Pregnancy modifies the large conductance Ca²⁺-activated K⁺ channel and cGMP-dependent signaling pathway in uterine vascular smooth muscle

Abstract

Regulation of uteroplacental blood flow (UPBF) during pregnancy remains unclear. Large conductance, Ca²⁺-activated K⁺ channels (BK_Ca), consisting of α- and regulatory β-subunits, are expressed in uterine vascular smooth muscle (UVSM) and contribute to the maintenance of UPBF in the last third of ovine pregnancy, but their expression pattern and activation pathways are unclear. We examined BK_Ca subunit expression, the cGMP-dependent signaling pathway, and the functional role of BK_Ca in uterine arteries (UA) from nonpregnant (n = 7), pregnant (n = 38; 56–145 days gestation; term, ∼150 days), and postpartum (n = 15; 2–56 days) sheep. The α-subunit protein switched from 83–87 and 105 kDa forms in nonpregnant UVSM to 100 kDa throughout pregnancy, reversal occurring >30 days postpartum. The 39-kDa β₁-subunit was the primary regulatory subunit. Levels of 100-kDa α-subunit rose ∼70% during placentation (P < 0.05) and were unchanged in the last two-thirds of pregnancy; in contrast, β₁-protein rose throughout pregnancy (R² = 0.996; P < 0.001; n = 13), increasing 50% during placentation and approximately twofold in the remainder of gestation. Although UVSM soluble guanylyl cyclase was unchanged, cGMP and protein kinase G_1α increased (P < 0.02), paralleling the rise and fall in β₁-protein during pregnancy and the puerperium. BK_Ca inhibition not only decreased UA nitric oxide (NO)-induced relaxation but also enhanced α-agonist-induced vasoconstriction. UVSM BK_Ca modify relaxation-contraction responses in the last two-thirds of ovine pregnancy, and this is associated with alterations in α-subunit composition, α:β₁-subunit stoichiometry, and upregulation of the cGMP-dependent pathway, suggesting that BK_Ca activation via NO-cGMP and β₁ augmentation may contribute to the regulation of UPBF.

in mammalian species, fetal growth and well-being depend on successful placentation and the subsequent rise in maternal uteroplacental blood flow (UPBF) (31, 34–36). In sheep, UPBF increases >30-fold to >1,000 ml/min at term (34, 36, 41). This occurs in three steps (41). The first is preplacentation, when blood flow increases 100–150 ml/min and reflects vasodilation due to ovarian steroids (10–12, 36). This is followed by a further rise to ∼500 ml/min between 40 and 60 days of gestation, which is associated with angiogenesis and development of the maternal placental vascular bed (2, 32, 36, 41). In the final period, which occurs in the last two-thirds of pregnancy, UPBF rises to 1,000–1,500 ml/min by term and accounts for ∼25% of maternal cardiac output (34, 36, 41). This rise in UPBF is primarily vasodilation and an increase in maternal placental blood flow, which accounts for ∼90% of total UPBF by term (2, 35, 36, 52).

The mechanisms that contribute to the profound uterine vasodilation in the last two-thirds of pregnancy remain unclear. Uterine vascular synthesis of vasodilating prostaglandins, e.g., prostacyclin, increases in pregnant sheep (19, 20), but their inhibition does not alter basal UPBF (22, 25). Uterine vascular nitric oxide (NO) synthesis also increases, paralleling enhanced expression of endothelial NO synthase (eNOS) and synthesis of cGMP (7, 23, 39, 50, 57). However, acute uterine NOS inhibition decreases basal UPBF modestly (24, 39). Estrogen is a potent uterine vasodilator in nonpregnant and pregnant sheep (21, 36, 40) and is markedly elevated in pregnancy (4, 5), but its contribution to the rise in UPBF remains unclear. It is possible that each of these modulators contributes to the rise in UPBF, but their importance and the signaling cascades remain to be determined.

Large conductance, Ca²⁺-activated K⁺ channels (BK_Ca) are expressed in vascular smooth muscle (VSM) and modify vasorelaxation, vasoconstriction, and myogenic responses (16, 17, 27, 49, 51). They consist of four α-subunits that form the pore and are derived from the Slo gene and as many as four regulatory β-subunits that modify channel activity and sensitivity to activators, resulting in substantial channel diversity (3, 29, 55). The α-subunit may undergo posttranslational modifications, further adding to the functional diversity (16, 17, 49). At least four β-subunits exist, but β₁ and β₂ predominate in VSM (16, 17). BK_Ca contribute to estrogen- and NO-induced vasodilation in several vascular beds, including the ovine uterus (38, 44, 58), e.g., 17β-estradiol (E₂) induces vasodilation via NO-induced increases in VSM cGMP and channel phosphorylation and/or directly activates the channel via the β₁-subunit (9, 16, 17, 44, 53, 55). E₂ also increases uterine VSM (UVSM) β₁-subunit expression in nonpregnant ewes (26), which alters subunit stoichiometry and thus BK_Ca function. Furthermore, BK_Ca inhibition with tetraethylammonium chloride (TEA) at concentrations ≤1 mM (27) has specificity for the channel and 0.35 mM decreases basal UPBF 70–80% in the last third of ovine pregnancy (43). Thus BK_Ca appear to be an excellent downstream candidate for regulating UPBF.

Although BK_Ca appear to contribute to UPBF regulation in pregnancy (38, 43), the expression pattern and subunit stoichiometry are unknown. Thus we examined 1) pregnancy-related changes in α-subunit isoform; 2) α- and β-subunit expression and stoichiometry in nonpregnant, pregnant, and postpartum ewes; 3) the cGMP-dependent signaling pathway for BK_Ca activation; and 4) the role of BK_Ca in UVSM relaxation-contraction responses.

METHODS

Tissue preparation.

Tissues were collected from seven nonpregnant, 38 pregnant (56–145 days gestation; term, ∼150 days), and 15 postpartum (2–54 days) sheep of mixed Western breed. Animals were euthanized with intravenous pentobarbital sodium (120 mg/kg). The abdomen was opened, the intact uterus was removed in block, and first and second generation uterine arteries were quickly dissected from both uterine horns and placed in sterile chilled physiological saline solution. While on ice, the adventitia and intraluminal blood were removed, and segments were opened along the long axis, and the endothelium was removed with a soft cotton swab, frozen in liquid nitrogen, and stored at −80°C (26, 48). Additional samples were prepared for immunohistochemistry, as described below, while a 2.5–3.0 cm segment was placed in physiological salt solution of (in mM) 120.5 NaCl, 4.8 KCl, 1.2 MgSO₄, 1.2 NaH₂PO₄, 20.4 NaHCO₃, 1.6 CaCl₂, 10 dextrose, and 1.0 pyruvate at room temperature and transported within 30 min to the muscle physiology laboratory for contraction-relaxation studies. The studies described were approved by the Institutional Animal Care and Use Committee at the University of Texas Southwestern Medical Center at Dallas.

Function studies.

Samples collected from six ewes at 54–58 days of gestation (the completion of placentation) and five at 142–146 days (term, ∼150 days) were used to study uterine artery function. Segments were cut into 4 to 5 mm rings, and the endothelium was removed by carefully turning the tip of a small forceps in the lumen. Endothelium removal was verified histologically, and the basal lamina and underlying VSM were intact. Each ring was placed on a stirrup attached to a transducer (Grass FT-03, Quincy, MA; and iWorx FT-302, Dover, NH) to measure force generation in a 25-ml volume chamber as previously reported (14, 45). The PBS containing (in mM) 120.5 NaCl, 4.8 KCl, 1.2 MgSO₄, 1.2 NaH₂PO₄, 20.4 NaHCO₃, 1.6 CaCl₂, 10.0 dextrose, and 1.0 pyruvate in the chamber was maintained at 37°C throughout the experiment by a circulating water pump. After a 30-min equilibration period, rings were progressively stretched to obtain optimal length (L_o), as determined with progressive increases in KCl-induced contractile responses with 65 mM KCl. Unresponsive rings were discarded. Studies with each ring were performed in duplicate. Dose-response curves were constructed at L_o with KCl (10–120 mM) and cumulative doses of phenylephrine (PE; 10⁻⁸ to 10^−4.5 M), an α₁-agonist, to examine nonreceptor and adrenergic receptor-mediated responses, respectively. To determine the role of BK_Ca in uterine artery contraction-relaxation responses, we employed the channel inhibitor TEA, which has specificity for the BK_Ca at concentrations ≤1.0 mM (27). Our laboratory previously used this agent to study channel function in intact sheep (38, 43, 44) and human uterine arteries (45). At the end of the dose responses and after returning to baseline, rings were exposed to 10⁻⁶ M PE followed by sodium nitroprusside (SNP; 10⁻⁷ M), a NO donor, to obtain control contraction-relaxation responses with activated channels. The rings were washed and allowed to return to baseline tension for a minimum of 30 min. At that time, duplicate rings were exposed to one of three doses of TEA (0.2, 0.5, and 1.0 mM) followed by 10⁻⁶ M PE, 10⁻⁷M SNP, and a wash as described above. In some instances, we examined the effect of PE alone after the first exposure to determine whether there was tachyphylaxis; this did not occur (P > 0.1). Data were recorded on an electronic data-acquisition system (Summit ACQuire and Summit DaStar; Gould Systems, Valley View, OH) in grams of force generated at L_o. The medial thickness and length of each segment was measured at the completion of studies, and responses were expressed as Newtons per squared meters (10³ × N/m²).

Western blot analysis.

Samples of endothelium denuded uterine arteries (30 mg) were weighed and homogenized in 40× volumes of SDS buffer as previously described (26). Homogenates were centrifuged at 10,000 g for 2 min. The supernatant was removed, and an aliquot was used to determine cellular or soluble protein by BCA reagent (Pierce, Rockford, IL). Bromophenol blue and 2-mercaptoethanol were added to aliquots, and equal amounts of soluble protein (20 or 30 μg) were loaded on 7.5% polyacrylamide minigels, subjected to SDS-PAGE, and proteins were electrophoretically transferred to nitrocellulose paper at 100 V for 1 h. Blots were incubated overnight with antisera to the BK_Ca α-subunit (1:300; α_{1,098–1,196}; Chemicon International, Temecula, CA) and regulatory β-subunits using commercial antisera β₁: (1:500; Calbiochem, San Diego, CA), β₂ (1:500; Calbiochem) and β₄ (1:200; Alomone Labs, Jerusalem, Israel). After 1 h incubation with anti-rabbit immunoglobulin G conjugated with horseradish peroxidase (1:2,000), immunoreactive protein was visualized by chemiluminescence. Samples were compared by scanning densitometry using arbitrary units. Ovine cerebellum was used as a positive internal control and nonimmune rabbit serum to show antibody specificity.

Placental estrogen synthesis increases during pregnancy, resulting in elevated circulating estrogen levels (4, 5). In nonpregnant ewes, daily E₂ increases basal uterine blood flow, vascular synthesis of NO and cGMP, and BK_Ca activation (26, 43, 47). This is an important pathway for BK_Ca activation, which includes soluble guanylyl cyclase (sGC), cGMP, and cGMP-dependent protein kinase (PKG_1α) (13, 30, 33, 44, 56, 58). Thus we sought to determine whether there were changes in UVSM sGC, PKG_1α, and PKG_1β expression during the ovine reproductive cycle. We analyzed sGC (G4405; Sigma, St. Louis, MO), PKG_1α (PK10; Calbiochem), and PKG_1β (a gift from Dr. H. K. Surks, Tufts-New England Medical Center, Boston, MA) by immunoblot analysis using 1:1,000 or 1:750 dilution of specific antisera. Equal amounts of soluble protein (30 μg) were loaded on 10% polyacrylamide minigels, subjected to SDS-PAGE, and proteins were electrophoretically transferred to >nitrocellulose paper as described above.

cGMP assay.

Although uterine cGMP synthesis increases in pregnancy (39), it is unclear where the synthesis occurs and whether it is related to increases in UVSM sGC. Thus we measured UVSM cGMP contents by radioimmunoassay as previously described (48). Arteries were collected and frozen at −80°C. At the time of assay, samples were thawed, homogenized in 1 ml of 6% trichloroacetic acid, and spun at 2,500 g at 4°C for 15 min, and the supernatant was removed and assayed using the commercial radioimmunoassay (NEX-125; Perking Elmer, Boston, MA). Intra-assay variability was 1.3%, and inter-assay variability was 17%. The data are presented as picomoles per milligrams wet weight.

RT-PCR.

To examine the regulation of BK_Ca α- and β₁-subunit expression in UVSM, we examined their respective mRNA using semiquantitative RT-PCR as previously reported (26). Denuded arterial samples were thawed in TRIzol Reagent (Invitrogen Life Technologies, Gaithersburg, MD) and homogenized, and total cellular RNA was extracted and resuspended in 20 μl 0.1% diethyl pyrocarbonate in water. The concentration and purity were measured at 260/280 nm optical density. RT was performed with 2 μg total RNA as previously described (39). The reaction was incubated at room temperature for 10 min, 37°C for 1 h, and terminated at 95°C for 5 min. PCR was performed on 2.5 μl of RT product (cDNA) with specific primers designed from nucleotide sequences retrieved from the GenBank (synthesized by Invitrogen Life Technologies). Malate dehydrogenase (MDH) was unchanged during pregnancy (P > 0.1, ANOVA) and was chosen as the reference gene to perform semiquantitative RT-PCR, normalizing values for BK_Ca α- and β₁-subunit mRNA to MDH. The primers for MDH were: forward 5′-TCCCAGCAGCAACGGGTGT-3′ and reverse 5′-AAATCTTCGGGGTGACAACC-3′. BK_Ca α-subunit primers were: forward 5′-CAGCATTTGCCGTCAGTGTCCT-3′ and reverse 5′-CATGCCTTTGGGTTATTTTTCC-3′; β₁-subunit primers were: forward 5′-TCACCTACTACATCCTGGTCACGA-3′ and reverse 5′-GGAATTTGGCTCTGACCTTCTCCA-3′. Conditions for the PCR reaction for the α- and β₁-subunits were optimized, and analyses were performed as previously reported (39). To compare values, the targeted PCR products were run on the same gel.

Immunohistochemistry.

Intact segments of first and second generation uterine arteries were immediately fixed in 4% paraformaldehyde for 6 h at room temperature and embedded in paraffin as previously described (26, 48). Sections were mounted on slides, deparaffinized, hydrated, incubated with avidin-biotin blocking agent for 30 min, and incubated overnight at room temperature with polyclonal antibodies to the BK_Ca α- or β₁-subunits (1:30). After endogenous peroxidases were quenched with 3% H₂O₂ in H₂O for 30 min, immunostaining was detected with standard streptavidin-biotin-horseradish peroxidase and hematoxylin counter staining.

Statistical analyses.

Data were analyzed using linear and nonlinear regression analyses to examine overall changes across the reproductive periods and ANOVA to examine where within the time periods, e.g., pregnancy, changes occurred, using pairwise multiple comparison procedures. Student's t-test was used where appropriate. Data are reported as means ± SE.

RESULTS

BK_Ca α-subunit.

More than one isoform of the BK_Ca α-subunit has been reported, suggesting tissue-dependent expression of splice variants (16, 17, 26). We determined whether pregnancy alters the α-subunit species in UVSM and validated this by performing preabsorption studies with the peptide α_{1,098–1,196}, which is a 39-kDa-specific protein moiety derived from the α-subunit and serves as a control for the immunoblots and the preabsorption studies. The predominant protein bands in nonpregnant uterine arteries were at 83–87 and 105 kDa with a minor band at 100 kDa (Figs. 1A and 2A); these bands were absent after preincubation (Fig. 1B). Notably, the 83–87 and-2 105-kDa protein species were not present during pregnancy, and the 100-kDa protein predominated (Figs. 1A and 2A) until >30 days postpartum, when the protein species were identical to those in nonpregnant UVSM (Fig. 2A). Again, all bands were absent after preabsorption (Fig. 1B). The sheep cerebellum had two bands at ∼100 kDa, resembling that seen in human, ovine, and mouse myometrium (26), which were also absent after preabsorption (Fig. 1B, lane 6). Importantly, the 39-kDa α-subunit peptide α_{1,098–1,196}, which as noted above serves as a control, was also absent after preabsorption (Fig. 1B). Thus we used the 100-kDa band to determine changes in channel density and α-subunit expression during ovine pregnancy and to make comparisons before and after pregnancy.

Fig. 1.

Preabsorption studies of the protein variants of the large conductance, Ca²⁺-activated K⁺ channels (BK_Ca) α-subunit in first and second generation uterine artery smooth muscle and the effects of pregnancy. Each lane was loaded with 20 μg of soluble protein (see methods). Lane representation is as follows: lane 1, nonpregnant; lanes 2 and 3, pregnant at 97 and 130 days gestation, respectively; lane 4, 25 days postpartum; lane 5, ovine cerebellum; and lane 6, α-subunit control peptide (α_{1,098–1,196}), 50 μg. A: Western immunoblot without preabsorption. B: after preabsorption with the competing peptide. The 39-kDa protein in lane 6 was a positive control.

Fig. 2.

Representative immunoblots comparing the protein variants of BK_Ca α- and β₁-subunits expressed in first and second generation uterine artery smooth muscle from nonpregnant (n = 3), pregnant (n = 6), and postpartum (n = 4) sheep. A: immunoblot analysis for subunit protein. B: analysis of densitometry in arbitrary units for the 100-kDa variant of the α-subunit (B.1) and 39-kDa β₁ protein (B.2) during different reproductive periods. Values are means ± SE. Different letters represent differences between groups at P ≤ 0.05 by ANOVA; comparisons are on a single gel.

Comparisons of α-subunit expression between groups can only be accurately assessed on a single immunoblot. When UVSM protein from nonpregnant, pregnant, and postpartum ewes were examined on a single immunoblot, the 100-kDa protein in pregnant UVSM increased 64–81% compared with nonpregnant levels (P < 0.05, ANOVA; Fig. 2) but did not differ from arteries obtained at 2–17 days postpartum. To determine whether changes in α-subunit protein occurred during pregnancy, we performed a single immunoblot with samples obtained throughout the last two-thirds of pregnancy, i.e., 50–146 days (Fig. 3). There was no change in α-subunit protein during this time (R² = 0.01; P = 0.8; n = 14).

Fig. 3.

Expression of the 100-kDa variant of the α-subunit and 39-kDa β₁-subunit in first and second generation uterine artery smooth muscle in the last two-thirds of ovine pregnancy. A: immunoblot analysis for subunit protein. B: analysis of densitometry in arbitrary units for the 100-kDa α-subunit variant (B.1) and β₁-subunit (B.2) during gestation; n = 3 to 4 for each time period. Values are means ± SE. For the α-subunit, P > 0.1 by ANOVA; regression analysis: R = 0.01, P = 0.9, n = 14. For the β₁-subunit, different letters represent significant differences between groups at P < 0.001 by ANOVA.

BK_Ca β-subunit.

Four regulatory β-subunits are associated with the BK_Ca (16, 17, 55); however, they have not been systematically examined in all tissues and have not been studied during pregnancy. The β₄-subunit was not detected in UVSM from any reproductive period. The β₂-subunit was present in UVSM from all reproductive groups, but expression was minimal and unchanged (P > 0.1, data not shown). The β₁-subunit was expressed as a single protein species at 39 kDa and was also present in UVSM from all reproductive groups (Figs. 2 and 3). The β₁-subunit was first analyzed in samples from the three reproductive groups on a single immunoblot as described above. β₁-protein increased >50% from nonpregnant to pregnant sheep, paralleling the rise in the α-subunit, but did not differ between pregnant and postpartum animals at 2–34 days (P < 0.001, ANOVA; Fig. 2). However, when the UVSM samples analyzed for the α-subunit during the last two-thirds of pregnancy were probed for β₁ on a single immunoblot, β₁-protein progressively rose between 50 and 146 days gestation (P < 0.001, ANOVA; Fig. 3B2). Notably, this rise was linear over the last two thirds of gestation (R² = 0.996; P < 0.001; n = 13), and protein expression increased approximately twofold by term.

RT-PCR.

Using the same uterine artery samples analyzed for protein expression, we performed semiquantitative RT-PCR (see methods) across reproductive periods. There was no significant change in α- or β₁-subunit mRNA before, during, or after pregnancy (Fig. 4B) when analyzed by regression analysis (R = 0.1; P ≥ 0.05; n = 23) or ANOVA (P > 0.1; Fig. 4, B1 and B2) after separating into groups of nonpregnant, term pregnant >135 days of gestation, and postpartum ≤30 days.

Fig. 4.

Semiquantitative RT-PCR for α- and β₁-subunit mRNA in first and second generation uterine artery smooth muscle from nonpregnant (n = 4), pregnant (n = 6), and postpartum (n = 4) sheep. A: results of RT-PCR for the α- and β₁-subunits and the reference gene malate dehydrogenase (MDH), which was unaffected by pregnancy (P > 0.1). B: densitometry of α- (B.1) and β₁- (B.2) subunit mRNA in arbitrary units expressed as the ratio with MDH. There was no change in either subunit during the reproductive cycle by regression analysis (R = 0.1, P ≥ 0.05, n = 23) or ANOVA (P > 0.1) for multiple groups.

Immunohistochemistry.

To determine the site(s) of channel expression within the uterine artery wall and estimate relative channel density, we performed immunohistology for the α- and β₁-subunits at 56, 78, and 146 days of gestation. There was no endothelial immunostaining for either subunit (Fig. 5). However, medial immunostaining for both subunits was evident throughout pregnancy, and the apparent intensity of immunostaining for both subunits increased over time; however, the immunostaining appears to be substantially greater and more diffuse for the β₁-subunit at term.

Fig. 5.

Immunohistochemistry of first generation uterine arteries for BK_Ca α- and β₁-subunit expression at 56 (A and D), 78 (B and E), and 146 (C and F) days of ovine gestation, respectively, at ×20 magnification.

cGMP-dependent signaling.

Samples of denuded uterine artery obtained from each reproductive period were assayed for sGC, cGMP, PKG_1α, and PKG_1β protein. Although the levels of sGC protein were unchanged during the reproductive periods (R = 0.08; P = 0.8; n = 14; not shown), cGMP contents, expressed as a percentage of nonpregnant levels, gradually rose, with a transition at 86 days and an increase of >80% at term (P = 0.017, ANOVA; Fig. 6). PKG_1β was present in all samples (Fig. 7A) but unchanged throughout the reproductive periods (P > 0.1; data not shown). In contrast, PKG_1α protein increased nearly twofold by term (R = 0.86; n = 11; P = 0.002; Fig. 7B1) and decreased gradually after delivery, returning to nonpregnant levels by 30 days postpartum (R = 0.83; P = 0.008; n = 9). This was quantitatively analyzed by performing immunoblots with 5 to 6 samples each from nonpregnant, near-term pregnant, and 30–37 days postpartum ewes. There was a nearly sixfold rise in PKG_1α protein by term gestation (P < 0.001; Fig. 7B2), followed by a 74% fall at 1 mo postpartum (P < 0.001).

Fig. 6.

The change in cGMP contents in first and second generation uterine artery smooth muscle during ovine pregnancy. Values are means ± SE expressed as levels relative to the average measured in 4 nonpregnant (NP) sheep, which serves as 100%. Different letters signify significant differences between groups at P < 0.017 by ANOVA. d, Day.

Fig. 7.

Effects of pregnancy and parturition on cGMP-dependent protein kinase 1α (PKG_1α) and PKG_1β expression in uterine artery vascular smooth muscle. A: immunoblot analysis for PKG_1α and PKG_1β protein. B: densitometry in arbitrary units for PKG_1α protein. PKG_1β was unchanged (P > 0.1, ANOVA). B1: pattern of PKG_1α changed across reproductive periods, rising during pregnancy (R² = 0.95, P < 0.001, n = 6) and falling during the puerperium (R² = 0.83, P = 0.008, n = 9), achieving nonpregnant (NP) levels ≥30 days postpartum. B2: comparison of nonpregnant (n = 5), term pregnant (n = 6), and postpartum (n = 5) ewes at ∼30 days. *P < 0.001 comparing nonpregnant with term; **P < 0.001 for term vs. postpartum (t-test). Values are means ± SE.

Channel function.

The role of BK_Ca in uterine artery function over the course of pregnancy is unknown. Since it may modulate both contraction and relaxation responses (16, 17, 27, 49, 51), we performed contraction-relaxation studies in segments of first generation denuded uterine artery rings at ∼55 days and ∼145 days gestation (term, ∼150 days), examining the effect of BK_Ca inhibition with TEA at ≤1.0 mM, doses that confer channel specificity (27). At both times in pregnancy, there were dose-dependent vasoconstrictor responses to KCl and PE (P < 0.001, ANOVA; data not shown); however, the control responses to 10⁻⁶ M PE at term exceeded those seen at 55 days, stresses increasing an additional 3.6-fold by term pregnancy (Fig. 8; P = 0.004). Although channel inhibition with 0.2, 0.5, and 1.0 mM TEA increased uterine artery responses to PE at both gestational ages (P < 0.05), there was no dose effect (P = 0.63). Thus we combined the data for further analyses. Following TEA exposure, PE responses were enhanced ∼50% at 55 days and ∼26% at term (P < 0.001). Although it appears that there might have been a time effect of channel inhibition on PE-induced contractions, this was not significant due to the variability in responses.

Fig. 8.

Comparison of the effects of BK_Ca inhibition with tetraethylammonium chloride (TEA) on phenylephrine (PE)-induced contractions at 55 days (d; A) and term (B) gestation. *P < 0.001 compared with control by t-test; **P = 0.004 compared with control at 55 days gestation by t-test. Values are means ± SE. See Channel function for description of the analyses.

Knowing that NO might play a role in regulating UPBF, we then examined the relaxation responses of denuded uterine arteries to SNP, a NO donor, at both time points in pregnancy. In contrast with the contraction responses, there was no time- or gestational age-dependent difference in the control relaxation responses to SNP, stresses decreasing 47 ± 7% and 66 ± 8% (P = 0.7) in uterine arteries precontracted with 10⁻⁶ M PE at 55 days and term, respectively (Fig. 9). BK_Ca inhibition with TEA did not have a dose effect (P = 0.47) on NO-mediated relaxation at 55 days (Fig. 9A); however, TEA reduced relaxation responses ∼50%, 47 ± 7% vs. 24 ± 4% (P = 0.002, t-test). In contrast, at term, when uterine vasodilation is near maximum (36, 41), channel inhibition with TEA dose dependently (P < 0.001, ANOVA) reduced the relaxation responses 41% and 65% with 0.5 and 1.0 mM TEA, respectively.

Fig. 9.

Comparison of the effects of BK_Ca inhibition with TEA on nitroprusside-induced relaxation (10⁻⁷ M) after precontraction with 10⁻⁶ M PE at 55 days (d; A) and term (B) gestation. *P = 0.002 compared with control (C) by t-test. Different letters for groups in B signify significant differences at P < 0.001 by ANOVA. Values are means ± SE. See Channel function for description of the analyses.

DISCUSSION

Normal pregnancy is characterized by a marked rise in UPBF, levels increasing greater than thirtyfold by term gestation (34, 36, 41). The major component of this rise occurs in the last two-thirds of pregnancy and is primarily mediated by vasodilation (2, 36, 41, 52), but the underlying mechanisms are unclear. Recent evidence suggests several vasodilators contribute to the rise in UPBF, e.g., estrogen and NO, and may act through signaling pathways that activate BK_Ca (17, 38, 43, 57, 58). In the present study, we examined BK_Ca subunit expression and observed pregnancy-related modifications in the α-subunit resulting in a predominant 100-kDa isoform, the expression of which increases during placentation, but not thereafter. We also observed that the β₁-subunit is the predominant regulatory subunit in UVSM, and its expression also increases during placentation; however, unlike the α-subunit, expression increases further during the last two-thirds of pregnancy, paralleling the rise in UPBF and altering channel stoichiometry. This was associated with increases in UVSM cGMP and cGMP-dependent kinase, PKG_1α, as well as evidence that BK_Ca contribute to uterine artery contraction-relaxation responses. Notably, these changes are reversed at ≥30 days postpartum. These are the first data characterizing uterine artery BK_Ca expression during pregnancy and provide evidence of parallel enhancement of the cGMP-dependent pathway that might serve as a signaling pathway for progressive channel activation. Thus we present new insights into the role of BK_Ca in the rise in UPBF during pregnancy; however, the initiator of this signaling pathway has yet to be determined.

BK_Ca consist of four α-subunits encoded by the Slo gene that form the channel pore (16, 17, 49, 51). Several α-subunit isoforms have been reported and are believed to represent posttranslational modifications (16, 26, 49). However, this may reflect the techniques used for tissue collection and/or preparation for analysis. In the present study, tissues from all reproductive groups were similarly collected and prepared for study, and importantly, they were examined on the same immunoblot. There were three α-subunit species in nonpregnant UVSM, 83, 100, and 105 kDa, confirming an earlier report (26). During pregnancy, there was an absence of the 83-kDa protein and marked decrease in the 105-kDa protein, both reappearing ≥30 days after delivery, suggesting that some agent peculiar to pregnancy was involved. The preabsorption assays suggest that each isoform is likely derived from the α-subunit, again confirming earlier findings (26), and that they may represent unique posttranslational modifications of the α-subunit during ovine pregnancy. Alternatively, there may be selective increases in clearance or repression of the 83- and 105-kDa protein variants. Although the 100-kDa protein rose during placentation, this does not appear to equal the fall in the other two proteins, suggesting total channel density may actually fall in pregnancy. It is unclear what triggers the change in α-subunit, but estrogen is an unlikely candidate (26). It also is unclear whether this is unique to UVSM, since we did not study peripheral VSM. Notably, only the 83-kDa protein is seen in UVSM from older nonpregnant women (45). This could reflect age differences, the absence of estrogen, or changes unique to sheep. It is important to determine how these changes in the α-subunit occur and alter BK_Ca function.

The BK_Ca β-regulatory subunits are derived from separate gene products. β₁ and β₂ are associated with smooth muscle, whereas the others are primarily in nervous tissue (16, 17, 49, 55), explaining the absence of β₄ in UVSM. β₂ was present in ovine UVSM, but the levels were low and unchanged throughout the reproductive cycle. Interestingly, its expression is elevated in UVSM from older nonpregnant women, consistent with an age-related expression pattern (28, 45). Only a single species of the 39-kDa β₁-subunit protein was seen, and it was highly expressed in each reproductive period. At the end of placentation, when UPBF has increased approximately fourfold, the relative increases in α- and β₁-subunit are similar; thus β₁:α stoichiometry may be unchanged. Unlike the α-subunit, β₁-expression increased during the last two-thirds of gestation, paralleling the rise in UPBF. Although channel density may have been unchanged at this time, the stoichiometry is markedly altered, increasing the β₁-to-α ratio. This difference in subunit expression pattern was evident in the immunohistochemistry by disproportionate changes in immunostaining. Our data for β₁ mRNA suggest the rise in expression is posttranslational; however, the mechanisms responsible for the differences in subunit expression can only be speculated on at present. Since UPBF increases approximately fourfold during placentation (41), the proportionate rise in subunit protein in the proximal uterine arteries might be flow dependent (17). This, however, does not explain why the α-subunit is unchanged, whereas UPBF increases another threefold and the β₁-subunit continues to increase. The latter may be due to increases in placental estrogen synthesis that begin with placentation and continue throughout the remainder of pregnancy (4, 5). This is supported by the finding that E₂ selectively increases β₁ expression in uterine arteries from nonpregnant ewes without altering the α-subunit (26) and evidence that placental estrogen is directly transferred to the uterine artery during pregnancy (18). These findings need further study.

The parallel rise in UPBF and β₁-subunit permit enticing speculation about the regulation of UPBF in the last two-thirds of ovine pregnancy. Exogenous and endogenous estrogens are potent uterine vasodilators in nonpregnant and pregnant ewes (21, 40, 46, 47). This is mediated by increasing uterine artery NOS expression and activation, NO and cGMP synthesis, and BK_Ca activation (7, 23, 39, 44, 48, 50, 54, 57). Furthermore, estrogens activate BK_Ca directly via the β₁-subunit (9, 17, 53). Thus one thesis is that placental estrogen accounts for the altered β₁:α-subunit stoichiometry, which enhances channel activation and potentiates the direct effects of estrogen, resulting in BK_Ca-mediated vasodilation; conformation requires electrophysiological studies. Alternatively, placental estrogen increases uterine artery NOS, augmenting the NO-cGMP-PKG pathway, which increases BK_Ca phosphorylation and activation, thereby increasing UPBF (13, 30, 33, 55). Although UVSM sGC was unchanged, cGMP contents rose in the last two-thirds of pregnancy, suggesting enzyme activity increased (15). This was associated with increases in PKG_1α protein by term, which could be activated by cGMP and in turn phosphorylate BK_Ca, accounting for the rise in UPBF that parallels the rise in the β₁-subunit (13, 30, 33). PKG_1β was present but unchanged. The presence of both PKG isozymes is not unusual (13, 56), but it is notable that PKG_1α is activated by levels of cGMP tenfold less than that required for PKG_1β (6). Thus the rise in PKG_1α may amplify BK_Ca activation during pregnancy (30). A third possibility is that both pathways are upregulated, representing a redundancy that insures the rise in UPBF during the period of exponential fetal growth. Irrespective of the signaling pathways involved, local BK_Ca inhibition with TEA decreases UPBF 70–80% in the last third of ovine pregnancy, supporting the thesis that the BK_Ca is a major regulator of basal UPBF (43). The challenge is to determine which pathway can be manipulated to develop therapeutic strategies to reverse decreases in UPBF seen in abnormal pregnancies, e.g., maternal hypertension (31, 35).

BK_Ca modify and contribute to contraction-relaxation responses in VSM (16, 17, 27). This was examined in uterine artery rings at 55 days and term. There was a gestational age-related increase in PE responses consistent with earlier studies that suggest that α-stimulation is an essential mechanism for maintaining UVSM tone in pregnancy (1, 8, 37). Importantly, BK_Ca inhibition with TEA enhanced the contraction responses to PE, suggesting BK_Ca modify adrenergic responses and may contribute to the attenuated uterine vascular adrenergic responses seen in normal pregnant women and sheep (36, 43) and possibly the myogenic responses observed by Cox et al. (8). BK_Ca inhibition also decreased relaxation responses to NO at both ages, but at term it was dose dependent, suggesting an age-related increased role in NO-mediated uterine vasodilation. These findings are strikingly similar to those recently seen in vascular rings from nonpregnant women (45) and in intact pregnant sheep (42).

In the present study we characterized UVSM BK_Ca expression during the ovine reproductive cycle, identifying alterations in the pore-forming α-subunit that persisted throughout pregnancy and disappeared when ewes began to cycle. β₁-subunit expression increased, altering channel stoichiometry and possibly enhancing BK_Ca resting potential, activation, and function and contributing to the regulation and rise in UPBF. Although the NO-cGMP-PKG_1α pathway was upregulated, its contribution is unclear, as are other proximal signaling events. It is possible placental estrogen modifies channel expression and activation. The potential for pharmacological intervention in the presence of a decrease in UPBF has not previously existed but may be in the horizon if a similar system exists in women.

GRANTS

These studies were supported by National Institute of Child Health and Human Development Grant HD-008783-33 (awarded to C. R. Rosenfeld) and the George L. MacGregor Professorship.