INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

UTERINE ARTERY (UA), but not systemic (renal) artery (RA) endothelial, endothelial nitric oxide (NO) synthase (eNOS) protein (29, 40) and mRNA (29) are elevated during the follicular phase of the ovarian cycle, in which the endogenous estrogen-to-progesterone ratio and uterine blood flow (UBF) are elevated (9, 11, 26). A low basal UBF level and a low estrogen-to-progesterone ratio characterize the luteal phase. Similar increases in eNOS protein levels were noted during the follicular phase of the menstrual cycle in women (31). Furthermore, these alterations in eNOS levels in the UA endothelium can be mimicked by exogenous administration of 17-estradiol and/or progesterone (35, 36, 40).

NO (27-29, 37, 47) and its physiological second messenger cGMP (17, 27, 37) are increased during pregnancy. UA, but not systemic artery endothelial, eNOS protein (28, 29, 31, 47) and mRNA (1, 29) expression and activity (27, 31, 46) also are elevated by pregnancy. UA eNOS protein levels are unchanged during late gestation, from day 110 to 142 of ovine pregnancy (28), although nitrite or nitrate (NO_x) levels increase with advancing gestational ages during the third trimester (29). The timing of either the gestational increase in UA eNOS expression or the circulating NO_x levels during the second trimester of pregnancy is not known.

Acute (7, 32, 39) and chronic (13) increases in shear stress upregulate eNOS expression. Furthermore, arteries respond to chronic increases in blood flow by increasing luminal diameter, a process that results in reduction of wall shear stress toward normal (18, 19, 38, 48). Because UBF is high during the follicular phase and pregnancy, this likely results in an acute and chronic increase in shear stress, respectively. However, during gestation, artery diameter is also substantially increased (12), which serves, in part, to normalize shear forces. It is not known if eNOS is differentially regulated by flow-mediated changes in shear stress in the follicular phase versus pregnancy, which are physiological states characterized by acute versus chronic rises in UBF.

There is evidence suggesting that heterogeneity of function exists within different sizes of vessels. A gradient of progesterone concentrations (34, 43, 44) and estrogen receptor densities (25) have also been observed in the uterine and aortic vasculatures, respectively. These steroid hormones are known to regulate UA but not systemic arterial endothelial eNOS expression (35, 36, 40). It is unknown whether the steroid hormone and/or receptor gradients regulate eNOS differently in different sized UA. Finally, the tunica media of arteries consists almost entirely of vascular smooth muscle (VSM), which decreases with reductions in vessel diameter, and therefore, as a result, may be differentially regulated by the endothelium or vice versa in a paracrine fashion (for a review, see Ref. 24).

In the present study, we investigated the hypothesis that regional differences in responses of eNOS protein expression are observed with both decreases in vessel diameter and during the ovarian cycle and pregnancy, because it is expected that shear stress-mediated mechanisms are regionally altered. The specific objectives of this study were as follows: 1) to evaluate regional expression of eNOS in the uterine circulation; 2) to compare any such eNOS expression gradients in luteal versus follicular phase ewes and in second versus third trimester pregnant ewes; and 3) to determine if earlier observations that UA endothelial eNOS levels are upregulated during the follicular phase and pregnancy are seen throughout the uterine vascular tree and how this relates to circulating NO_x levels.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Synchronization of follicular and luteal phase ewes. As described previously (26, 29, 40), at estrus (day 0), mixed Western breed ewes were randomly paired into two groups, follicular (day 1 to 0, n = 6) or luteal (day 10, n = 6). Luteolysis was induced on day 8 postestrus by two injections (7.5 mg im, 4 h apart) of PGF₂ (Lutalyse, Upjohn; Kalamazoo, MI) to produce follicular phase ewes, and ewes assigned to the luteal phase group were given equivalent volume (1.5 ml) injections of saline 4 h apart. Follicular phase ewes were euthanized 44 h after the first injection of PGF₂, and the paired luteal phase ewe was also euthanized on the same day, ~44-46 h after the first injection of saline. Follicular phase animals had regressing blanched, avascular corpora lutea (5.5 ± 0.7 mm) and large preovulatory follicles (>6 mm). The luteal phase sheep had larger, highly vascularized corpora lutea (8.5 ± 0.9 mm).

Procurement of tissues. Procedures for animal handling and protocols for experimental procedures were approved by the University of Wisconsin-Madison Research and Animal Care and Use Committees of both the Medical School and the College of Agriculture and Life Sciences and follow the recommendations of the "Report of the American Veterinary Medical Association Panel on Euthanasia." Nonpregnant (follicular, n = 6; luteal, n = 6) and pregnant (midpregnant, n = 6, 82 ± 1 days gestation; late pregnant, n = 6, 127 ± 3 days gestation; term = 145 ± 4 days gestation) ewes were euthanized with pentobarbital sodium (50-70 mg/kg). The fetal weights, crown rump lengths, and body mass index of the midpregnant sheep (0.403 ± 0.011 kg, 25.41 ± 0.29 cm, and 6.24 ± 0.12 kg/m², respectively) were less (P < 0.01) than those of the late pregnant ewes (2.23 ± 0.18 kg, 41.35 ± 1.20 cm, and 12.60 ± 0.41 kg/m², respectively). RA, internal iliac arteries (II), and UA [primary (UA 1°), secondary (UA 2°), and tertiary (UA 3°) branches] were dissected free of connective tissue, fat, and veins and then rinsed free of blood in phosphate-buffered saline (8 mM sodium phosphate, 2 mM potassium phosphate, and 0.15 M NaCl, pH 7.4; Sigma; St. Louis, MO). The external diameter (in mm) of each artery was measured at multiple sites (8-12) along the length of the artery using a micrometer. The arteries were then opened longitudinally, and the endothelium/tunica intima was gently scraped (4-6 times) from the artery using a curved-end spatula and 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 phenylmethylsulfonyl fluoride, 5 µg/ml leupeptin, and 5 µg/ml aprotinin; all from Sigma). We (28) have previously reported and validated this method of isolating endothelial protein to be relatively free of VSM contamination. The endothelial isolated proteins were snap frozen immediately upon collection in liquid nitrogen and stored at 20°C.

Western blot analysis of endothelial isolated proteins. The protocol for SDS-PAGE on 7.5% gels and Western immunoblot analysis has been described previously (28, 29, 35, 40). Endothelial isolated proteins (10 µg/lane) were separated by size on polyacrylamide gels (100 V, 1.5 h, MiniProtean II, Bio-Rad) before transfer to Immobilon P membranes (100 V, 2 h). The blots were designed to include either luteal versus follicular or midpregnant versus late pregnant comparisons. Each blot included a human umbilical vein endothelial cell (HUVEC) lysate as a positive control/standard and Rainbow molecular weight markers (Amersham; Arlington Heights, IL). The membrane eNOS was visualized using enhanced chemiluminescence as described by Amersham and exposure to Hyperfilm for 15 min. The primary mouse anti-eNOS monoclonal antibody (Transduction Laboratories; Lexington, KY; 1:750 dilution) was directed against sheep anti-mouse serum (Amersham; 1:3,000 dilution). A positive band for samples corresponding to the HUVEC standard was detected at ~135-140 kDa, confirming the presence of eNOS. Levels were quantified by using a Bio-Rad scanning transmission densitometer (model 670) coupled with Bio-Rad Molecular Analyst software. Factored arbitrary units were used to account for the blot-to-blot variability and to normalize the data. Data are presented as the total sum of the arbitrary units of each blot multiplied by each individual arbitrary unit and then divided by their representative HUVEC standard.

NO_x measurements. Plasma samples were obtained from the jugular vein just before euthanization from the sheep in the four treatment groups. NO_x levels were measured as total NO₂-NO₃ using a Sievers NO analyzer (model 280) as we previously described (1, 29, 47).

Statistical analysis. When necessary, data were log₁₀ transformed to meet assumptions of normal distribution and equal variance. Data were analyzed by two-way ANOVA followed by Bonferroni's multiple-comparison tests with both "vascular generation" and "treatment group" being highly significant; however, the interaction of these two variables was marginal. Therefore, further differences among physiological treatment groups within regional segments of the uterine circulation were analyzed by Student's t-test or one-way ANOVA (SigmaStat; Jandel Scientific; San Rafael, CA) as appropriate. Data are presented as means ± SE, and significance was accepted at P < 0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Regional changes in vessel diameter in cycling and pregnant ewes. We examined the effects of the ovarian cycle and pregnancy status on the vessel external diameter measured at multiple sites (8-12) along the length of each artery (Table 1). In the four reproductive treatment groups, the external diameter decreased progressively from the II to UA 3° (P < 0.001). The RA external diameter did not differ from that of the II in either the nonpregnant state or during late pregnancy. Each branch of the UA collected from luteal phase sheep had consistently larger diameter than those from follicular phase sheep (P < 0.02). As expected, the external diameter increased significantly during pregnancy in each branch of the UA compared with the nonpregnant sheep (P < 0.001). In contrast, the diameter of the RA was unaltered by pregnancy. No clear pattern was observed for the diameter of the II in nonpregnant versus pregnant sheep. There were no differences in external diameter in mid- versus late pregnant ewes except for UA 1° (P < 0.05), which was slightly larger in late pregnancy.

Regional changes in eNOS protein expression in UA endothelial isolated protein from cycling and pregnant ewes. In Fig. 1, a representative Western immunoblot of UA 1°, UA 2°, and UA 3° endothelial isolated proteins for each reproductive treatment group is shown. We observed that luteal and late pregnant, but not follicular or midpregnant, ewes exhibited a regional, decreasing expression gradient of eNOS. To further quantify regional changes of eNOS protein expression, densitometry was performed. In luteal phase ewes, RA, II, and UA 1° eNOS levels did not differ; however, eNOS protein expression was decreased by a total of 60.7 ± 9.8% in UA 2° and UA 3° compared with the II and UA 1° (Fig. 2) (P < 0.05). In contrast to the luteal phase, no regional eNOS protein expression gradient was observed in either the follicular phase or midpregnant sheep. However, in late pregnancy when comparing UA 1°, UA 2°, and UA 3°, there also appeared to be a regional expression gradient, which is further supported by the observation that UA 2° and UA 3° eNOS protein expression decreased significantly (51.2 ± 12.6%) compared with II (P < 0.05) expression, which was similar to those of RA and UA 1°.

View larger version (33K): [in this window] [in a new window]   Fig. 1.   Representative Western immunoblots demonstrating regional expression of endothelial nitric oxide (NO) synthase (eNOS) in the uterine vasculature for each reproductive treatment group. The summary data (means ± SE) are also expressed [in densitometric arbitrary units (AU)] for each reproductive treatment group in Fig. 2. UA 1°, primary branch of the uterine artery (UA); UA 2°, secondary branch of the UA; UA 3°, tertiary branch of the UA.

View larger version (33K): [in this window] [in a new window]   Fig. 2.   Western immunoblot analysis comparing eNOS protein levels from luteal (LUT), follicular (FOL), midpregnant (MP), and late pregnant (LP) ewes. Endothelial isolated proteins (10 µg/lane) from the renal artery (RA), internal iliac artery (II), and branches of the UA are shown. Densitometric quantification was performed, and data are reported as factored AU divided by a human umbilical vein endothelial cell (HUVEC) standard. Data are means ± SE. Significant differences are shown by different letters within an artery type/generational branch (P < 0.05).

Effects of the ovarian cycle and pregnancy on plasma NO_x levels. NO_x levels were unaltered by the ovarian cycle (Fig. 3); however, the levels of plasma NO_x in late gestation sheep were substantially higher (P < 0.02) than either luteal (4.45-fold), follicular (4.47-fold), or midpregnant (3.00-fold) sheep, which were similar.

View larger version (9K): [in this window] [in a new window]   Fig. 3.   Effects of the ovarian cycle [LUT (n = 4) vs. FOL (n = 3) phase] and pregnancy [MP (n = 6) vs. LP (n = 6)] on circulating NO levels measured as total nitrite and nitrate (NOx). Values are means ± SE. Significant differences are shown by different letters. P < 0.02, LP > nonpregnant and MP, which were equal.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Our results demonstrate, for the first time, a regional decreasing eNOS protein expression gradient in the uterine circulation during the luteal phase and in late pregnant sheep. In the follicular phase during midpregnancy, eNOS expression is upregulated in smaller diameter UA. Herein, we also report for the first time that UA eNOS protein levels from mid- versus late pregnant ewes exhibited higher eNOS expression, whereas plasma NO_x levels were only significantly elevated in late gestation sheep, which demonstrates that differences in eNOS expression and NO production are not temporally synchronized and suggests that rises in eNOS levels proceed the activation of this enzyme.

In the present study, eNOS expression was upregulated in the smaller diameter UA during the follicular phase and midpregnancy, physiological states associated with elevated UBF and estrogen levels (9, 11, 26). Unexpectedly, we also observed higher eNOS expression in the midpregnant state compared with late gestation. Others have reported that pregnancy is characterized by upregulated UA endothelial eNOS expression (28, 29, 47) and specific activity (27, 31, 46). Estrogen also upregulates UA endothelial eNOS expression (35, 36, 40) and activity (41). Our previous report (28) of a lack of change in UA eNOS protein levels from day 110 to 142 of ovine pregnancy demonstrates eNOS actually is elevated in the second trimester and must fall somewhat during the third trimester. These observations also further underscore the important finding that there must be an increase in activation of eNOS protein due to changes in cell signaling (1, 5, 29) because plasma NO_x levels are elevated in late gestation compared with midpregnant and nonpregnant states. In support of this observation, we observed that the highest eNOS protein expression occurs in midgestation, without any change in NO_x levels compared with the nonpregnant state. Thus elevated estrogen in the follicular phase and estrogen plus progesterone during pregnancy may regulate eNOS expression, but other factors such as rising shear stress with elevations in UBF must also modulate the activation of this enzyme.

A decreasing eNOS expression gradient in the UA was observed especially in the luteal phase. We initially hypothesized that regional differences in shear stress may account for the observed eNOS gradient noted in the current study. It is noteworthy that there are no in vivo calculations of shear stress for the uterine vascular bed in the literature. Therefore, we estimated the physiological UA shear stress mathematically, where shear stress () was calculated in dynes per square centimeter and calculated from the Hagen-Poiseuille formula as follows:  = 4µQ/r³, where µ is the viscosity of blood (in Poise), Q is blood flow through the vessel lumen (in ml/s), and r is one-half of the externally measured diameter (in cm; Table 1). UBF and renal blood flow were derived experimentally from our laboratory using radioactive microspheres and Transonic flow probes in the nonpregnant sheep (luteal and follicular) and late pregnant sheep (6-8 animals/group), accounting for the number of branches present in the primary, secondary, and tertiary uterine circulation. Unlike the uterine vascular bed, ovine renal blood flows remain relatively stable during the ovarian cycle and late pregnancy and provide a good control within the same animal for comparison purposes. Shear stress in the renal vascular bed remained unaltered during these physiological states, averaging 16.0, 19.7, and 18.0 dyn/cm² in the luteal, follicular, and late pregnant sheep, respectively. Shear stress increased throughout the uterine vascular bed (UA 1°  UA 2°  UA 3°) in all three physiological states. During the luteal phase, the basal levels of shear stress averaged 2.5, 2.8, and 4.2 dyn/cm² for UA 1°, UA 2°, and UA 3°, respectively. These levels increase during the follicular phase to 14.0, 19.5, and 29.2 dyn/cm²; however, the greatest increases in shear stress were estimated to occur in the late pregnant ewe: 59.7, 77.9, and 101.0 dyn/cm² for UA 1°, UA 2°, and UA 3°, respectively. Thus because the level of shear stress increased in the smaller diameter UA relative to the primary UA, whereas eNOS levels were decreased, shear stress cannot be responsible for this eNOS gradient. The reason for this expression gradient in the luteal phase sheep remains unknown but may relate to local hormone gradients within the uterine tissues. Pope et al. (34) and Weems et al. (43, 44) demonstrated in unilaterally ovulating sheep and cattle that the uterine tissue and artery on the side ipsilateral to the corpus luteum contained greater quantities of progesterone than those more distal. The ability of progesterone to be concentrated in the uterine tissues occurs via a countercurrent exchange of steroid between the ovarian vein to the ovarian artery (42). The importance of progesterone in the regulation of UA eNOS was recently demonstrated in ovariectomized, exogenous steroid-treated ewes (35) as well as intact luteal phase sheep (29, 40). Thus the progesterone concentration gradient may be responsible for the decreasing eNOS expression gradient observed in luteal phase sheep. To our knowledge, an estrogen concentration gradient has not been reported in any vascular bed. However, an estrogen receptor density gradient was established in the baboon aorta (25) moving proximal to distal. Whether either estrogen-mediated mechanism is present in the uterine circulation remains to be elucidated.

A decreasing gradient of eNOS expression was also noted in the uterine vasculature of late gestation ewes. The observation that eNOS increased only slightly during late pregnancy is in contrast to our previous observations in which eNOS expression was quantified as pregnancy > follicular > luteal (28, 29, 40). Previously, UA endothelial-isolated proteins were obtained without accounting for regional differences. This apparent discrepancy may be accounted for by the dramatic remodeling of the uterine vasculature during pregnancy. Cipolla and Osol (3) reported that VSM and endothelial cells of rat UA undergo cellular hypertrophy and hyperplasia during pregnancy, which increases both the caliber and length of this artery. Our data and those of Fuller et al. (10) also suggest that the growth of the UA during pregnancy is due to increased diameter (Table 1). Furthermore, the internal radius and VSM content of UA in sheep (12) and the cross-sectional area and relative wall volume of rat UA (33) all increase during pregnancy. Because Weiner et al. (45) observed that the number of endothelial cells per micrometer squared is similar in UA from nonpregnant versus pregnant guinea pigs (12.3 ± 0.6 versus 13.2 ± 0.5 cells/µm²), we expect the total number of endothelial cells per micrometer squared to be consistent from each artery scraped but the relative proportion of endothelial isolated proteins derived from UA 1°, UA 2°, and UA 3° to differ substantially. Therefore, the mixing of endothelial isolated proteins from each region may have increased the differences between pregnant and nonpregnant groups we previously reported.

The external diameters of luteal phase UA, but not systemic arteries, were slightly higher than those of follicular phase UA. This is in contrast to the study by Ford et al. (8), which reported no difference in external diameter during these reproductive states; however, different-sized UA were not measured in their study. When we factor these diameters for the RA, differences between luteal and follicular phase sheep were no longer observed, supporting this previous report (8). Of potential importance, there are differences in the amount of tunica media present in small and large arteries. The tunica media of large arteries consists almost entirely of VSM, which progressively thins out with decreasing vessel diameter (24). Myatt et al. (30) demonstrated using immunohistochemistry a decreasing eNOS gradient in the human placental vasculature associated with a decreasing vessel diameter and thus VSM content. Our data suggest that differences in the amount of VSM are unlikely to represent the mechanism in which the eNOS expression gradient is regulated, because neither follicular phase nor late pregnant UA show such an eNOS expression gradient.

An alternate explanation for the upregulated eNOS expression of smaller diameter UA during the follicular phase and midpregnancy is their responsiveness to flow (23). Small resistance vessels may be more sensitive to flow changes than large arteries. Hull et al. (16) showed that a 10-fold increase in flow caused a 9% increase in diameter in the femoral artery but also a larger 15% increase in the smaller, downstream saphenous artery. In the coronary circulation, a 30% dilation of arterioles with resting diameters of 65 µm during flow enhancement (22) occurred, whereas large coronary arteries dilated only 3-10% (2, 6, 14). On the other hand, estrogen (4) and pregnancy (15) can enhance the reactivity of resistance vessels to flow/shear stress. It is not known if the presence of estrogen in follicular and midpregnant ewes (26), both which are associated with elevated UBF, alters the responsiveness of the uterine vasculature to flow.

Increases in UBF and thus shear stress during the follicular phase and midpregnancy are associated with upregulation of eNOS protein expression in UA 2° and UA 3°. Although UBF continues to rise in late pregnancy, a decreasing eNOS expression gradient is observed. The normalization of shear forces due to the increased diameter during late pregnancy may downregulate eNOS expression in these UA. Other investigators reported that chronic increases in blood flow result in increased diameter, a process that normalizes shear stress (18, 19, 48). Furthermore, flow-mediated dilation is enhanced in small subcutaneous (4) and myometrial arteries (20) during human pregnancy and in the presence of N-nitro-L-arginine methyl ester (L-NAME), an inhibitor of NO; the response to increasing shear stress was similar to that of nonpregnant women. Interestingly, 17-estradiol stimulated flow-mediated relaxation in small subcutaneous arteries from postmenopausal women but not premenopausal or pregnant women. This response was also abolished by L-NAME (21). Of physiological and clinical importance, preeclamptic pregnancy is associated with reduced flow-mediated vasodilation (4) associated with a lower circulating estrogen level and attenuated release of NO. However, this response may be restored via exogenous estrogen treatment (15, 20, 21). Thus the increase in eNOS expression in the smaller diameter UA observed in midpregnant ewes, which is followed by a decreasing expression gradient in late gestation and a concomitant rise in NO_x levels and an increase in diameter of UA 1°, may represent an adaptive mechanism that serves to normalize shear stress to physiological levels.