INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

IT IS DOCUMENTED THAT arterial blood pressure and peripheral resistance decrease during pregnancy in both humans and rats, whereas blood flow, in many vascular beds, and cardiac output are increased (25). Vascular resistance to many exogenous vasoactive substances accompanies these hemodynamic alterations. Decreased responses to vasoconstrictors are also observed in isolated blood vessels, indicating that the mechanisms responsible are not dependent on cardiovascular reflex pathways (6, 18). It has been shown that the decreased responsiveness of blood vessels to vasopressor agents is not dependent on a downregulation of their receptors or on an increased release of endogenous vasodilators (5, 36). It has been shown that endothelium-dependent relaxation through nitric oxide could be involved in the blunted responses to vasoconstrictors that accompany pregnancy (25). However, in isolated blood vessels denuded of their endothelium, a decreased reactivity to vasopressors still persists when obtained from pregnant animals (29, 36).

It was previously shown that Ca²⁺ influx through voltage-operated calcium channels was functionally impaired in aortic rings and mesenteric resistance vessels of pregnant compared with nonpregnant rats (29, 35). This supported the results previously reported by Ezimokhai et al. (10). In the cited studies, we have demonstrated that BAY K 8644, a dihydropyridine calcium channel activator, induced by itself concentration-dependent contractions in aortic rings of nonpregnant rats, contractions that were almost absent in aorta of pregnant rats. This differential response to BAY K 8644 was overcome by preincubating aortic rings from nonpregnant and pregnant rats in different low concentrations of KCl (5 and 10 mM, respectively). These results suggested that the blunted responses of blood vessels of pregnant rats could be dependent on altered membrane potential. Indeed, Meyer et al. (20) reported that smooth muscle cells of mesenteric resistance arteries of pregnant rats were hyperpolarized by 7 mV compared with nonpregnant rats. It is considered that pregnancy, via a yet unidentified factor or mechanism, induces an alteration of resting membrane potential and/or an alteration of the mechanism implicated in the depolarization of the cell membrane upon stimulation by vasoconstrictors.

Membrane potential is largely regulated by its K⁺ permeability, and the relative contribution of the different K⁺ channels differs from one organ to another. It is thought that K⁺ channels would be the most important channels in regulating the membrane potential in smooth muscle (2). When open, they maintain or help to recover a basal state of polarization and decrease the depolarizing effects produced by Ca²⁺ or Na⁺ influx or by Cl efflux. This effect seems to affect particularly the voltage-gated ion channels (7, 27).

K⁺ channels are selective to K⁺ and are ubiquitous. They are found in pancreatic -cells, heart, and vascular smooth muscle (4, 22, 27). It is known that K⁺ channels are implicated in many physiological mechanisms such as regulation of vascular tone, cellular excitability, and hormone secretion (11, 19). K⁺ channels contain at least five well-characterized members. Among them, two are of particular interest in the regulation of myotropic responses in vascular smooth muscles, the ATP-sensitive K⁺ channels, commonly referred to as K_ATP channels, and the high-conductance calcium-activated K⁺ channels, or BK_Ca channels (1-3, 21).

In most organs, except pancreas, K_ATP channels are inhibited by physiological ATP concentrations (11, 37), and they are voltage and Ca²⁺ independent (9, 19, 28). They are blocked by sulfonylureas such as glibenclamide and opened by K⁺ channel openers such as cromakalim, which belongs to the benzopyran family (11). K_ATP channel openers have been shown to hyperpolarize myocytes of rat aorta by increasing the open probability of the channel (2, 12). Those channels are in major part regulated by the ADP-to-ATP ratio (38).

BK_Ca channels are Ca²⁺ and voltage dependent, and they show the highest conductance for K⁺ (11). They produce a negative feedback to cell membrane depolarization caused by a Ca²⁺ influx (2, 3). BK_Ca channels are targets for many vasoactive substances and are thought to be implicated in the long-term regulation of blood flow and arterial blood pressure (31). They are blocked by peptides such as iberiotoxin, which blocks BK_Ca channels in a reversible bimolecular reaction, and opened by benzimidazols such as NS-1619 (11). Tetraethylammonium (TEA) is an agent with the potency to block both types of channels described above as well as inward rectifier potassium (K_V) channels (21).

K⁺ channels are major players in determining membrane potential. Because voltage-operated Ca²⁺ channels are dependent on this potential for their activity, we tested the effects of modulators of K⁺ channels on myotropic responses of aortic rings of nonpregnant and pregnant rats to vasoconstrictors such as phenylephrine (Phe), arginine vasopressin (AVP), and KCl.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Animals. Female Sprague-Dawley rats (Charles River Canada, St-Constant, Québec, Canada) aged 10-11 wk were mated with males. The morning on which spermatozoa were found in vaginal smears was labeled day 1 of pregnancy. The pregnant females were placed in individual cages until used on the eve of parturition (22nd day of gestation). Nonpregnant rats of the same age served as controls without considering the stage of the estrous cycle. The animals were housed in facilities of the Research Center at Sainte-Justine Hospital, which is accredited by the Canadian Council on Animal Care. The protocol was approved by the local animal care committee.

Organ bath assay. After decapitation, the thoracic aorta was rapidly removed and placed in cold Krebs bicarbonate solution (KBS). It was cleaned of fat and extraneous tissues and cut into four consecutive rings (2-3 mm) that were placed in individual jacketed organ baths (15 ml; Radnoti Glass, Monrovia, CA) maintained at 37°C. The endothelium of each ring was gently removed by rubbing the lumen with an 18-gauge needle. In each experiment, four rings of both nonpregnant and pregnant rats were used. They were equilibrated for 60 min under 2.0 g passive tension, the optimal tension for both groups of tissues (36), with frequent washing and tension adjustment. The tissues were bathed in KBS of the following composition (in mM): 118 NaCl, 4.65 KCl, 25 NaHCO₃, 2.5 CaCl₂, 1.18 MgSO₄, 1.18 KH₂PO₄, and 5.5 dextrose. The solution was bubbled with a mixture of 95% O₂-5% CO₂; pH was 7.4. After equilibration, the tissues were challenged with 1.0 µM Phe. At plateau response, carbachol (0.1 mM) was added to ensure removal of the endothelium. Tension was measured by force-displacement transducers and was recorded on a computerized data acquisition system using the Work Bench software (both from Kent Scientific, Litchfield, CT). The experiments with NS-1619 were performed under sodium lamps to prevent photodegradation of the substance.

Experimental protocol. In a first set of experiments, cumulative concentration-response curves to TEA (0.3-30 mM) were performed on aortic rings without and with preincubation with low concentrations of KCl (5 or 10 mM).

In the other experiments, cumulative concentration-response curves to Phe (10⁹ to 10⁴ M), AVP (10¹¹ to 10⁷ M), and KCl (2-100 mM, as hypertonic solution) were consecutively obtained on the same tissues. The second and third curves were chartered 1 h after the return to baseline. In each experiment, four aortic rings of both nonpregnant and pregnant rats were assayed simultaneously. One aortic ring of each rat served as a control, whereas each of the other three were preincubated 10 min before measuring the curve with a different concentration of cromakalim (0.3, 0.6, 3 µM), glibenclamide (0.1, 1, 10 µM), NS-1619 (3, 10, 30 µM), or iberiotoxin (10, 30, 100 nM). Each ring was exposed to only one concentration of the same inhibitor or activator throughout the experiment. The sequence of agonist stimulations was changed each day to avoid any effect of the previous curve on the subsequent ones.

Data analysis. Each concentration-response curve was analyzed by computer fitting to a four-parameter sigmoid curve using the Prism 2.01 Program (GraphPad, San Diego, CA) to evaluate the EC₅₀ and the maximum asymptote of the curve (E_max). Different curves of the same protocol were compared by two-way ANOVA on mean pD₂, the negative logarithm of the EC₅₀, and on mean E_max. Data are expressed as mean experimental points with their SE along the best curve fitted to these points.

Drugs and chemicals. All salts employed in these experiments were of analytical grade obtained from Fisher Scientific (Montréal, Québec, Canada). AVP was obtained from Peninsula Laboratories (Belmont, CA). Phe hydrochloride, carbamylcholine chloride (carbachol), TEA chloride, cromakalim, and iberiotoxin were purchased from Sigma Chemical (St. Louis, MO). Glibenclamide and NS-1619 were purchased from RBI (Natick, MA). Cromakalim was prepared in 70% ethanol; NS-1619 and glibenclamide were prepared in DMSO. Similar dilution of the vehicle was applied to aortic rings serving as control in each experiment.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Effects of TEA. Concentration-response curves to TEA were measured on aortic rings of nonpregnant rats, but a response was hardly obtained in pregnant animals (Fig. 1A). When 10 mM KCl was added in the bathing solution, a small increase in tone was obtained reaching 0.33 ± 0.15 and 0.06 ± 0.01 g in aorta of nonpregnant and pregnant rats, respectively. Under these conditions, TEA induced concentration-dependent contractions in both groups of aortic rings, but responses to TEA were significantly smaller in tissues of pregnant compared with nonpregnant rats (Fig. 1B). The contractile effects of TEA were similar for the two groups when the rings of nonpregnant and pregnant rats were preincubated with 5 (0.05 ± 0.02 g) and 10 (0.06 ± 0.01 g) mM KCl, respectively (Fig. 1C). This suggests that the direct effects of TEA on aortic rings is dependent on membrane potential since it can be modulated by low concentrations of depolarizing KCl.

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Concentration-response curves to tetraethylammonium (TEA) in aortic rings from nonpregnant and pregnant rats. Curves were obtained in the absence of KCl (A), in the presence of 10 mM KCl for both groups of rats (B), and in the presence of 5 and 10 mM KCl for the nonpregnant and pregnant rings, respectively (C). Curves are best fits to mean ± SE experimental points (n = 12). Ordinate depicts responses in tension (g), whereas abscissa represents the logarithm of TEA concentrations (mol/l).

Resistance to vasopressor. In the present experiments, when we plotted all curves together, vascular reactivity for the three agonists (Phe, AVP, and KCl) was reduced in aortic rings of pregnant rats. This is shown in Table 1 by decreases in pD₂ and E_max in aorta of the pregnant vs. nonpregnant rats, except for the decrease in pD₂ of Phe during pregnancy that did not reach statistical significance.

Effects of cromakalim. The effects of cromakalim (0.3, 0.6, and 3 µM) were measured on concentration-response curves to Phe, AVP, and KCl on aortic rings from nonpregnant and pregnant rats (Fig. 2). Cromakalim is known to be a potent K_ATP channel opener.

View larger version (26K): [in this window] [in a new window]   Fig. 2.   Concentration-response curves to phenylephrine (Phe; A and B), arginine vasopressin (AVP; C and D), and KCl (E and F) in aortic rings from nonpregnant (A, C, and E) and pregnant rats (B, D, and F). Curves were obtained in the absence () or in the presence of different concentrations of cromakalim (, , ). Curves are best fits to mean ± SE experimental points (n = 10). Ordinate depicts responses in tension (g), whereas abscissa represents the logarithm of agonist concentrations (mol/l).

Cromakalim antagonized the effects of Phe by reducing both E_max and sensitivity (pD₂) to this agonist (Fig. 2, A and B). The antagonistic effect of cromakalim was concentration dependent on the E_max but not on pD₂. In this last parameter, the K_ATP channel activator (0.3 µM) decreased pD₂ from 7.77 ± 0.11 to 6.89 ± 0.06 (P < 0.01) in rings of nonpregnant rats and from 7.75 ± 0.05 to 7.07 ± 0.10 (P < 0.01) in tissues of pregnant animals. With increasing concentrations of cromakalim (0.6 and 3 µM), pD₂ to Phe did not decrease further. However, the reduction of E_max by cromakalim was not statistically different in aorta of pregnant compared with nonpregnant rats, whereas the reduction of pD₂ was two times larger in rings of the latter group (P < 0.05, by 2-way ANOVA).

Cromakalim also reduced the maximum response to AVP (Fig. 2, C and D). This reduction of the E_max to AVP was again concentration dependent and was more important in aorta of nonpregnant than in pregnant rats (P < 0.05, 2-way ANOVA). pD₂ to AVP was also decreased by cromakalim in the rings of nonpregnant animals (from 8.81 ± 0.04 to 8.37 ± 0.08 at 0.3 µM, P < 0.01) but not in tissues of pregnant rats [from 8.60 ± 0.05 to 8.54 ± 0.03 at 0.3 µM, not significant (NS)]. With larger concentrations of cromakalim (0.6 and 3 µM), pD₂ did not decrease further in either group of aorta.

Cromakalim did not show any significant effect on the maximum response to KCl in either group of aorta (Fig. 2, E and F), but it significantly decreased pD₂ to the depolarizing agonist two times as much in aorta of nonpregnant than of pregnant rats (respectively, from 1.83 ± 0.03 to 1.50 ± 0.03 and from 1.74 ± 0.02 to 1.55 ± 0.04 at 0.3 µM cromakalim, P < 0.01). Again, higher concentrations of cromakalim (0.6 and 3 µM) did not induce a significant larger reduction of pD₂ to KCl.

These data show that cromakalim is more potent to block the effects (either E_max or pD₂) of the three agonists in aorta of nonpregnant than of pregnant rats. There was a consistent effect of the K_ATP channel opener on pD₂ to the three agonists, but this appears not to be concentration dependent or to have already reached a maximal effect at 0.3 µM cromakalim. These data also suggest that the involvement of these channels (K_ATP channels) in myotropic responses to the three agonists is quantitatively different.

Effects of glibenclamide. Glibenclamide (0.1, 1.0, and 10 µM), a K_ATP channel blocker, did not show any effect on concentration-response curves to the three vasoconstrictors on the nonpregnant aorta (Fig. 3, A, C, and E). In the rings from pregnant rats, it significantly increased the pD₂ of Phe and AVP from, respectively, 7.67 ± 0.06 to 7.89 ± 0.05 (P < 0.05) and 8.68 ± 0.05 to 8.90 ± 0.06 (P < 0.05) at 1.0 µM glibenclamide (Fig. 3, B and D). Curves obtained in the presence of 0.1 and 10 µM glibenclamide are not shown for clarity. Gestation was the only factor to influence the maximal response to all agents, and this effect represents a reduction of this parameter in the pregnant compared with the nonpregnant group (P < 0.01). These results suggest that inhibition of K_ATP channels has no effect on the myotropic responses of the aorta of nonpregnant rats. However, these channels could contribute to myotropic responses in the aorta of the pregnant rats.

View larger version (20K): [in this window] [in a new window]   Fig. 3.   Concentration-response curves to Phe (A and B), AVP (C and D), and KCl (E and F) in aortic rings from nonpregnant (A, C, and E) and pregnant rats (B, D, and F). Curves were obtained in the absence () or in the presence of 1.0 µM glibenclamide (). Curves are best fits to mean ± SE experimental points (n = 10). Ordinate depicts responses in tension (g), whereas abscissa represents the logarithm of agonist concentrations (mol/l). Curves obtained in the presence of 0.1 and 10 µM glibenclamide are not shown for clarity.

Effects of NS-1619. Figure 4 describes the effects of NS-1619 (3-30 µM), a BK_Ca channel opener. Figure 4, A and B, shows that NS-1619 decreased the pD₂ of Phe in a similar concentration-dependent manner in both groups of rats (from 7.95 ± 0.09 to 7.22 ± 0.09 and from 7.91 ± 0.06 to 7.15 ± 0.10 in nonpregnant and pregnant rats, respectively, at 30 µM, P < 0.01). Alternatively, it did not have any significant effect on the E_max (by 2-way ANOVA) despite an apparent effect at 30 µM.

View larger version (26K): [in this window] [in a new window]   Fig. 4.   Concentration-response curves to Phe (A and B), AVP (C and D), and KCl (E and F) in aortic rings from nonpregnant (A, C, and E) and pregnant rats (B, D, and F). Curves were obtained in the absence () or in the presence of different concentrations of NS-1619 (, , ). Curves are best fits to mean ± SE experimental points (n = 10). Ordinate depicts responses in tension (g), whereas abscissa represents the logarithm of agonist concentrations (mol/l).

Figure 4, C and D, shows that E_max to AVP was significantly decreased by NS-1619 in a concentration-dependent manner in both groups of rats (P < 0.01). This effect was more important at the higher concentration of drug used (30 µM) and was of greater importance in aortic rings of pregnant rats, at least with 10 µM of the activator. NS-1619 also decreased pD₂ to AVP, but this was only significant in aorta from nonpregnant compared with pregnant rats (at 30 µM, from 8.87 ± 0.05 to 8.56 ± 0.05, P < 0.01, and from 8.70 ± 0.04 to 8.64 ± 0.11, respectively, NS). Figure 4, E and F, depicts the effect of NS-1619 on KCl concentration-response curves. Both pD₂ and E_max were significantly decreased at the highest concentration of NS-1619 used, and this was equivalent in aorta of nonpregnant and pregnant rats.

Effects of iberiotoxin. Iberiotoxin (10-100 nM) is a potent BK_Ca channel blocker. Its effects were also measured on concentration-response curves to Phe, AVP, and KCl (Fig. 5). We observed that it had a more potent effect in aortic rings of nonpregnant than of pregnant rats for the three agonists. At the highest concentration used (100 nM), it significantly increased the E_max of Phe (Fig. 5A) in the nonpregnant group (P < 0.05) but not in the pregnant group of rats (Fig. 5B). Iberiotoxin also increased the pD₂ of Phe in both groups (from 7.77 ± 0.05 to 8.23 ± 0.04 and from 7.81 ± 0.04 to 8.10 ± 0.04, respectively, at 100 nM, P < 0.01).

View larger version (25K): [in this window] [in a new window]   Fig. 5.   Concentration-response curves to Phe (A and B), AVP (C and D), and KCl (E and F) in aortic rings from nonpregnant (A, C, and E) and pregnant rats (B, D, and F). Curves were obtained in the absence () or in the presence of different concentrations of iberiotoxin (, , ). Curves are best fits to mean ± SE experimental points (n = 10). Ordinate depicts responses in tension (g), whereas abscissa represents the logarithm of agonist concentrations (mol/l).

The pD₂ to AVP (Fig. 5, C and D) was increased in a concentration-dependent manner by iberiotoxin in the nonpregnant group (from 8.83 ± 0.03 to 9.31 ± 0.06 at 100 nM, P < 0.01). A significant increase in E_max to AVP was also observed with 100 nM iberiotoxin. The increase in the pD₂ of AVP of aortic rings from pregnant rats was of the same amplitude for all concentrations of iberiotoxin used (from 8.74 ± 0.03 to 8.85 ± 0.03 at 100 nM, P < 0.05). An increase in the pD₂ of the aorta of nonpregnant rats was also observed with KCl (from 1.77 ± 0.03 to 2.22 ± 0.14 at 100 nM, P < 0.05) but not in tissues of pregnant rats. E_max to KCl was not changed by iberiotoxin in either group of rats (Fig. 5, E and F). In Fig. 5, it is clearly shown that iberiotoxin showed more pronounced effects on aortic rings of nonpregnant than pregnant rats.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The present investigation was undertaken to determine the involvement of potassium channels of K_ATP and BK_Ca types in the blunted responses to vasoconstrictors that accompany normal pregnancy. Such an involvement of K⁺ channels was suggested by previous observations indicating that the blunted contractile response of aortic rings of pregnant rats to BAY K 8644, a voltage-operated calcium channel activator, can be overcome by differentially depolarizing the smooth muscle of rat aorta from nonpregnant and pregnant rats with low concentrations of KCl (29). Our results confirmed these findings with TEA, a nonselective K⁺ channel blocker, using a similar protocol (Fig. 1). The present results (Table 1) confirm the previously reported blunted responses to many vasoconstrictors during pregnancy (34) that were also observed in resistance-sized vessels (5, 35). Because aorta exhibit the same characteristics as those vessels (29, 35), we thus conclude that it is a reliable model to study the blunted reactivity to vasopressor agents during pregnancy. Our results also show that myotropic actions of Phe, AVP, and KCl were inhibited by activation of K_ATP and BK_Ca channels with cromakalim and NS-1619, respectively. The former activator was more potent than the latter in this respect and was also significantly more effective in aortic rings of nonpregnant than pregnant rats. We have also observed that K_ATP channel blockade (with glibenclamide) did not influence myotropic responses to the three vasopressors used on aortic rings from nonpregnant rats but slightly potentiated the sensitivity of pregnant aorta to Phe and AVP. Alternatively, BK_Ca channel blockade substantially potentiated the responses to the three agonists, potentiation that was much smaller in aortic rings of pregnant compared with nonpregnant rats. The present data suggested that, in unstimulated conditions, K_ATP channels are in an activated state, favoring hyperpolarization of vascular smooth muscle cells during pregnancy. However, BK_Ca channels appear to be more importantly involved in regulating myotropic responses of the smooth muscle cells, since their blockade potentiated vascular responses to vasoconstrictors more effectively than did the K_ATP blocker glibenclamide. Moreover, they also potentiated more efficiently vascular responses to vasopressors in tissues of nonpregnant than pregnant rats.

Effects of TEA. Our results with TEA, a nonselective K⁺ channel blocker, indicate that the aorta of pregnant rats is somewhat resistant to the inhibition of these channels compared with nonpregnant animals. However, this resistance can be overcome by preincubating the rings in different low concentrations (5-10 mM) of depolarizing KCl. These results are similar to the ones previously reported by our laboratory with an activator of voltage-operated calcium channels, BAY K 8644 (29). Using aortic segments of spontaneously hypertensive (SHR) and coarcted Sprague-Dawley rats, Rusch et al. (30) observed a potentiation of the direct contractile effects of TEA linked with the high blood pressure. They did not however observe any contraction in response to TEA (up to 10 mM) in aorta of normotensive Wistar-Kyoto (WKY) rats and in the abdominal aorta (not exposed to high blood pressure) of coarcted hypertensive rats. Later, Rusch and Runnells (32) reported that this enhanced effect of TEA in aorta of hypertensive animals is linked to an increased density of maxi-K⁺ (BK_Ca) currents. These data are compatible with the results obtained in the present investigation in which opposite reduction of this contractile effect of TEA was observed. This suggests that analogous reduction of the TEA-sensitive K⁺ channels accompanies pregnancy, in which a reduction of the blood pressure is observed (33).

The mechanisms involved in the decreased responses to TEA in aorta of pregnant compared with nonpregnant rats are still unknown. Because TEA can induce contraction in aortic rings of nonpregnant animals but not in pregnant ones, we suggest that K⁺ channels are indeed involved in hyperpolarization of vascular smooth muscle during pregnancy, as reported in the mesenteric resistance arteries (20). The similar results observed with different concentrations of KCl, 5 and 10 mM for aortic rings from nonpregnant and pregnant rats, respectively (Fig. 1C), provide more evidence for a difference in membrane potential.

K_ATP channels. Dumont and Lamontagne (7) showed that concentration-response curves to AVP on thoracic aortic rings of male rats were more affected by lemakalim, a K_ATP channel opener, than the ones to Phe. Our experiments with cromakalim showed similar results on aortic rings of both nonpregnant and pregnant rats. We demonstrated that cromakalim (0.3-3 µM), another K_ATP channel opener, had an effect on both the pD₂ and E_max of Phe and AVP, with a greater effect on the latter. It was proposed that cromakalim, by opening K_ATP channels, hyperpolarizes the membrane, decreases the open probability of L- and T-type calcium channels, and inhibits inositol triphosphate formation (27). The different magnitude of the effects of cromakalim on responses to Phe and AVP suggests that the coupling mechanisms for these two vasoconstrictors in vascular smooth muscles are somewhat different and provides strong evidence that Phe is more dependent on intracellular calcium to induce contraction than would be AVP. Cromakalim was not able to affect the E_max of KCl (Fig. 2, E and F), but Quast (26) showed that the effects of K_ATP channel opening cannot be observed if the concentration of KCl used to depolarize the membrane is higher than 50 mM. Indeed we observed that, at 35 mM KCl, cromakalim is able to decrease the maximal response in a concentration-dependent manner in both groups of rats (data not shown).

BK_Ca channels. BK_Ca channels have been shown to be present in most types of smooth muscle, and their dual modulation by membrane potential and intracellular Ca²⁺ is compatible with a postulated role in the regulation of myogenic tone. NS-1619 is a potent activator of these channels, although its selectivity, in respect to inhibition of dihydropyridine-sensitive calcium channels, has been challenged (8, 13). In tracheal smooth muscle cells in culture, Macmillan et al. (17) showed that NS-1619 is a potent activator of BK_Ca channels. In the experiments reported here, NS-1619 reduced the maximum response to the three agonists in a concentration-dependent manner. This was particularly important with the highest concentration of NS-1619 used (30 µM). However, the activator showed more important inhibition on aortic rings of pregnant than nonpregnant rats, in opposition to what we observed with cromakalim. This suggests a different involvement of BK_Ca compared with K_ATP channels in the regulation of myotropic responses of vascular smooth muscle, since activation of each type of channel gave opposite relative responses in the two groups of rats. This is particularly true for KCl responses for which cromakalim reduced importantly the responses to small concentrations of depolarizing agents (Fig. 2) while NS-1619 markedly affected maximum responses (Fig. 4). This suggests that, with the former activator, hyperpolarization had already taken place, and larger concentrations of KCl were needed to induce the required depolarization. Activation of BK_Ca with NS-1619 appeared to recruit some counteracting mechanisms that oppose the developing contraction induced by depolarization.