelium isolated by collagenase disbursement. Values are comparisons of eNOS transcripts (copies/µg total RNA). Values are means ± SE; n = 6 ewes/nonpregnant group and n = 8 for pregnant sheep. Bars with different superscripts are significantly different. P = 0.01, OVEX < LUT < FOL < PREG.

Effects of the ovarian cycle and pregnancy on plasma NO_x levels. We also evaluated the effects of ovariectomy, the ovarian cycle, and pregnancy on circulating NO_x levels measured as total NO₂ and NO₃ (Fig. 3). Neither ovariectomy nor stage of the ovarian cycle altered NO_x levels in systemic plasma. However, pregnancy substantially elevated (P < 0.01) NO_x levels of plasma by nearly sixfold over luteal values. As shown in Fig. 4, there were also effects of both gestational age during the third trimester and fetal numbers (single, twin, triplet) on circulating plasma NO_x levels. We observed that there was a progressive increase from 110 to 130 days in NO_x levels. These days were chosen to span the days of gestation for which we evaluated UA eNOS expression, and 8 of the 33 plasma samples were from the animals used in the eNOS expression studies. In addition, regardless of gestational age, the singleton pregnancies (n = 13) had less NO_x plasma levels than did the twins (n = 15) and triplets (n = 5), which were statistically equivalent (Fig. 4). In eight additional chronically catheterized late-gestation sheep (110-120 days) in which uterine venous and arterial samples were simultaneously obtained, no significant venous-arterial concentration differences (2 ± 0.7 µM) were observed with uterine venous and arterial concentrations averaging 14 ± 1.5 and 15 ± 2.1 µM, respectively.

View larger version (10K): [in this window] [in a new window]   Fig. 3.   Effects of OVEX (n = 8), the ovarian cycle [FOL (n = 7) vs. LUT (n = 6) phase], and PREG (n = 33) on circulating nitric oxide (NO) levels measured as total NO2-NO3 (NOx). Values are means ± SE; bars with different letter superscripts are significantly different. P < 0.01, pregnant > all nonpregnant groups, which were equal.

View larger version (11K): [in this window] [in a new window]   Fig. 4.   Effect of gestational age (110-130 day's gestation; left) and fetal numbers [singles (n = 13), twins (n = 15), triplets (n = 5); right] on NO levels in circulating plasma measured as total NOx. Values are means ± SE; bars with different superscripts are significantly different, P < 0.01. Left: effects of the third trimester on circulating NOx. NOx levels in plasma increased progressively from 110 to 130. Right: effect of fetal number; with greater fetal numbers there was an increase in NOx in circulating plasma.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We demonstrate for the first time that eNOS mRNA levels are increased in UA endothelium during the follicular vs. luteal phase of the ovarian cycle, when the endogenous estrogen-to-progesterone ratio and UBF are elevated (9-11, 14, 17-19, 23, 37, 38). This observation is consistent with reports showing that eNOS protein (16) and endothelium-dependent NO-mediated ACh relaxation responses (2) in human uterine arteries are elevated during the menstrual cycle phases when estrogen levels are elevated. In contrast to our previous data (38), follicular- vs. luteal-phase increases in UA eNOS protein were very small and equivalent to the omental artery responses but were reflected by significant increases in mRNA levels in the follicular-phase and decreases in the ovariectomized ewes. This dissociation between levels of eNOS message and protein may reflect either translational control or, more likely, changes in mRNA vs. protein stability. Alternatively, this effect may also reflect the portion of the vascular tree we used in the current study (1° and 2°) to sample enough of the UA endothelium to obtain both protein and RNA. In recent preliminary observations, we have noted that the relative follicular-phase rise in UA eNOS protein is much more substantial in 2° and 3° rather than primary vessels (15).

In ovariectomized sheep the eNOS expression in uterine and omental artery endothelium was lower than that in the follicular-phase group. These data therefore suggest that estrogen or other ovarian factors are important in maintaining constitutive levels of eNOS protein expression. We recently reported that cyclooxygenase-1 mRNA, another UA endothelial gene, was reduced in UA endothelium of ovariectomized vs. luteal- or follicular-phase sheep (12). Because estrogen administration restores eNOS protein expression in uterine but not renal artery endothelium (34a, 38), we believe that estrogen plays a primary role in the maintenance of eNOS mRNA and protein levels, although combined estrogen plus progesterone therapy further stimulates these increases (34a).

Uteroplacental blood flow increases 30- to 50-fold during pregnancy to maintain normal oxygen and nutrient delivery for optimal fetal growth and development (see reviews in Refs. 17, 27, 32, 36). Angiogenesis/vasculogenesis during early to midgestation also contributes to increased UBF (14, 17, 27, 32, 47). However, because the last third of gestation is the period of greatest absolute increase in uterine perfusion, occurring after completion of new vessel growth, this indicates that maintenance of vasodilation in newly developed vessels is crucial for increases in UBF. Recent studies have shown that local uterine arterial inhibition of NOS activity using L-NAME given to late pregnant sheep decreases the uterine venous NO second message, cGMP levels (33), and UBF (28).

We confirmed that pregnancy substantially increases in vivo steady-state eNOS mRNA and protein levels in UA endothelium (3, 25, 29, 36, 42, 43), consistent with observations that eNOS specific activity is greater in endothelium-intact uterine arteries from pregnant vs. nonpregnant guinea pigs (41), sheep (24), and women (29). Endothelium- and NO-dependent cGMP production and NO_x production were also elevated in uterine arteries by pregnancy (24, 29, 42, 43). Functional studies using isolated vessels have shown decreased responses to vasoconstrictor agents (e.g., norepinephrine) in intact (not denuded) uterine arteries from pregnant vs. nonpregnant states via an NO-mediated mechanism (36, 42, 43). The endothelium, not VSM, contains the vast majority of the NO synthase activity and eNOS levels in the uterine vascular wall (3, 24, 25, 29, 42; Fig. 1). Thus these pregnancy-related responses are cell specific but are also unique to the uterine vasculature since systemic artery endothelium showed only minimal or no significant changes in either NO synthase activity or eNOS protein expression.

Neither ovariectomy nor the ovarian cycle significantly affected circulating NO_x. In contrast, in women follicular (proliferative)-phase levels of NO_x were elevated (5, 34). This may be a species difference, or the latter studies may not have properly controlled the patient's diet (7). It is also possible that local minor changes in UA eNOS expression with ovariectomy and the ovarian cycle could not manifest a significant rise in total systemic NO_x levels. Alternatively, because pregnancy is associated with dramatic increases in receptor coupling via the extracellular signal-regulated kinase 1/2 signaling pathway in UA endothelial cells (3), our negative systemic NO_x data may be partly explained by a lack of such an adaptive signaling response during these three nonpregnant physiological states.

We observed nearly sixfold elevations in systemic plasma NO_x levels in pregnancy as reported by others studying rats (6) and sheep (44, 45). In the sheep studies, blood NO_x levels were evaluated in pregnant sheep at 110-124 (44) and 140-143 days gestation (45); however, gestational differences or the presence of multiple fetuses was not commented on. The third trimester is physiologically very dynamic, and elevations in uteroplacental blood flows are associated with substitutional increases in the metabolic demands of fetal growth (see reviews in Refs. 9, 17, 27, 36). We observed that systemic plasma NO_x levels increased progressively from 110 to 130 days gestation (0.7-0.9 gestation) when fetal growth velocity is greatest and that multiple vs. singleton fetal gestations had greater increases in plasma NO_x levels. Greater fetal number will substantially increase total placental and uterine weight, and thus the total mass of the fetoplacental and uteroplacental vasculature will be increased because of angiogenic changes (17, 27, 32, 47) that could contribute to the elevations in systemic plasma levels of NO_x. A placental source of NO secreted across the fetoplacental to the maternal circulation is possible, since the ovine placenta expresses eNOS and produces NO, and they both rise during the third trimester (46). However, we and others (44) did not observe uterine venous-arterial concentration differences, possibly because of the very long half-life of NO_x metabolites (5 h), which would abrogate any small venous-arterial NO_x differences. Previously, we (25) did not observe a change in UA eNOS levels throughout the third trimester, regardless of fetal numbers, and so herein suggest that there must be an increase in the activation of eNOS protein due to changes in signaling (3). This is supported by Xiao et al. (42, 43) who showed that NO_x production was higher in pregnant than in nonpregnant perfused uterine arteries under both basal, but more so under stimulated (A23187 and ATP), conditions and that inhibition of eNOS or endothelium removal increased norepinephrine-induced contractions of uterine arteries from pregnant but not nonpregnant sheep.

In summary, we have shown that elevations of eNOS mRNA and/or protein are seen during the follicular phase and pregnancy but that changes in circulating NO_x are only seen in pregnancy. In our recent studies of UA endothelial cells, we demonstrated that pregnancy specifically increased growth factor receptor and heptahelical receptor signaling, which substantially impacted NO production (3). Thus pregnancy-specific increases in NO_x levels reported herein likely reflect both increased eNOS protein, controlled at the level of eNOS mRNA, and receptor coupling/signaling to cause activation of eNOS.