INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

THE ENDOTHELIUM INFLUENCES the contractile function of vascular smooth muscle (VSM) via the release of a variety of vasoconstrictor and vasodilator agents [prostaglandins, thromboxanes, reduced nitric oxide (NO'), endothelium-derived hyperpolarizing and contracting factors (EDHF, EDCF)]. Changes in the basal production of NO', in particular, can alter smooth muscle responsiveness to vasoconstrictors (7, 15, 20). In addition, increasing quantities of NO' may be produced in response to vasoconstrictor agents, thus providing a brake on the process of contraction (7a). We have some insight into the cellular mechanisms by which NO' acts on membrane ion channels to produce smooth muscle relaxation. There is considerable evidence that NO' can activate K⁺ channels, resulting in membrane hyperpolarization, inhibition of voltage-dependent Ca²⁺ current, and smooth muscle relaxation (18). NO' may also inhibit Ca²⁺-channel activity via cGMP (5, 10, 14, 16). We propose that these mechanisms by which NO' produces relaxation may be supplemented by an ability to interfere with agonist-induced activation of Cl current, thus interfering with depolarization and preventing contraction.

We have presented evidence that agonist-induced contraction of vascular smooth muscle depends on the activation of a Cl current that has the permeability sequence I > Br > Cl > methanesulfonate (MS) (8). This Cl current may be activated by the release of stored intracellular Ca²⁺. It is proposed that the ensuing depolarization is essential for producing contraction-sustaining entry of Ca²⁺ through voltage-dependent Ca²⁺ channels. Inhibiting or enhancing the activation of depolarizing Cl current may be as important to regulation of VSM membrane potential as is the activation or inhibition of K⁺ current.

Because Cl conductance appears to be quite low at rest (3) and is activated to a large degree by agonists (2, 6, 17), it may be important to distinguish between those mechanisms that cause the relaxation of an established contraction and those that may prevent vasoconstriction. These may represent two quite distinct cellular processes. If Cl current is responsible for a portion of sustained agonist-induced depolarization, then, similar to the activation of K⁺ current, inhibition of Cl current may relax an established contractile response. Neither K⁺ channel activation nor Cl channel inhibition will have any effect, however, if the cell is not first depolarized by the agonist and contraction initiated. Interfering with this process may provide a potent mechanism for regulating vascular contractility. Understanding this distinction between vasodilatation and suppressed contractility will require an in-depth understanding of the cellular processes involved in VSM contraction, including the mechanisms regulating depolarization in response to an agonist.

The current studies were designed to determine whether the agonist-induced Cl current of rat aortic VSM is regulated by two endothelial-derived factors that inhibit contraction, cyclooxygenase products of arachadonate metabolism (prostacyclin) or NO'. We have also addressed the issue of how NO' controls the sensitivity of the VSM response to adrenergic stimulation. The results suggest that NO' may regulate both resting Cl conductance and the ability of agonists to activate the Cl current required for depolarization and contraction. Furthermore, it appears that NO' regulates vascular sensitivity to catecholamines by altering this voltage-dependent (nifedipine-sensitive) portion of the contractile response. This may occur by a combination of several mechanisms including 1) controlling the response of Ca²⁺ channels to a given degree of depolarization, 2) preventing depolarization via inhibition of Cl channels, or 3) limiting depolarization via activation of K⁺ channels.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Rings of thoracic aorta were obtained from adult male Sprague-Dawley rats and prepared for isometric force recording in a manner identical to that described in our preceding companion article (8). Control and low-Cl buffers were also prepared in the same manner. Low-Cl buffer in this study refers to 8 mM Cl substituted with 130 mM MS. The endothelium was removed from some rings by gently rolling the tissue around the end of a finely serrated steel forceps. Indomethacin treatment consisted of a 1-h exposure at 10⁶ M, which occurred during the 2-h equilibration period that preceded all experiments. NO synthesis was inhibited by incubation in N-nitro-L-arginine (L-NNA; 10⁴ M) for 20 min before a response was elicited. Dose-response experiments were performed cumulatively, and the response to each concentration of norepinephrine (NE) was observed for 10 min before the next higher concentration was added. This time period was generally adequate to allow a plateau of the contractile response. All drugs and all salts for the preparation of physiological salt solution were obtained from Sigma Chemical.

Data are displayed as means ± SE. Calculations of the half-maximal effective dose (ED₅₀) were performed by linear regression of each dose-response curve following logit transformation of the response data and log transformation of the agonist concentrations. The resulting linear equations were then solved for the dose producing the half-maximal response. Statistical analysis of group differences was performed using Student's t-test, and n values represent the number of animals in each group. A P value of <0.05 was considered to be statistically significant.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

There are two readily recognized phases to VSM contraction. The initial contractile response has been attributed to activation of the contractile proteins by Ca²⁺ released from the sarcoplasmic reticulum. Our hypothesis maintains that a second important function of this Ca²⁺ pool is to activate a depolarizing Cl current. This depolarization results in the influx of extracellular Ca²⁺, which contributes to the second, or maintained phase, of contraction. To assess the effect of endothelial factors on both phases of contraction, tension was recorded at both an early and a late time point. As previously demonstrated (8), the response of intact aortic rings to an approximately 20% effective dose (~ED₂₀) of NE is potentiated in low-Cl buffer (Fig. 1, A and B). The effect is most impressive at the 3-min, or peak, time point but is also statistically significant during the maintained phase of contraction (20 min). This effect of low-Cl buffer is not altered by inhibiting prostaglandin production with indomethacin (10⁶ M). Even in normal buffer, rubbed or L-NNA-treated rings contract more vigorously to a single dose of NE than do intact rings. A contraction to ~20% of the response to 120 mM KCl (ED₂₀) was achieved at a significantly lower concentration of NE in tissues with an impaired ability to synthesize NO' [intact rings, 3.0 ± 0.5 × 10⁸ M NE: 19.5 ± 2.7 at 3 min, 15.2 ± 2% at 20 min (n = 9); indomethacin-treated rings, 8.3 ± 2.6 × 10⁹ M NE: 21.2 ± 4.3% at 3 min, 12.4 ± 3.0% at 20 min (n = 6); rubbed rings, 1.2 ± 0.3 × 10⁹ M NE: 16.0 ± 2.0% at 3 min, 23.4 ± 3.3% at 20 min (n = 5); and L-NNA-treated rings, 1.6 ± 0.4 × 10⁹ M NE: 27 ± 7.1% at 3 min, 23.6 ± 4.4% at 20 min (n = 5)]. There was no significant difference among these contractile responses at the 3-min, or peak, time point. After 20 min, the response was significantly larger in rubbed or L-NNA-treated rings. This difference was due to a significant drop-off in tension from the peak to the maintained phase of contraction in the intact and indomethacin-treated rings. This drop-off may be attributed to NE-induced NO' production activated either directly, via endothelial receptors, or indirectly, by stretch. Rubbed and L-NNA-treated rings have a markedly potentiated response to NE in low-Cl buffer at both 3 and 20 min. This effect is most prominent on the maintained phase of contraction. The tension at the 20-min time point exceeds that at 3 min. This is in marked contrast to intact and indomethacin-treated rings, which show an even more remarkable time-dependent drop-off in tension in low-Cl buffer than in normal buffer. Figure 1B shows typical responses of isolated aortic rings with and without endothelium in normal and low-Cl buffer.

View larger version (22K): [in this window] [in a new window]   Fig. 1.   A: low-Cl buffer potentiates contractile response to norepinephrine (NE). Tension recorded at both 3 min (peak, open bars) and 20 min (steady state, solid bars) is significantly increased in intact rings. This effect is not altered by pretreatment with indomethacin. Removal of endothelium by rubbing or inhibition of reduced nitric oxide (NO') synthase with N-nitro-L-arginine (L-NNA) results in much greater potentiation of NE response by low Cl. * P < 0.05, tension in low-Cl buffer vs. tension in control; + P < 0.05, tension at 3 min vs. tension at 20 min in control rings. B: typical tension recordings of response of intact and rubbed rings to NE in control buffer (138 mM Cl) and in presence of 8 mM Cl. Solid vertical bars (left) represent the size of response of that ring to 120 mM KCl. The most remarkable effect of disrupting NO' production is the large increase in steady-state tension recorded at 20 min.

As previously reported (8), low-Cl buffer does not alter contractile responses to K⁺ in intact rings (Fig. 2). In these vessels, 18 mM K⁺ produced ~20% of the maximum response to KCl regardless of the Cl concentration of the buffer. In rings treated with L-NNA, the contractile response to K⁺ is potentiated. This potentiation increases with the duration of exposure to L-NNA (data not shown). In L-NNA-treated rings, 8 mM K⁺ (KCl in normal buffer and K-MS in low-Cl buffer) was used as the agonist because responses to lower concentrations were inconsistent. To control for the time-dependent effect of L-NNA, two rings were prepared from each animal (8 rings from 4 animals). Responses were obtained from both rings in both normal and low-Cl buffer, with one ring being exposed to 8 mM K⁺ in normal buffer first and the other ring exposed to low-Cl buffer first. The responses of the two rings from each animal were averaged, and therefore all eight rings were included in the statistical calculations. In contrast to the effect seen with NE, there was no effect of low-Cl buffer on K⁺-induced contraction when NO synthase was inhibited.

View larger version (14K): [in this window] [in a new window]   Fig. 2.   Low-Cl buffer does not alter contractile responses to K+ at either the 3-min (open bars) or 20-min (solid bars) time point. In intact rings, 18 mM K+ produces ~20% of maximum response to KCl regardless of Cl concentration of buffer. In rings treated with L-NNA, contractile response to K+ is potentiated; therefore 8 mM K+ was used. There was no significant effect of low-Cl buffer on K+-induced contraction even when NO synthase was inhibited.

Because of the dramatic ability of NO' to alter the potentiation of NE-induced contraction via low Cl, we wished to demonstrate that the previously documented effects of Cl-channel blockers [DIDS (10³ M), anthracene-9-carboxylic acid (10³ M), and niflumic acid (10⁴ M)] and Cl transport inhibitors [bumetanide (10⁵ M) and bicarbonate-free buffer (10 mM HEPES)] on NE-induced contraction (8) were not related to activation of NO' release by the endothelium. We therefore repeated these experiments in L-NNA-treated rings (Fig. 3). Inhibition of NO' production did not impair the ability of these interventions to suppress contractile responses to NE. These compounds appear to exert their effect independent of NO.

View larger version (18K): [in this window] [in a new window]   Fig. 3.   Ability of Cl-channel blockers and Cl transport inhibitors to interfere with NE-induced contraction is not altered by L-NNA and therefore is not dependent on endothelial release of NO'. Control responses to NE or KCl were obtained that were ~80% of response to 120 mM KCl [pair of bars (NE) or bar (KCl) at left for each labeled intervention]. Open bars represent highest tension achieved in the first 5 min; solid bars represent tension at 20 min. Repeat responses (bar or pair of bars at right for each labeled intervention) were obtained following a 20-min incubation in normal buffer only (Time control), bumetanide (105), or HEPES (Bumet; 10 mM) or a 10-min incubation in DIDS (103 M), anthracene-9-carboxylic acid (A-9-C; 103 M), or niflumic acid (104 M). * P < 0.05 vs. control.

Low-Cl buffer alone produced little or no contractile response in intact rings (n = 11), and no response was ever noted at the 3-min time point, whereas 3 of 11 rings showed very small contractions (2-9% of the response to 120 mM KCl) at the 20-min time point (Fig. 4A). In contrast, rubbed rings (n = 21) contracted consistently to low-Cl buffer, but to a variable degree. Examples of the spectrum of response to low Cl in rubbed rings are shown in Fig. 4B. These responses to low-Cl buffer were not inhibited by phentolamine (10⁵ M, n = 3, data not shown). It is possible that the three intact rings that did have small responses to low Cl underwent inadvertent partial endothelial damage during isolation.

View larger version (17K): [in this window] [in a new window]   Fig. 4.   A: low-Cl buffer alone produces little or no contractile response in intact rings. No response was ever noted at the 3-min time point (open bars). In rubbed rings a contractile response of inconsistent size is invariably seen. Solid bars, 20-min time point. B: typical contractile responses of rubbed rings to low-Cl buffer. Solid vertical bars at left for each response represent the size of the response of that ring to 120 mM KCl. Note variability of both maximal tension generated and time course of response.

Intact aortic rings underwent very small contractions in response to L-NNA (10⁴ M) that developed slowly [0% (3 min) and 3.1 ± 1.2% (20 min) of response to 120 mM KCl]. After a 10-min exposure to L-NNA, these rings contracted to low-Cl buffer to a degree similar to that of rubbed rings (Fig. 5A). These contractions were also quite variable from ring to ring, suggesting that the variability seen in the response to low Cl in rubbed rings was not a function of incomplete denuding but rather was characteristic of the contractile response to low Cl. Both the response to L-NNA and the low Cl-induced contraction are completely inhibited by nifedipine (10⁶ M, n = 6), suggesting that the low Cl response is produced by membrane depolarization that either did not occur or did not result in contraction in the presence of intact endothelial NO' production (Fig. 5B).

View larger version (15K): [in this window] [in a new window]   Fig. 5.   A: intact aortic rings contract in response to L-NNA (104 M), although these contractions are very small. After a 10-min exposure to L-NNA, rings contract to low-Cl buffer to a degree similar to that of rubbed rings (see Fig. 3). Open bars, 3-min time point; solid bars, 20-min time point. B: typical response of an intact aortic ring to L-NNA followed by low-Cl buffer. Solid vertical bar (left) represents the size of the response of that ring to 120 mM KCl. Low-Cl-induced contractions were completely inhibited by nifedipine (106 M).

Figure 6 shows cumulative dose-response curves to NE in intact rings before and after treatment with L-NNA (n = 7) and in rubbed rings (n = 6). Data in Fig. 6A is plotted as tension (g), whereas Fig. 6B depicts the data normalized as a percentage of the maximal response of each ring to NE. The rings achieved a greater maximal response to NE than to KCl. The maximal response to NE in intact rings was 7.8 ± 0.29 g, and after treatment with L-NNA this significantly increased to 8.35 ± 0.32 g (P < 0.05). The response to 120 mM KCl of these same rings was 4.45 ± 0.20 g. There was no significant difference among the log ED₅₀ for intact (7.86 ± 0.07), L-NNA-treated (8.00 ± 0.07), or rubbed rings (7.93 ± 0.11). In intact rings, there was a smaller response to NE when it was added as a single dose to produce an ~ED₂₀ response in the low Cl experiments (3.0 ± 0.5 × 10⁸ M) than to the same concentration of NE when achieved in a cumulative fashion during a dose-response experiment. The reason for this is not readily apparent, but the difference was not evident in rubbed or L-NNA-treated rings and therefore may be related to how these methods of NE exposure affect endothelial NO' release.

View larger version (18K): [in this window] [in a new window]   Fig. 6.   Concentration-response curves to NE in rat aortic rings. A: control rings () had an intact endothelium. These responses are compared with those from rubbed () and L-NNA-treated rings (). Rubbing appeared to slightly decrease maximal tension response. B: responses were normalized as percentages of maximal tension generation to NE. There was no significant change in sensitivity in rubbed or L-NNA-treated rings.

Cumulative dose-response curves to KCl (Fig. 7) were performed in intact rings before and after exposure to L-NNA (n = 5) and in rubbed rings (n = 4). Rubbing or treatment with L-NNA (10⁴ M) increased the sensitivity (Fig. 7B) of the rings to K⁺-induced depolarization (ED₅₀: 17.2 ± 1.2 mM in intact rings; 12.6 ± 0.9 mM in L-NNA-treated rings; 13.9 ± 2.5 mM in rubbed rings). L-NNA treatment also increased the maximum response (Fig. 7A) to K⁺ in paired experiments (4.57 ± 0.35 g in intact rings, 5.42 ± 0.40 g in L-NAA-treated rings). As previously stated, we find that there is a time-dependent increase in the response to K⁺ following rubbing or application of L-NNA. These experiments were performed 20 min after the first and only exposure of the rings to L-NNA.

View larger version (20K): [in this window] [in a new window]   Fig. 7.   Concentration-response curves to KCl. A: control rings () had an intact endothelium. These responses are compared with those from rubbed () and L-NNA-treated rings (). B: response are normalized as percentages of maximal tension generation to KCl. Both rubbed and L-NNA-treated rings were significantly more sensitive to K+-induced depolarization than was control. This suggests that sensitivity to depolarization is under influence of NO' and is important for determining agonist sensitivity.

Nifedipine (10⁶ M) was used to separate the voltage-dependent portion from the voltage-independent portion of the contractile response to NE. In the presence of nifedipine, both the maximal response (control 7.96 ± 0.29 g, nifedipine 2.28 ± 0.29 g, n = 5) (Fig. 8A) and the sensitivity (log ED₅₀: control 7.86 ± 0.7, nifedipine 7.37 ± 0.12) (Fig. 8B) to NE are markedly diminished. Treatment of these same rings with nifedipine plus L-NNA caused a large increase in their maximum ability to generate force (4.23 ± 0.51 g); however, there was no recovery of the diminished sensitivity (log ED₅₀: nifedipine + L-NNA 7.47 ± 0.03, not significantly different from nifedipine alone).

View larger version (17K): [in this window] [in a new window]   Fig. 8.   Concentration-response curves to NE in nifedipine-treated (106 M) rings with and without L-NNA (104 M). A: control rings () had an intact endothelium. These responses are compared with those from rings treated with nifedipine alone () or nifedipine plus L-NNA (×). Non-voltage-dependent component of NE-induced contraction is quite small. B: responses are normalized as percentages of maximal tension generated to NE. Sensitivity to the agonist is remarkably diminished by nifedipine. Although inhibition of NO' synthase does increase the magnitude of voltage-independent response, sensitivity is unaltered by L-NNA. These data suggest that the magnitude of the voltage-independent portion of response to NE is unlikely to be a major determinant of agonist sensitivity.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

We have previously shown (8) that NE-induced contraction of the rat aorta is potentiated by low-Cl buffer. We now demonstrate that the magnitude of this potentiation is dependent on endothelial NO. Rubbing or L-NNA increased the size of the response to a single low (ED₂₀) concentration of NE. After the dose of NE was reduced to account for this effect, repeating the response in 8 mM Cl buffer remarkably enhanced the magnitude of contraction. This effect was particularly pronounced on the maintained (20 min) phase of contraction. Contraction at this time point was mildly potentiated by low-Cl buffer when the endothelium was intact (8) but was increased severalfold following disruption of endothelial NO' production. The effects of rubbing and L-NNA were similar, although the control responses in the rubbed rings were slightly smaller and the potentiation by low Cl somewhat greater. Any difference between the effects of rubbing and L-NNA would suggest that other endothelial factors such as EDHF may also impact on the response to low Cl. Recent data indicate that EDHF acts through activation of K⁺ channels (18) and therefore may not directly affect the response to low Cl. Prostaglandins do not seem to influence the ability of NE to activate Cl current, because pretreatment with indomethacin does not alter the effect of low-Cl buffer. K⁺-induced contraction is not altered by low Cl conditions because elevated extracellular K⁺ does not activate a Cl conductance.

Low-Cl buffer alone does not elicit contraction of intact rings but consistently causes a variable degree of contraction in rubbed or L-NNA-treated rings. This suggests that the resting Cl conductance of intact vessels is low enough that even a dramatic change in the Cl equilibrium potential does not produce enough depolarization to cause contraction. In the absence of endothelial NO', the same change in Cl gradient elicits a contractile response that is completely blocked by nifedipine and therefore is due to depolarization. These findings suggest that either more Cl channels were open at the time of the Cl gradient change or more Ca⁺ current was activated by the same degree of depolarization. There is ample support from the literature for the regulation of Ca²⁺ channel conductance by cGMP (5, 10, 14, 16). Our results raise the possibility that, in addition to this effect on depolarization-induced Ca²⁺ entry, NO' may also suppress the degree of depolarization induced by an agonist by inhibiting the activation of Cl channel conductance.

We can only indirectly address the question of direct regulation of Cl channel conductance by NO' on the basis of our studies, which assume that contractile responses are proportional to depolarization. If NO' has a direct interaction with Cl channels, then one might expect rubbing or L-NNA to increase the contractile response to NE/low Cl more than the response to K⁺. These two stimuli represent alternative methods of producing membrane depolarization, and if the effect of NO' is limited to the Ca²⁺ channel, then they should be similarly affected by NO' withdrawal. Data from the dose-response to KCl shows that 12 mM K⁺ produced an ~ED₂₀ response in intact rings (22.5 ± 3.5%). After exposure to L-NNA, this response is approximately twice as large (46.3 ± 6.2%). In the NE/low-Cl studies, in the presence of L-NNA, we used a NE concentration of 1.6 ± 0.4 × 10⁹ M to achieve a peak response of 27 ± 7.1% and a maintained response of 23.6 ± 4.4%. This dose of NE is barely a threshold concentration in an intact ring (10⁹ M NE produced 3.6 ± 1.2% of response to 120 mM KCl, Fig. 6), and a higher dose was required to achieve an ~ED₂₀ response in intact rings (3.0 ± 0.5 × 10⁸ M). Repetition of the response to NE in L-NNA plus low Cl (Fig. 1) produced a peak response of 74.4 ± 9.2% and a maintained response of 85.2 ± 8.3%. The maintained response is approximately triple that seen in intact rings. This is even more impressive in view of the higher average concentration of NE used in the intact rings. Unfortunately, this comparison is not completely fair. The response to NE is not simply a function of depolarization. Many additional factors contribute to the magnitude of this response, and we cannot completely separate out the voltage-dependent component. Direct measurement of changes in membrane potential will be required to completely answer this question. The possible ways in which NO' may be acting to influence our results are summarized in Fig. 9.

View larger version (21K): [in this window] [in a new window]   Fig. 9.   Model cells outlining possible mechanisms by which endothelial NO may alter Ca2+ entry into vascular smooth muscle (VSM). We have performed companion studies (8) delineating the contribution of Cl currents to agonist-induced contraction of VSM. Removal of endothelium by rubbing or L-NNA treatment allows a much greater response to low-Cl buffer. Several possible explanations may exist: 1) low-Cl buffer increases NO' release from endothelial cells, thereby suppressing contraction of intact rings; 2) NO' suppresses L-type Ca2+ channel function, thereby minimizing contractile response to increased depolarization; or 3) NO' tonically suppresses Cl conductance so that, after withdrawal of NO', there are more open Cl channels and more depolarization is produced when extracellular Cl is lowered. In presence of an agonist, NO' may also alter the degree to which Cl conductance can be activated. cNOS, constitutive NO synthase.

One additional factor that could have an impact on the interpretation of our results and that must be considered is the possibility that low-Cl buffer can alter endothelial NO' production directly. If low-Cl buffer increased basal NO' production or augmented an NE-/stretch-induced increase in NO' production, this could explain the comparatively diminished ability of low Cl to augment NE-induced contraction in intact rings. In fact, the literature suggests that just the opposite is true. Endothelial NO' production appears to be directly proportional to the sustained level of intracellular Ca²⁺ achieved following stimulation by a vasoactive agent (12). This Ca²⁺ enters the cell from extracellular sources through a nifedipine-insensitive, voltage-independent channel (1). Nevertheless, Ca²⁺ influx is apparently controlled by membrane potential in that endothelial cell hyperpolarization increases the driving force for Ca²⁺ influx and thereby increases the sustained level of intracellular Ca²⁺ (11). Conversely, depolarization by high K⁺ depresses the agonist-induced sustained increase in intracellular Ca²⁺ (9, 11, 21). Endothelial cells clearly possess Cl channels (4, 13, 19), and reduction of extracellular Cl (to 20 mM) was found to markedly depress the ATP-induced increase in sustained level of Ca²⁺ in human aortic endothelial cells (21). On the basis of these results, one would predict that low-Cl buffer would inhibit rather than augment the sustained release of NO'. This would clearly not explain our results.

The dose-response curve to K⁺ is more affected by disruption of NO' production than is the dose-response curve to NE. The change in the maximal response to NE with L-NNA was significant but quite small. We have no clear explanation for the difference between the response to a single low concentration of NE and the response to cumulatively added NE. Whatever the reason for the difference, it may be a function of NO' production, because the difference is not seen in rubbed or L-NNA-treated rings. The more pronounced effect of NO' on KCl-induced contractions may reflect the fact that although this response is completely coupled to depolarization, only a portion of the response to NE is voltage-dependent. We were able to accentuate this portion of the response by lowering extracellular Cl and thereby observed a large change in the response to a relatively low concentration of NE.

We have made an attempt to define the voltage-dependent portion of the dose response to NE and to determine how that portion of the response to NE contributes to the sensitivity and magnitude of contraction. Complete inhibition of depolarization-induced Ca²⁺ influx through voltage-dependent channels (10⁶ M nifedipine) dramatically reduces both the magnitude of contraction and the sensitivity to NE in intact rings (Fig. 8). Addition of L-NNA approximately doubles the size of the contractile response to NE in the presence of nifedipine but has absolutely no effect on sensitivity. This result suggests that NO' has significant effects on vascular contraction that are independent of Ca²⁺ influx, but these intracellular events do not determine sensitivity. The remarkable L-NNA-induced change in sensitivity to K⁺ suggests that Ca²⁺ channels are an important target for regulation of vascular sensitivity by NO'; however, changes in intracellular Ca²⁺ handling may also contribute to this. The large rightward shift in the sensitivity to NE induced by nifedipine suggests that agonist-induced depolarization makes an important contribution to the contractile response even at low agonist concentrations of NE. If this depolarization is indeed Cl dependent, Cl homeostasis may play an important role in determining how a given tissue responds to an agonist.

We can speculate that cellular Cl handling and the endothelial regulation of Cl currents may also have pathophysiological significance. The importance of Cl currents may be accentuated in regions of localized endothelial dysfunction such as the coronary artery of atherosclerotic patients. Under the diminished influence of NO', a sudden rise in circulating catecholamines might elicit increased localized Cl-dependent depolarization and a more sustained contractile response, resulting in ischemia. If this is indeed the case, selective inhibition of Cl channel conductance may prove to be a useful strategy for controlling coronary vasospasm.

In summary, we have demonstrated that the endothelium, via NO', suppresses the degree to which NE-induced contraction is potentiated by low-Cl buffer. In addition, low-Cl buffer does not produce contraction in intact rings but consistently does so after endothelial NO' production is disrupted. These data suggest that NO' can inhibit the opening of voltage-dependent Ca²⁺ channels by an agonist. This inhibition may be direct (via cGMP inhibition of Ca²⁺ channels) or indirect via prevention of the activation of depolarizing Cl currents.