INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

CHRONIC HYPOXIA during the course of pregnancy is one of the most common insults to the maternal cardiovascular system and fetal development and is thought to be associated with increased risk of preeclampsia and fetal intrauterine growth restriction (20, 21, 34). Whereas the precise regulation of uterine blood flow is important for both growth and survival of the fetus and for maternal cardiovascular well being, the adaptive mechanisms of the uterine vasculature to chronic hypoxia are not clear. In pregnant women residing at a high altitude (3,100 m), uterine blood flow at 36 wk decreased compared with that of women at low altitude (1,600 m), because of a decreased vessel diameter resulting from a structural remodeling of the uterine artery (35). In contrast, the blood flow velocity was higher in the women living in a high altitude, which helped to compensate for the reduced diameter, and may have resulted from downstream vasodilation (35). In the guinea pig, chronic high-altitude hypoxia did not diminish the pregnancy-associated reduction in contractile sensitivity to phenylephrine but enhanced basal nitric oxide (NO) activity in the nonpregnant uterine artery and the pregnant mesenteric artery (31).

Endothelial NO plays a key role in modulating vascular reactivity. The effects of chronic hypoxia on endothelial NO synthesis and release have been investigated in both cultured cells and in vivo, but the results are controversial (1, 15, 16, 18, 24, 30). In rat pulmonary vasculature, chronic hypoxia increased endothelial NO release and upregulated endothelial NO synthase (eNOS) and inducible NO synthase gene and protein expression (8, 15, 27). In contrast, in the rat aorta, it has been shown that chronic hypoxia results in a decrease in eNOS protein and mRNA and results in impaired endothelium-dependent relaxation (30). In a recent study, White et al. (32) demonstrated that ACh-mediated relaxation of the uterine arteries of the near-term pregnant guinea pig was the same in those animals kept in a hypobaric chamber at a simulated high altitude (3,962 m) throughout gestation as in those animals kept at low altitude (1,600 m). Nonetheless, the effect of the NO synthase (NOS) inhibitor N^G-nitro-L-arginine (L-NNA) on the relaxation response to ACh was decreased in the uterine arteries of simulated high-altitude pregnant guinea pigs compared with the low-altitude controls. On the basis of these findings, the authors suggested that the stimulatory effect of pregnancy on NO in the guinea pig was diminished at high compared with low altitude. Nevertheless, the effect of high-altitude chronic hypoxia on uterine artery endothelial NO synthesis and release and eNOS gene expression was not examined.

Contrary to the finding by White et al. (32) in the guinea pig, we have recently demonstrated that long-term (~110 days), moderate high-altitude (3,820 m) exposure increases plasma nitrate levels in the near-term pregnant but not in nonpregnant sheep (38). The previous studies with the same animal model demonstrated that in the face of high-altitude chronic hypoxemia with maternal arterial PO₂ decreased from 102 to 64 mmHg, intrauterine growth restriction of the fetus was not observed (9). We reasoned that changes in uterine arterial vascular tone may play a key role in the adaptation of pregnant sheep to moderate high altitude. Indeed, we have demonstrated that long-term, moderate high-altitude hypoxia attenuates uterine artery contractility in pregnant sheep by suppressing vascular smooth muscle pharmacomechanical coupling (5-7, 37).

The present study was designed to test the hypothesis that chronic exposure to moderate high-altitude hypoxia in vivo increases eNOS gene expression and endothelium-dependent vasorelaxation in the uterine artery of pregnant sheep. The specific objectives of the study were to determine in the uterine artery to what extent chronic hypoxia increases 1) calcium ionophore A23187-mediated relaxation, 2) basal and agonist-stimulated endothelial NO release, and 3) eNOS protein and mRNA levels in the endothelium. To determine the potential effect of chronic hypoxia on endothelium-independent vasorelaxation, sodium nitroprusside (SNP)- and 8-bromoguanosine 3',5'-cyclic monophosphate (8-BrcGMP)-induced relaxations of the uterine artery were also examined. The studies were performed in both nonpregnant and pregnant sheep to determine whether the effect of chronic hypoxia is specific to pregnant animals. To examine the tissue specificity, the effect of chronic hypoxia on eNOS protein expression was also determined in femoral and renal arteries.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Tissue preparation. The same animal model as previously described (5, 7, 9, 38) was used in the present study. Time-dated pregnant sheep were obtained from Nebeker Ranch in Lancaster, CA [altitude: ~300 m; with arterial PO₂ (Pa_O2) being 102 ± 2 mmHg]. Uterine arteries were obtained from nonpregnant and near-term (~140 days of gestation) pregnant sheep. To induce chronic hypoxia, the nonpregnant and pregnant (30 days of gestation) animals were transported to Barcroft Laboratory, White Mountain Research Station, in Bishop, CA (altitude, 3,820 m; Pa_O2, 64 ± 2 mmHg) and kept there for ~110 days, whereas the control animals were maintained near sea level (~300 m). The animals were transported (~6 h) to the laboratory immediately before the studies. To minimize the potential effect of stress caused by transportation, tissues were collected the next day after the animals had rested overnight. Control studies of this preparation have shown no difference in responses among those tissues collected the next day and 4 days later. Anesthesia of the animals was rapidly induced with intravenous injection of thiamylal (10 mg/kg). The ewes were then intubated, and anesthesia was maintained on 1.5-2.0% halothane in oxygen throughout surgery. An incision was made in the abdomen and the uterus exposed. The uterine, renal, and femoral arteries were removed without stretching and were placed into a modified Krebs solution (pH 7.4) composed of (in mM) 115.21 NaCl, 4.70 KCl, 1.80 CaCl₂, 1.16 MgSO₄, 1.18 KH₂PO₄, 22.14 NaHCO₃, and 7.88 dextrose. After the tissues were removed, the animals were euthanized with T-61 (Hoechst-Roussel; Somerville, NJ). All of the procedures and protocols used in the present study were approved by the Animal Research Committee of Loma Linda University and were used in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Contractile studies. Fourth (pregnant) and third (nonpregnant) branches of main uterine arteries and femoral arteries were cut into ~2-mm ring segments. In some rings, the endothelium was removed by gentle rotation of the artery rings on an approximately sized, rough-surfaced, blunt hypodermic needle, as described previously (7). Validation of endothelium removal was demonstrated by the elimination of endothelium-dependent relaxation induced by ATP and by the examination of endothelial integrity by using en face silver staining. Contractile responses of arterial rings were quantified in the Krebs solution in tissue baths at 37°C, as described previously (7). Isometric tensions were measured. After 60 min of equilibration in the tissue bath, each ring was stretched to the optimal resting tension as determined by the tension developed in response to potassium chloride (120 mM) added at each stretch level. Concentration-response curves were obtained by cumulative addition of the agonist in approximately one-half log increments. For relaxation studies, the tissues were precontracted with submaximal concentration (1 µM) of phenylephrine, and then the calcium ionophore A23187, SNP, or 8-BrcGMP, respectively, were added in a cumulative manner.

NO measurement. From the relaxation study, 0.1 ml of the bath solution was collected before and after each dose of the calcium ionophore A23187. Samples were flash-frozen in liquid N₂ and stored at 80°C. NO was measured by chemiluminescence method, as described previously (33). Because of the instability of NO in oxygenated physiological solution, it is rapidly converted to nitrite and further to nitrate. Nitrite and nitrate are relatively stable in the solution and are readily reduced back to NO in vanadium III-HCl solution. The samples (100 µl) were injected into the gas purge vessel containing 5 ml vanadium III-HCl to react for 1 min and reduce nitrite and nitrate in the sample back to NO. To achieve high reduction efficiency, the reduction was performed at 90°C. The NO in the sample was then "stripped" into the head space of the gas purge vessel by bubbling it with helium (12 ml/min) for 60 s. The NO in the head space was drawn into a NO analyzer (model 270B, Sievers Instruments, Boulder, CO) and mixed with ozone (O₃) in the front of a cooled, red-sensitive photomutiplier tube (Hamamatsu). Signals from the detector were analyzed by an online computer as area under the peak. The measurement reflected the combined concentrations of nitrite, nitrate, and NO (NO_x) of each sample.

Western blot analysis of eNOS. The freshly isolated endothelial cells from the combined fourth, third, and second branches and main uterine arteries and from the femoral and renal arteries were used. The endothelium was isolated by gently scraping from the vessel lumen as previously described (17, 33). The cells were then solubilized by sonication in lysis buffer composed of 150 mM NaCl, 50 mM Tris · HCl, 10 mM EDTA, 0.1% Tween 20, 0.1% -mercaptoethanol, 0.1 mM phenylmethylsulfonyl fluoride, and 5 µg/ml aprotinin, pH 7.4. After centrifugation, proteins were quantified in the supernatant by the method of Bradford (3). Samples with equal protein (10 µg) were loaded on a 7.5% polyacrylamide gel with 0.1% SDS and were separated by electrophoresis at 100 V for 1 h. Proteins were then transferred onto an Immobilon-P membrane at 30 V for 45 min at room temperature by using a semidry blotter (Bio-Rad). The Immobilon-P membrane was probed by mouse eNOS monoclonal antibody (1:750, Transduction; Lexington, KY), followed by the secondary horseradish peroxidase-conjugated goat anti-mouse (1:2,000) antibody (Amersham; Arlington Heights, IL). Actin was used as a loading control. Because the effect of pregnancy and/or chronic hypoxia on actin content is unclear, eNOS was not normalized individually to the actin signal. Rather, an external eNOS standard from bovine coronary artery endothelial cells was used to normalize sample eNOS. Proteins were visualized with enhanced chemiluminescence reagents (Amersham), and the blots were exposed to Hyperfilm. Results were quantified by the scanning densitometer (model 670, Bio-Rad).

Measurement of eNOS mRNA by RT-PCR. eNOS mRNA was quantified by coupled RT-PCR amplification in a single tube assay as described previously (2). Total cellular RNA was extracted from freshly isolated endothelial cells by using a RNAqueous kit (Ambion) following the manufacturer's protocol. The forward and reverse primers used for targeting amplification from part of the ovine eNOS protein-coding region (GenBank accession no. U76738) were 5'-TGTGGCTGTCTGCATGG-3' and 5'-TGGCTGGTAGCGGAAGG-3', respectively. The final product included 300 bases. The RNA (0.1 µg) from each sample was incubated in a 50-µl final volume containing 1× PCR buffer; 2 mM MgCl₂; 10 nmol each of dATP, dCTP, dTTP, and dGTP; and 30 pmol of each forward and reverse temperature-matched primer. Amplification was performed in the presence of 0.1 µl of avian myeloblastosis virus (AMV) reverse transcriptase (2.5 U) and 1 µl of Taq polymerase (5 U) except for RT-controls, which contained only Taq polymerase. The program used was annealing at 62°C for 10 min, RT at 50°C for 10 min, denaturing at 94°C for 2 min, and amplifying for 28 cycles (each cycle consisted of 94°C for 30 s, 62°C for 30 s, and 72°C for 30 s). Final products were extended to full length by incubation at 72°C for 30 s. A standard curve contained known numbers of eNOS cDNA target sequence. At the end of the assay, 15 µl of products were separated on 2% agarose gel. The transcript levels of eNOS were visualized by SYBO Gold staining and analyzed by using densitometry, as described previously (12). The results were calculated from the standard curve of 10⁴ to 10¹⁰ copies eNOS cDNA plasmid run in each assay.

Data analysis. Concentration-response curves were analyzed by computer-assisted nonlinear regression to fit the data by using Prism (GraphPad Software, San Diego, CA). Half-maximal effective concentration (EC₅₀) values for an agonist in each experiment were taken as the molar concentration at which the contraction-response curve intersected 50% of the maximum response and were expressed as pD2 (log EC₅₀) values. For NO data analysis, the area under the peak was continuously integrated during sample measurement by using the data acquisition software WorkBench (Kent Scientific, Litchfield, CT). Results were expressed as means ± SE, and the differences were evaluated for statistical significance (P < 0.05) by ANOVA, followed by Newman-Keuls post hoc test. Where appropriate, we used two-way ANOVA with altitude and pregnancy as the two factors.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Effect of chronic hypoxia on basal endothelial activity. Norepinephrine-induced dose-dependent contractions of the uterine arteries from control and chronic hypoxic pregnant animals are illustrated in Fig. 1. In Fig. 1, removal of endothelial cells shifted the norepinephrine dose-response curves to the left. The pD₂ values were significantly different (P < 0.05) between denuded and intact tissues in both control (7.29 ± 0.13 vs. 6.91 ± 0.11, n = 8) and hypoxic (7.09 ± 0.03 vs. 6.45 ± 0.04, n = 10) animals. Moreover, the augmentation of the pD₂ values between endothelium-denuded and -intact tissues was significantly higher in hypoxic (0.64 ± 0.084) than in control (0.38 ± 0.056) animals (P < 0.05).

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Norepinephrine-induced, concentration-dependent contractions of pregnant uterine arteries. The contractions were obtained with intact endothelium and denuded uterine arteries from control (B) and chronic hypoxic (A) pregnant sheep. Data are means ± SE of 8-10 animals, and pD2 values are presented in the text.

Effect of chronic hypoxia on endothelium-dependent relaxation. The endothelium-dependent relaxation induced by calcium ionophore A23187 in uterine arteries from control and hypoxic sheep are shown in Fig. 2A. Calcium ionophore A23187 produced dose-dependent relaxations of uterine arteries precontracted with 1 µM phenylephrine in both nonpregnant and pregnant animals. The relaxation induced by A23187 was significantly increased in pregnant compared with nonpregnant tissues (maximum response: 83.9 ± 14.3 vs. 55.7 ± 6.7%). Whereas chronic hypoxia did not change A23187-induced relaxation in nonpregnant uterine arteries, it significantly increased A23187-induced relaxation of pregnant uterine arteries (Fig. 2A). In contrast, the endothelium-independent relaxations induced by SNP and 8-BrcGMP were not altered by chronic hypoxia (Fig. 2, B and C). Removal of endothelial cells abolished A23187-induced relaxation but had no effect on SNP and 8-BrcGMP-induced relaxations.

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Agonist-mediated relaxations of pregnant uterine arteries. Concentration-dependent relaxations induced by calcium ionophore A23187 (A), sodium nitroprusside (B), and 8-bromoguanosine 3',5'-cyclic monophosphate (8-BrcGMP) (C) were obtained with phenylephrine (1 µM)-precontracted uterine arteries. Data are means ± SE of 4-10 animals. *P < 0.05, comparison of the corresponding values between control and hypoxic groups.

Effect of chronic hypoxia on NO release. In pregnant animals, the basal level of NO release was significantly higher in hypoxic (59.8 ± 10.5 pmol/100 µl, n = 3) than in control (32.1 ± 1.8 pmol/100 µl, n = 11) uterine arteries (P < 0.05). As shown in Fig. 3, the calcium ionophore A23187 produced concentration-dependent increases in NO release from the uterine arteries, which were significantly higher in hypoxic than in control tissues (maximum response: 178.6 ± 36.5 vs. 56.3 ± 3.6 pmol/100 µl). In contrast to the pregnant animals, chronic hypoxia did not change either basal (17.4 ± 6.4 vs. 20.1 ± 1.8 pmol/100 µl) or A23187 (1 µM)-induced (28.3 ± 9.3 vs. 46.2 ± 14.3 pmol/100 µl) NO release in the nonpregnant uterine arteries.

View larger version (15K): [in this window] [in a new window]   Fig. 3.   Calcium ionophore A23187-induced endothelial nitric oxide (NO) releases in pregnant uterine arteries. NO was measured as nitrite, nitrate, and NO (NOx) by using chemiluminescence assay described in MATERIALS AND METHODS. Data are means ± SE of 3-11 sheep. *P < 0.05, comparison of the corresponding values between control and hypoxic groups.

Effect of chronic hypoxia on eNOS protein expression. Figure 4 shows the effect of chronic hypoxia on eNOS protein expression in the endothelium of uterine arteries. The representative Western immunoblots showed that the monoclonal antibody for eNOS detected a single band at the expected size of 140 kDa (Fig. 4, top). Two-way ANOVA with pregnancy as one factor and altitude as the other indicated that the differences in pregnancy (P < 0.0001) and altitude (P = 0.0113) were significant. Pregnancy increased eNOS protein expression in uterine artery endothelial cells, in agreement with previous observations (16). Futhermore, of direct relevance to this study, we showed that chronic hypoxia augmented pregnancy-induced increase in eNOS protein expression. Whereas there was no difference for eNOS protein levels in the nonpregnant uterine artery endothelium between control and hypoxic animals, chronic hypoxia resulted in an onefold increase of eNOS protein in pregnant uterine artery endothelial cells (Fig. 4, bottom). In comparison, neither pregnancy nor hypoxia affected eNOS protein expression in the endothelium of femoral and renal arteries (Fig. 5).

View larger version (33K): [in this window] [in a new window]   Fig. 4.   Endothelial NO synthase (eNOS) protein expression in uterine arteries. Top: Western analysis of eNOS was performed in freshly isolated endothelial cells of uterine arteries obtained from four groups of sheep: control nonpregnant (cnps), control pregnant (cps), hypoxic nonpregnant (hnps), and hypoxic pregnant (hps). Immunoblot illustrates eNOS bands detected by the monoclonal antibody at the expected size of 140 kDa. Bottom: summary findings for quantitative densitometry of samples from 6 to 12 animals. Data are percent density of the standard (std; 10 µg of protein from bovine coronary artery endothelial cells), loaded with the samples. Actin bands show an equal loading for the samples. *P < 0.05, pregnant vs. nonpregnant; +P < 0.05, hypoxia vs. control.

View larger version (16K): [in this window] [in a new window]   Fig. 5.   eNOS protein expression in femoral (A) and renal (B) arteries. Western analysis of eNOS was performed in freshly isolated endothelial cells of femoral and renal arteries obtained from cnps, cps, hnps, and hps groups. Data are percent density of the standard (10 µg of protein from bovine coronary artery endothelial cells) loaded with the samples and are means ± SE of 5 animals in each group.

Effect of chronic hypoxia on eNOS mRNA levels. The levels of eNOS mRNA in the uterine artery endothelium were analyzed by using a RT-PCR mass assay as previously reported (2). The calibration curve, using known copy numbers of eNOS cDNA templates, showed an excellent linear correlation between densities and copy numbers in the range of 10⁴ to 10¹⁰ copies. The effect of chronic hypoxia on eNOS mRNA levels in the uterine artery endothelium is shown in Fig. 6. Two-way ANOVA with pregnancy as one factor and altitude as the other indicated that the differences in pregnancy (P < 0.0001) and altitude (P = 0.0002) were highly significant. In agreement with the elevated eNOS protein levels, pregnancy was also observed to significantly increase eNOS mRNA levels in the uterine artery endothelium as previously described (2). Chronic hypoxia, however, resulted in a threefold to ninefold increase in eNOS mRNA in the nonpregnant and pregnant uterine artery endothelium, respectively (Fig. 6).

View larger version (33K): [in this window] [in a new window]   Fig. 6.   eNOS mRNA level in uterine artery endothelium. eNOS mRNA levels were quantified by RT-PCR, as described in MATERIALS AND METHODS, in freshly isolated endothelial cells of uterine arteries obtained from cnps, cps, hnps, and hps groups (top). Representative gel with eNOS bands at 300 bases is shown. Bottom: summary findings for quantitative densitometry of samples from 9 animals in each group. *P < 0.05, pregnant vs. nonpregnant; +P < 0.05, hypoxia vs. control.

View larger version (14K): [in this window] [in a new window]   Fig. 7.   Relationship of eNOS protein and mRNA in uterine artery endothelium. The steady-state eNOS protein and mRNA levels from Figs. 4 and 6 are expressed as increases in the values obtained from control nonpregnant uterine artery endothelium. A linear correlation is shown between the mRNA and protein among cnps, cps, hnps, and hps sheep. The slope (0.12 ± 0.01) defines the apparent eNOS mRNA translational efficiency.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The present study has demonstrated that long-term, moderate high-altitude hypoxia during the course of pregnancy in sheep increases the following: 1) eNOS protein and mRNA in uterine artery endothelium, 2) basal and calcium ionophore A23187-induced NO release from the uterine artery, and 3) endothelium-dependent relaxation of phenylephrine-precontracted uterine artery rings. It is unlikely that the increased endothelium-dependent relaxation is because of changes at downstream signals of soluble guanylate cyclase or cGMP-dependent protein kinase, because chronic hypoxia had no effect on SNP- or 8-BrcGMP-mediated relaxation in the uterine artery.

The present finding that long-term, high-altitude exposure is associated with the upregulation of eNOS protein and NO production in pregnant sheep uterine arteries is consistent with our previous studies, in which we demonstrated in the same animal model that high altitude significantly increased plasma nitrate levels in pregnant but not in nonpregnant sheep (38). In addition, in agreement with our previous study (33), the present study has demonstrated that removal of the endothelium potentiates norepinephrine-induced contractions in pregnant but not in nonpregnant uterine arteries. More importantly, the present study showed that chronic hypoxia exaggerated the pregnancy-associated responses, suggesting an increase in either basal NO synthesis and release and/or some other changes in downstream mechanisms of the NO pathway. Though it is not clear about the role of ₂-adrenoceptors, if any, on sheep uterine artery endothelium, the present study cannot rule out the possibility that norepinephrine may stimulate NO release by binding to ₂-adrenoceptors. Similar findings (31) were obtained in the guinea pig, in which four groups of animals (low-altitude pregnant, nonpregnant, high-altitude pregnant, and nonpregnant) were studied, and the NOS inhibitor L-NNA potentiated phenylephrine-induced contractions (decreased EC₅₀) significantly only in the uterine artery of the high-altitude pregnant guinea pig. Whereas White et al. (31) clamed that L-NNA also raised contractile sensitivity in nonpregnant high-altitude uterine arteries of the guinea pig, the changes in the EC₅₀ (from 4.8 × 10⁸ to 3.4 × 10⁸ M) and the maximal contraction (from 1,444 ± 259 to 1,381 ± 232 mg) of phenylephrine-induced contractions in the absence and presence of L-NNA were not significant. On the basis of these findings, it would be more likely that chronic high-altitude hypoxia enhances basal NO activity in the pregnant uterine artery, but not in the nonpregnant uterine artery of the guinea pig as the authors concluded (31). Indeed, direct measurement of NO production in the present study indicated that basal NO synthesis and release was significantly elevated in hypoxic, compared with control, uterine arteries of pregnant sheep. The close correlation of onefold increases in both basal NO release and eNOS protein levels suggest that the increased basal NO release is predominantly because of the increased eNOS protein in the hypoxic arteries. On the other hand, the degree of increase of NO release induced by the calcium ionophore A23187 was higher than that of eNOS protein expression in hypoxic uterine arteries, suggesting a hypoxic-mediated increase in sensitivity of calcium signaling pathway of eNOS additional to the enhanced protein expression.

The finding that chronic hypoxia increased eNOS protein levels in the endothelium of pregnant uterine artery, but not femoral and renal arteries from either pregnant or nonpregnant animals, indicates a tissue-specific effect of chronic hypoxia on the uterine artery. The marked regional variability in eNOS protein expression during systemic in vivo chronic hypoxia has been demonstrated previously (15, 28, 30). In the rat pulmonary circulation, chronic hypoxia increased pulmonary eNOS protein and mRNA and augmented endothelium-dependent pulmonary arterial dilation (15, 27). However, in the rat aorta, prolonged exposure to hypoxia resulted in a decrease in aortic eNOS protein and mRNA and impaired endothelium-dependent relaxation (30). In a recent study (32), consistent with our findings, neither pregnancy nor altitude effected bradykinin-induced relaxation of thoracic arteries of the guinea pig. Contrary to our findings, White et al. (32) concluded that the stimulatory effect of pregnancy on NO in the guinea pig was diminished at high altitude compared with low altitude, on the basis of decreased effect of NOS inhibitor L-NNA on the relaxation response to ACh in the uterine arteries of simulated high-altitude pregnant guinea pig, compared with the low-altitude controls. However, the effect of high-altitude chronic hypoxia on uterine artery endothelial NO synthase and release and eNOS gene expression was not examined (32). Whereas the reason(s) for the different findings between the present study and the previous one (32) is not entirely clear, it is likely due, at least in part, to the differences in species (sheep vs. guinea pig) and in altitude (3,820 vs. 3,962 m) between the two studies. In addition, because the calcium ionophore A23187 used in the present study induces NO release and endothelium-dependent relaxation through a receptor-independent mechanism, our results do not examine the potential changes in endothelial receptors and coupling signals by chronic hypoxia, but rather focus on changes in NOS. Hence the increase in A23187-mediated, endothelium-dependent relaxation of the uterine artery by chronic hypoxia is likely because of an increase in NO release. This is confirmed by the direct measurement of A23187-induced NO releases in the uterine arteries. In a study by White et al. (32), the ACh-induced relaxation of the uterine artery was examined in the guinea pig, and hence it would be difficult to exclude entirely the effects of prolonged exposure to high altitude on endothelial muscarinic receptor function and/or postreceptor intracellular pathways that lead to NO release.

The mechanisms underlying this regional variability in eNOS protein expression in response to chronic hypoxia in vivo are not presently clear. It is not unlikely that hypoxia per se may have different regulatory effects on eNOS protein and mRNA in different vascular beds. However, studies (1, 16, 18, 24) in cultured endothelial cells have yielded conflicting results of the effects of hypoxia (24 h) on eNOS protein and mRNA. On the other hand, other factors as result of systemic in vivo hypoxia are also likely to play an important role in determining the selective responses of vascular beds to chronic hypoxia. Controversy exists concerning the relative roles of hypoxia per se versus hemodynamic factors associated with hypoxia, such as altered vascular shear forces or vascular remodeling, in mediating eNOS responses to chronic hypoxia. In the rat pulmonary circulation, reduction of blood flow to the left lung by left pulmonary arterial stenosis does not prevent the upregulation of eNOS after chronic hypoxia, suggesting that hypoxia may increase eNOS expression independently of hemodynamic changes associated with pulmonary hypertension (14). However, in the same animal model, Resta et al. (26) demonstrated that chronic hypoxic-induced increases in vascular mechanic forces and shear stress and/or vascular remodeling contributed to enhanced endothelium NO-dependent arterial dilation and the upregulation of arterial eNOS in the rat pulmonary circulation. Whereas it is not clear whether the uterine blood flow is altered in response to chronic hypoxia in pregnant sheep, human studies (35) have clearly demonstrated that uterine artery blood flow velocity (hence the shear stress) is significantly increased by chronic moderate high-altitude hypoxia in pregnant women. Shear stress increases eNOS mRNA and protein levels in cultured endothelial cells (25), and vessels exposed to chronic elevations in shear stress exhibit augmented endothelium-dependent relaxation (19). Given that the uterine circulation undergoes dramatic vascular remodeling and hemodynamic changes during pregnancy, and this process is particularly vulnerable to hypoxia, we speculate that the effect of chronic hypoxia on uterine vascular remodeling and hemodynamic changes during the course of pregnancy may play an important role in the upregulation of eNOS in uterine artery endothelium.

Our finding that pregnancy was associated with a significant increase in eNOS expression in the uterine artery endothelium confirms previous studies (2, 17, 33). Although chronic hypoxia increased eNOS mRNA in the endothelium of nonpregnant uterine arteries, it had no effect on eNOS protein, NO production, or endothelium-dependent relaxation. This suggests that the effect of chronic hypoxia is selective to pregnant uterine arteries. The reason that chronic hypoxia increased eNOS mRNA by threefold, but did not significantly change eNOS protein levels in the nonpregnant uterine artery endothelium, may have resulted at least in part from the low apparent translational efficiency (0.12) of eNOS mRNA determined in the present study. In fact, it has been shown that many mammalian genes translate relatively poorly (13). In pregnant uterine artery endothelium, chronic hypoxia was associated with a ninefold increase in steady-state eNOS mRNA levels, and thus resulted in a onefold increase in eNOS protein. The finding that chronic hypoxia did not change apparent translational efficiency of eNOS mRNA, but increased eNOS mRNA levels, suggests that increased eNOS protein expression in the pregnant uterine artery endothelium may not be regulated at the translational level. In addition, it is unlikely that increased steady-state eNOS protein levels found in hypoxic compared with control pregnant uterine artery endothelium is because of increased protein stability because that eNOS protein versus message is constant among four groups. However, it is not clear from the present study the extent to which increased steady-state mRNA levels are resulted from increased transcription or enhanced message stability.

To test whether uterine artery smooth muscle became more sensitive to NO by chronic hypoxia and thereby generated more cGMP, we examined the relaxation response produced by SNP in the uterine arteries. The finding that SNP-induced relaxations were the same in control and hypoxic uterine arteries indicates that the signal transduction pathways distal to NO are unchanged by chronic hypoxia. Furthermore, the finding that relaxation responses to 8-BrcGMP were not affected by chronic hypoxia, in either pregnant or nonpregnant uterine arteries, indicates that the cGMP-dependent kinase sensitivity is unaltered by chronic hypoxia. It has been demonstrated (2, 4, 17, 22, 23, 29, 33) that pregnancy-induced increase in uterine artery relaxation is caused by increased NO release but not signal transduction pathways distal to NO production. Collectively, our study suggests that chronic hypoxia augments pregnancy-induced vascular adjustments of the uterine artery by further enhancing endothelial NO production.

The physiological significance of the enhanced NO production of pregnant uterine arteries by chronic hypoxia is not fully characterized at present. Given that chronic hypoxia produces profound stress on maternal and fetal cardiovascular systems, the increased NO synthesis is likely to represent an adaptive mechanism to the stress of chronic hypoxia and may play an important role in the acclimatization of the mother and fetus to moderate chronic hypoxia. Among other effects, NO plays an important role in the regulation of vascular reactivity. Our previous studies demonstrated that moderate long-term high-altitude hypoxia attenuated uterine artery contractility in pregnant sheep by suppressing ₁-adrenoceptor-mediated inositol 1,4,5-trisphosphate [Ins(1,4,5)P₃] signaling pathway in the uterine artery (6). NO activates soluble guanylate cyclase and increases cGMP in vascular smooth muscle. Indeed, chronic hypoxia significantly increased cGMP levels in the uterine artery (36). cGMP, through activation of cGMP-dependent protein kinase, induces phosphorylation of the Ins(1,4,5)P₃ receptor in vascular smooth muscle and decreases the Ins(1,4,5)P₃ binding affinity (10, 11). We demonstrated that treatment of the uterine artery with the cGMP analogue 8-BrcGMP significantly decreases Ins(1,4,5)P₃ binding affinity to its receptors (36). Thus by suppressing contractile response and enhancing relaxation, the increased NO is likely to play an important compensatory role in maintaining the uterine artery in a dilated state under the stress of chronic hypoxia during the course of pregnancy and may contribute to the successful adaptation of the pregnant sheep and fetus to moderate high-altitude hypoxia.