INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

UNDER MOST CIRCUMSTANCES vascular smooth muscle (VSM) tone and contractility are tightly coupled to membrane potential (V_m) (29). This relationship is maintained by the dependence of contraction on the influx of Ca²⁺ through voltage-dependent, primarily L-type Ca²⁺ channels. Numerous studies have attempted to define the role of various K⁺ conductances (delayed rectifier, Ca²⁺ activated, ATP dependent) in the control of resting V_m, agonist-induced depolarization, and vasodilatation (4, 11, 18, 40). Conversely, although Cl channels are present in VSM (23, 27) and are activated by agonists (24, 32), there is minimal functional evidence defining the contribution of these events to contractile responses.

The manner in which VSM cells handle Cl sets up an ideal system for producing and maintaining depolarization. VSM cells accumulate Cl intracellularly through several processes (9, 13, 14), including Na⁺-K⁺-2Cl cotransport, Cl/HCO₃ exchange, and a third component, possibly an ATP-dependent transporter (16). Whereas resting V_m in VSM ranges from approximately 45 to 65 mV, estimates of the Cl equilibrium potential [E_Cl = 60 log (extracellular Cl concentration/intracellular Cl concentration)], measured using either radiolabeled Cl flux (25, 38) or ion-selective microelectrodes (16), range between 11 and 50 mV. In any given vascular tissue, E_Cl has always been measured to be roughly 15-30 mV more positive than V_m. Depolarization of the precise magnitude induced by vasoconstrictors can therefore be readily produced by an increase in Cl conductance.

Estimates of relative permeabilities at rest for Cl and K⁺ have varied widely from 0.09 in rat femoral artery (10) to 0.82 in rat portal vein (38); however, this ratio may change dramatically after exposure to a contractile agonist. Adrenergic stimulation has frequently been shown to increase total membrane conductance [rabbit carotid artery (28) and pulmonary artery (7), guinea pig mesenteric artery (3) and pulmonary artery (5)] consistent with the activation of a Cl current (decrease in membrane resistance). Occasionally, total membrane conductance has been found to decrease [guinea pig ear artery (22), pulmonary artery (35)]. These contrasting results may be due to functional differences between tissues. Alternatively, the interpretation of these results may be complicated by the syncytial nature of VSM. Changes in membrane resistance may be obscured or magnified by changes in the electrical coupling between cells.

Radiolabeled ion flux data are less dependent on cell-cell coupling. Norepinephrine (NE) dramatically increases the efflux of both ⁴²K⁺ and ³⁶Cl from rabbit pulmonary artery (7, 34). The computed increase in permeability for Cl exceeds that for K⁺ (21). Because depolarization moves V_m closer to E_Cl (diminishing the driving force for Cl movement), it is difficult to explain an increase in Cl efflux unless there is a significant increase in Cl conductance. Conversely, the observed increase in K⁺ efflux is difficult to reconcile with a mechanism of depolarization that is driven primarily by the alternative proposed mechanism, a decrease in K⁺ conductance.

Agonist-induced Cl currents have now been characterized in a number of vascular tissues. NE activates a Cl current in cells from portal vein (6, 32), mesenteric vein (36), and ear artery (2). Endothelin elicits a similar current in coronary, aortic, and mesenteric VSM cells (24, 37), as does vasopressin in cultured aortic cells (17, 37). These Cl currents are Ca²⁺ dependent (I_Cl,Ca) (27, 31) and appear to be activated initially by agonist-induced release of intracellular Ca²⁺ stores. The extent to which Ca²⁺ entry from extracellular sources can sustain activation of these channels is unknown. At the single-channel level they appear to have a low conductance of 1-2 pS (23). Another Cl channel that has been characterized in VSM at the single-channel level is a large conductance channel (340 pS) that is activated by protein kinase C inhibitors (and is therefore likely inhibited by protein kinase C-dependent phosphorylation) in cell-attached patches (33). VSM also contains a typical volume-activated Cl current (30).

Despite the abundance of evidence documenting the existence of agonist-induced Cl currents, the degree to which the resulting depolarization contributes to contraction is not known. We have employed a strategy involving 1) alteration of E_Cl, 2) inhibition of Cl transport, and 3) block of Cl channels to obtain conclusive functional evidence that Cl currents are critical to agonist-induced activation of vascular smooth muscle.

	METHODS

Top Abstract Introduction Methods Results Discussion References

Adult male Sprague-Dawley rats (250-300 g) were obtained from Harlan Sprague Dawley. The animals were killed by exposure to 100% CO₂ for 5 min, followed by cervical dislocation. Thoracic aortas were removed, cleaned of adherent connective tissue, and cut into 6-mm rings. The endothelium was left intact, and the rings were mounted in individual 10-ml isolated organ chambers (Radnoti Glass) using stainless steel triangles and were connected to an isometric force transducer (Kent Scientific) by 32-gauge stainless steel wire. Contractile responses were recorded with an eight-channel MacLab 8E and stored on a Power Macintosh 7200 computer. Passive stretch was set at 2.5 g, and the rings were allowed to equilibrate in physiological salt solution (PSS) at 37°C for 120 min before the start of experimentation. PSS was aerated with a mixture of 95% O₂-5% CO₂; the composition was as follows (in mM): 130 NaCl, 4.7 KCl, 1.18 KH₂PO₄, 1.17 MgSO₄ · 7H₂O, 14.9 NaHCO₃, 1.6 CaCl₂ · H₂O, 5.5 dextrose, and 0.03 CaNa₂-EDTA 0.03 (pH 7.30). The standard low-Cl buffer was prepared by substituting NaCl with 130 mM NaOH and titrating the pH of the buffer to 7.30 with methanesulfonic acid while the solution was being aerated. The measured osmolality of the low-Cl buffer was 292 mosmol/kg compared with 293 mosmol/kg for the control buffer (5500 vapor pressure osmometer). For the concentration-response curve to extracellular Cl (Fig. 4), the various Cl concentrations were achieved by mixing standard PSS with low-Cl buffer in fixed ratios. Low-Cl buffers substituted with Br or I were made by replacing 130 mM NaCl completely with NaBr or NaI. HEPES buffer was made by substituting 10 mM HEPES for bicarbonate and titrating the pH of the solution to 7.30 using 1 M NaOH. When both HEPES and bicarbonate were used, pH was once again 7.30.

In all experiments in which extracellular Cl was altered suddenly, the low-Cl buffer was prewarmed and preaerated in a 37°C constant temperature bath before the solution was poured directly into the drained organ chamber. This was done to avoid depletion of intracellular Cl before exposure to the contractile agent. Control responses for these experiments were also recorded after the agonist-containing control buffer was poured into the bath.

At the beginning of each experiment a contractile response to 120 mM KCl was recorded. Subsequent contractile responses are normalized by expression as a percentage of this initial maximal response to KCl. All agonist-induced contractions were recorded for 20 min. When KCl was used to induce contraction, it was added directly to the buffer from a 1 M stock without adjusting for changes in tonicity. When K⁺ was used as an agonist in low-Cl conditions, it was added from a 1 M stock of K-methanesulfonate. This stock was prepared by titrating the pH of 1 M KOH to 7.3 with methanesulfonic acid. Ca²⁺-free PSS was prepared by omission of CaCl₂ without the use of an additional chelating agent. The bath was washed three times in this buffer over an ~5 min period before addition of the agonist.

There are two readily recognized phases to VSM contraction. The initial contractile response has been attributed to direct activation of the contractile proteins by Ca²⁺ released from the sarcoplasmic reticulum (SR). 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 is responsible for the second, or maintained, phase of contraction. To make a functional assessment of both phases of contraction, the contractions are measured at both an early and a late time point. Lower concentrations of agonist [~20% effective dose (ED₂₀)] generally produce a clear early peak of contraction (~3 min) that exceeds the tension measured once contraction has stabilized. These data (see Figs. 1, 2, 4, 5, and 10) are displayed as paired bars, peak contraction (highest tension recorded at 3 min), and tension at the 20-min time point. In experiments in which higher concentrations of agonist are used [~80% effective dose (ED₈₀)] to study inhibition of contraction (see Figs. 6-9), there is generally no early peak seen. Tension increases rapidly over the first 5 min and then more slowly until a stable level is achieved at ~20 min. These data are displayed as tension at the 5- and 20-min time points.

NE, serotonin (5-HT), DIDS, and ryanodine were dissolved directly into aqueous solution. BAY K 8644 and bumetanide were prepared as stock solutions in ethanol, whereas ethacrynic acid, tamoxifen, niflumic acid, anthracene-9-carboxylic acid (A-9-C), and cyclopiazonic acid (CPA) were prepared as stock solutions in DMSO. The drugs were added to the buffer at no greater than a 1,000:1 dilution, yielding a final solvent concentration of 0.1%. No individual ring was used for more than a single pharmacological intervention (drug or combination of drugs) to avoid potential effects of incomplete washout. All drugs and salts for the preparation of PSS were obtained from Sigma Chemical with the exception of BAY K 8644 (Calbiochem).

Data are displayed as means ± SE, and n represents the number of rats in each group. Statistical analysis of group differences was performed using Student's t-test. A value of P < 0.05 was considered to be statistically significant.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Contractile responses to NE and 5-HT but not to KCl are potentiated in low-Cl buffer. The concentrations of these contractile agents were titrated in each individual ring to achieve a control response that was ~20% of the response to 120 mM KCl. Figure 1A shows the effect of low-Cl buffer on the peak and maintained phases of contraction in response to these agents. The potentiating effect of low-Cl is more dramatic at the 3-min time point but remains significant after 20 min. The force recording shown in Fig. 1B demonstrates the typical pattern of phasic contractions that occur during the initial phase of contraction in low-Cl buffer. Mean concentrations used were 2 × 10⁸ M for NE (n = 9), 1.6 × 10⁶ M for 5-HT (n = 6), and 18 mM for K⁺ (n = 5). Although both NE and 5-HT elicit SR Ca²⁺ release and can activate I_Cl,Ca, K⁺ cannot. All of these experiments (including control responses) were performed by exposing the rings to NE and to the altered extracellular Cl concentration at the same moment by rapidly draining the vessel chamber and replacing the buffer with prewarmed, premixed low-Cl buffer. Low-Cl buffer does not cause any contractile response by itself (data not shown).

View larger version (16K): [in this window] [in a new window]   Fig. 1.   Contractile responses to norepinephrine (NE) and serotonin but not to K+ are potentiated by low-Cl buffer. A: both peak (3 min, open bars) and maintained (20 min, solid bars) contractile responses to NE and serotonin are significantly larger than control in low-Cl buffer while responses to K+ are unaltered. * P < 0.05. B: typical response of an aortic ring to NE in control (138 mM) and low-Cl (8 mM) buffer. The change in Cl concentration and addition of agonist are accomplished simultaneously by pouring on prewarmed and preaerated buffer containing the drug at the desired concentration. A follow-up control response demonstrates that the effect of low-Cl buffer is completely reversible.

The results from Fig. 1 suggest that Cl current contributes more to agonist-induced activation of VSM during the initial phase of contraction. Alternatively, the potentiating effect of low-Cl buffer may diminish with time as intracellular Cl falls and cannot be readily replenished. To distinguish between these two possibilities, rings were first contracted in normal 138 mM Cl buffer with ~ED₂₀ NE and allowed to reach a stable, maintained level of tension (20 min). The buffer was then rapidly exchanged in an identical manner to that used before and was replaced with buffer containing the same concentration of NE and either 138 or 8 mM Cl [substituted with methanesulfonate (MS)]. The results are shown in Fig. 2. The contractile response is again dramatically potentiated by low-Cl buffer, and this effect diminishes over 5-10 min; however, tension remains significantly greater than that of the control 20 min later. These results suggest that Cl channels remain open during the sustained phase of contraction and that intracellular Cl is rapidly depleted in low-Cl buffer.

View larger version (18K): [in this window] [in a new window]   Fig. 2.   Low-Cl buffer also potentiates NE-induced contraction during maintained phase of contraction. A: there is no significant change in tension when bath is drained after 20 min of contraction and replenished with fresh buffer with the same concentration of NE (control). If extracellular Cl is lowered when buffer is changed [8 mM Cl substituted with methanesulfonate (MS)], the contractile response is potentiated (low Cl). Open bars, first initial peak response to NE; shaded bars, first maintained response to NE; solid bars, second initial peak response to NE (fresh buffer); hatched bars, second maintained response to NE (fresh buffer). * P < 0.05 compared with same time point during control protocol. B: typical recording of a single ring showing both control (left) and low-Cl (right) responses.

BAY K 8644 (5 × 10⁵ M, n = 5), a direct Ca²⁺-channel activator that causes Ca²⁺ influx, produced a transient contractile response in normal Ca²⁺ buffer. This response was unaltered by low-Cl buffer (Fig. 3). The activity of BAY K 8644 is voltage dependent, and one might expect a larger response if low-Cl buffer produced a significant depolarization. However, BAY K 8644 does not cause release of SR Ca²⁺ and may not produce sufficient elevation of intracellular Ca²⁺ to activate I_Cl,Ca. NE (10⁸ M, n = 6) in the absence of extracellular Ca²⁺ produced a similar transient contractile response due to the release of intracellular Ca²⁺ stores. Again, this contraction was not effected by low-Cl conditions (Fig. 3). Even if a Cl current was activated by NE, in the absence of extracellular Ca²⁺ no Ca²⁺ influx can result from depolarization regardless of the magnitude.

View larger version (22K): [in this window] [in a new window]   Fig. 3.   BAY K 8644 in normal Ca2+ buffer and NE (108 M) in nominally Ca2+-free buffer both produce transient contractile responses (open bars). When these responses are repeated in low-Cl buffer (hatched bars), there is no change in sizes of responses.

The potentiating effect of low-Cl buffer on NE-induced contractions is dependent on the degree to which the extracellular Cl concentration is lowered. Figure 4 represents a concentration-response curve to lowering extracellular Cl (from 138 to 8 mM) using the same dose of NE at all Cl concentrations (n = 4). The peak of contraction is significantly potentiated in 41 and in 8 mM Cl, whereas the maintained response is enhanced only in 8 mM Cl.

View larger version (19K): [in this window] [in a new window]   Fig. 4.   Effect of lowering extracellular Cl is concentration dependent. Peak responses (open bars) are significantly augmented when extracellular Cl is lowered to 41 or 8 mM. Maintained responses (20 min, solid bars) are larger only when Cl concentration is lowered to 8 mM. * P < 0.05 compared with response in normal, 138 mM Cl.

Substitution of Cl with MS should make E_Cl more positive due to its very low permeability (no inward MS current). Substitution with an anion having a significant ability to permeate the VSM anion channels would allow inward movement down their concentration gradients (infinitely large at the instant of substitution). This movement will contribute to V_m and would be predicted to produce hyperpolarization. Figure 5 compares the effect on the peak contractile response to NE when MS (n = 5), Br (n = 5), or I (n = 5) are used to replace Cl. Once again in these tissues, Cl substitution with MS resulted in a significant potentiation of the peak NE-induced contraction. However, Br and I significantly suppressed the contractile response to NE, with I producing a more profound effect than Br.

View larger version (15K): [in this window] [in a new window]   Fig. 5.   Size of peak response to NE is dependent on the anion used to substitute for Cl. Substitution with MS for Cl results in potentiation of contraction. Substitution with Br or I causes suppression of initial peak response. * P < 0.05 compared with Cl.

Inhibitors of Cl transport suppress contractile responses to NE (Fig. 6). Rings were first contracted with a dose of NE titrated to produce an ~ED₈₀ response (mean NE dose 4.9 × 10⁸ M). After the control response was obtained, the rings were incubated for 20 min in either PSS alone (time control, n = 4), bumetanide (10⁵ M, n = 5), 10 mM HEPES (0 mM HCO₃, n = 5), ethacrynic acid (10⁵ M, n = 5), bumetanide + HEPES (n = 4), ethacrynic acid + HEPES (n = 5), bumetanide + ethacrynic acid (n = 6), or bumetanide + HEPES + ethacrynic acid (n = 6). Time alone had no effect on the magnitude of the response to NE, because the second response to NE was unchanged from control. Both bumetanide and HEPES buffer significantly suppressed NE-induced contraction, whereas ethacrynic acid produced very mild inhibition at the 5-min time point only. Bumetanide and HEPES together produced a greater inhibitory effect than either agent alone, and this combination produced as much inhibition of contraction as all three agents together.

View larger version (17K): [in this window] [in a new window]   Fig. 6.   NE-induced contractions are inhibited by Cl transport inhibitors. Control responses to NE were obtained that were ~80% of response to 120 mM KCl (ED80; pairs of bars at left for each labeled intervention). In contrast to responses at ~20% of response to 120 mM KCl (ED20), responses of this magnitude fail to show a consistent initial peak tension, and therefore force is displayed as the highest tension in the first 5 min, generally as tension at the 5-min time point (open bars), and as tension at the 20-min time point (solid bars). Control contractions were of similar magnitude in all groups. Repeat responses were obtained following a 20-min incubation in physiological salt solution only (Time, n = 4), 105 M bumetanide (Bum, n = 5), 10 mM HEPES buffer without bicarbonate (HEPES, n = 5), 105 M ethacrynic acid (Eth, n = 5), or a combination of interventions [Bum + HEPES (Bum/Hep), n= 4; Eth + HEPES (Eth/Hep), n = 5; Bum + Eth (Bum/Eth), n = 6; or all 3 together (Bu/He/Et, n = 6).] These responses are displayed as pairs of bars at right for each labeled intervention. There is no effect of Time alone. Eth mildly inhibits contraction at 5-min time point. Bum and HEPES significantly suppress contractions to NE. Bum/Hep appears to be more potent than either agent alone. * P < 0.05 compared with control response.

Figure 7 demonstrates that it is the absence of bicarbonate, not the presence of HEPES, that suppresses NE-induced contractions. When both HEPES and bicarbonate are included as buffers (n = 4), NE-induced contractions are unaltered. To control for nonspecific effects of bumetanide or HEPES, experiments similar to those in Fig. 6 were carried out using K⁺ as the contractile agent (Fig. 8). K⁺-induced contractions were obtained that were similar in size to those achieved with NE (~ED₈₀); the mean concentration of K⁺ required was 38 mM. Neither bumetanide (n = 5) nor HEPES buffer (n = 5) had any significant inhibitory effect on K⁺-induced contractions.

View larger version (12K): [in this window] [in a new window]   Fig. 7.   Effect of HEPES buffer on NE-induced contractions is due to absence of bicarbonate, not to presence of HEPES. When both 10 mM HEPES and the usual 14.9 mM bicarbonate were included in buffer (right), control contractile responses to NE (left) were not altered. Open bars, 5 min; solid bars = 20 min.

View larger version (48K): [in this window] [in a new window]   Fig. 8.   KCl-induced contractions are not altered by low-Cl conditions. Each bar represents tension measured at 20-min time point. Only a single time point was recorded because responses to this dose of KCl never show an initial peak. Open bars represent control contractions, and hatched bars represent tension obtained in response to same dose of KCl after 20 min of exposure to either 105 M bumetanide (n = 5) or 10 mM HEPES buffer (n = 5). Neither had any effect on contractions induced by KCl.

Cl-channel blockers inhibit contractile responses to NE (Fig. 9). Four different compounds with the previously documented ability to inhibit anion currents were assayed for their ability to inhibit contractile responses to NE or K⁺. The design of these experiments was identical to that of the experiments in Fig. 6. After a control ~ED₈₀ response to the agonist and subsequent washout, 10-min incubations in either DIDS (10³ M; n = 5 for NE, n = 6 for KCl), A-9-C (10³ M; n = 5 for NE, n = 6 for KCl), niflumic acid (10⁴ M; n = 5 for NE, n = 4 for KCl), or tamoxifen (10⁵ M; n = 6 for NE, n = 5 for KCl) were performed before reexposure to the agonist. DIDS, A-9-C, and niflumic acid all markedly inhibited NE-induced contractions. Of these three compounds, only DIDS was without effect on KCl-induced contraction. Tamoxifen had no significant effect on contractile responses to either vasoconstrictor.

View larger version (17K): [in this window] [in a new window]   Fig. 9.   Cl-channel blockers inhibit NE-induced contractions. Control responses to NE or KCl were obtained that were ~80% of response to 120 mM KCl (bar or pair of bars at left for each labeled intervention). For NE responses, open bars represent highest tension recorded in first 5 min and solid bars represent tension at 20 min. Only tension at 20 min is recorded for KCl-induced contractions. After a 10-min incubation in DIDS (103 M), anthracene-9-carboxylic acid (A-9-C; 103 M), niflumic acid (104 M), or tamoxifen (105 M), repeat responses were obtained (bar or pair of bars at right for each labeled intervention). DIDS, A-9-C, and niflumic acid all inhibited responses to NE, whereas A-9-C and niflumic acid also inhibited KCl-induced contractions. * P < 0.05 compared with control response.

We wanted to further test the hypothesis that the effect of altering E_Cl on NE-induced contraction was dependent on the release of intracellular Ca²⁺. We therefore used either CPA (10⁷ M, n = 5) to inhibit uptake of Ca²⁺ into the SR or ryanodine (10⁵ M, n = 5) to prevent the release of SR Ca²⁺ (Fig. 10). It was not possible to employ a combination of these two agents because this combination of drugs consistently resulted in large contractions. After an initial control ~ED₂₀ response to NE was obtained, rings were continuously exposed to CPA or ryanodine for the remainder of the experiment. Two subsequent control responses to the same dose of NE were elicited to establish a meaningful control response in the presence of the drugs. The effect of ryanodine on the control responses was apparent in that the larger initial peak response to NE was lost and contractions were slower and larger at the 20-min time point. Both CPA and ryanodine diminished the size of control responses to NE. A final response to NE was then obtained in 8 mM Cl buffer (MS substituted) containing CPA or ryanodine. This response to NE tended to be larger than the second control response obtained in the presence of the drugs; however, the increase in the peak response to NE seen in low-Cl buffer was no longer statistically significant (CPA 2nd control = 13.4 ± 2.4, CPA/low Cl = 29.2 ± 9.4, P = 0.11; ryanodine 2nd control = 7.2 ± 2.2, ryanodine/low Cl = 19.6 ± 8.1, P = 0.08). One can assess the effect of these drugs on the potentiation due to low Cl by comparing potentiation as a percentage of the 120 mM KCl response. For example, if the control peak is 21% of the KCl response and the peak in low Cl is 57% of the KCl response, then there is an increase of 36%. Under control conditions the peak response to NE is increased in low Cl by 35.6 ± 4.7% of the response to 120 mM KCl (Fig. 1, n = 9). In CPA, low Cl increases the peak response by 7.6 ± 6.4% compared with the initial drug-free control response and by 15.8 ± 8.6% compared with the second response (n = 5). Similarly, for ryanodine the increase compared with the drug-free control is 0.8 ± 7.9% and that compared with the second drug-treated control is 12.4 ± 5.9% (n = 5). All of these percent increases are statistically smaller in the presence of the drugs (unpaired t-test). In contrast to their suppression of the peak response, these drugs appear to augment the ability of low Cl to potentiate the maintained phase of contraction.

View larger version (18K): [in this window] [in a new window]   Fig. 10.   Effect of cyclopiazonic acid (CPA; 107 M) (A) or ryanodine (Ryan; 105 M) (B) on NE-induced contractions. Data represent peak (open bars) or 3-min (#1) and 20-min (#2) time points (solid bars). First, a control response was obtained that was ~20% of response to 120 mM KCl. Rings were then incubated for 20 min in CPA or ryanodine, and these drugs remained in the bath for the rest of the experiment (including during washout between responses). Two subsequent contractile responses were recorded in presence of drugs to obtain an assessment of the effect of these compounds on NE-induced contractions in normal Cl buffer. Subsequent responses were diminished in size, and ryanodine clearly altered time course of contraction by suppressing initial peak response. The final response was obtained in low-Cl (8 mM) buffer. * P < 0.05 compared with second control response in presence of drugs.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

We have used a variety of approaches to gather data in support of the concept that NE-induced VSM contraction is dependent on activation of a Cl current. Cl currents appear to contribute to the maintenance of tension not only during the initial phase of the response, when I_Cl,Ca has been shown to be activated by the release of SR Ca²⁺, but also during the maintained phase of contraction. We have shown that NE-induced contraction is potentiated in low-Cl buffer (MS substituted) regardless of when extracellular Cl is lowered during the contractile response. In contrast, responses to KCl, BAY K 8644, and NE in the absence of extracellular Ca²⁺ are not affected by altering extracellular Cl. These other stimuli lack either the ability to release Ca²⁺ stores and thereby activate Cl current (KCl, BAY K 8644) or the ability for the depolarization produced by that current to facilitate inward Ca²⁺ flux (NE in 0 Ca²⁺). The potentiating effect of low Cl on NE-induced contraction is concentration and anion dependent. Unlike MS substitution, I or Br substitution causes suppression of the peak response to NE. These data suggest that I and Br pass through the channels that are activated by NE more readily than does Cl. This implied relative conductivity sequence (I > Br > Cl > MS) is consistent with that recorded for spontaneous inward Cl currents characterized in rat portal vein (39). Drugs that interfere with either Cl transport (bumetanide, HEPES) or Cl-channel function (DIDS) inhibit contractile responses to NE but not to KCl. Although the endothelium was left intact in these experiments, the effect of these agents is not due to stimulation of endothelial nitric oxide release (26). Finally, the Cl current responsible for this depolarization is at least partially dependent on agonist-induced release of SR Ca²⁺ stores, because CPA and ryanodine alter the potentiating effect of low-Cl buffer. These findings are summarized in Fig. 11.

View larger version (13K): [in this window] [in a new window]   Fig. 11.   Model smooth muscle cell outlining proposed mechanism by which Cl currents produce vascular smooth muscle depolarization. Important structures are labeled with the pharmacological tool affecting their function. See text for detailed discussion.

We propose that the mechanism of the potentiating effect of low-Cl buffer is related to the associated change in E_Cl. MS is a very impermeable inert anion that has been used previously without discernible biological effects aside from those related to the altered Cl concentration (27, 42). We changed Cl concentrations abruptly and either simultaneously with or 20 min after exposure to the contractile agent. In both instances, the degree of potentiation decreased with time but remained significant even after 20 min. Previous studies have in fact suggested that intracellular Cl begins to fall measurably within 1-2 min following reduction of extracellular Cl to 15 mM (15). In evaluating the dose-response to Cl (Fig. 4), it is important to note that although the changes in extracellular Cl were made on a linear scale, the effect of these changes on E_Cl is logarithmic. For this reason, changes in extracellular Cl will have a more profound effect on E_Cl at lower Cl concentrations. For example, with the assumption of an intracellular Cl of 44 mM, as has been measured in the rat femoral artery (14), changing extracellular Cl from 139 to 106 mM will change E_Cl from 30 to 23 mV (7 mV), whereas dropping extracellular Cl from 41 to 8 mM changes E_Cl from +2 to +44 mV (42 mV). This explains why there is little effect of lowering extracellular Cl until relatively low concentrations are reached.

While it is impossible for serum Cl ever to fall low enough for alterations in E_Cl to effect vascular reactivity in vivo, large changes in E_Cl may be achieved by even modest elevations in intracellular Cl, and these changes may have physiological significance. Davis et al. (14) have shown that mineralocorticoid-induced (DOCA-salt) hypertension in the rat is associated with a significant rise in intracellular Cl in femoral artery VSM (34 vs. 52 mM in bicarbonate-free buffer). This difference can be attributed to an increase in the activity of the Na⁺-K⁺-2Cl cotransporter. If one assumes a serum Cl concentration of 100 mM, this change in intracellular Cl concentration translates to a shift in E_Cl from 29 to 18 mV. This will produce more depolarization when Cl conductance dominates V_m, as it may when a contractile agonist binds. This phenomenon may explain at least a portion of the shift in sensitivity to a variety of agonists that is seen in DOCA-salt hypertension.

Producing a depolarizing anion current requires Cl to be accumulated intracellularly by transport against its electrochemical gradient. Interfering with this accumulation leaves Cl passively distributed, and hence, no depolarizing Cl current can flow regardless of how many channels are open. VSM accumulates Cl via Na⁺-K⁺-2Cl cotransport and Cl/HCO₃ exchange (13). Bumetanide specifically inhibits Na⁺-K⁺-2Cl cotransport (8), whereas the absence of bicarbonate ion interferes with Cl/HCO₃ exchange (13). In guinea pig femoral artery this interference occurs without altering intracellular pH due to a large contribution of Na⁺/H⁺ exchange to pH regulation (1). Although we have not measured intracellular pH in our tissues, HEPES buffering had no effect on KCl-induced contraction, suggesting that the contractile capability of the rings remained intact. The lack of effect of HEPES on NE-induced contraction in the presence of bicarbonate suggests that, indeed, its effects can be attributed to the absence of bicarbonate. The ability of bumetanide and HEPES to inhibit NE-induced contraction suggests that both of these transporters play a vital role in maintaining the ability of VSM to respond to vasoconstrictors. Ethacrynic acid had only a tiny effect on contractile responses to NE. This may be related to compensatory activity of the other two transport mechanisms or may indicate that this third transporter is not important in aortic tissue.

A number of compounds have been shown to block Cl channels; however, none have proven ideal for sorting out the contribution of Cl currents to cell activation. DIDS, A-9-C, and niflumic acid all potently inhibit spontaneous transient inward currents produced by I_Cl,Ca in rabbit portal vein (19, 20). They do this at lower concentrations (IC₅₀: 2.1 × 10⁴ M for DIDS, 3 × 10⁴ M for A-9-C, and 2.3 × 10⁶ M for niflumic acid at 50 mV) than those at which they inhibit agonist (NE, caffeine)-induced Cl current (IC₅₀: 7.5 × 10⁴ M for DIDS, 6.5 × 10⁴ M for A-9-C, and 6.6 × 10⁶ M for niflumic acid). In addition, these compounds have no effect on spontaneous transient outward currents produced by Ca²⁺-activated K⁺ currents (I_K,Ca), but they clearly augment agonist-induced K⁺ currents. These data suggest that these compounds, in addition to blocking Cl currents, may also enhance release of Ca²⁺ from the SR (19, 20). We have looked at the ability of each of these agents to inhibit contraction. The concentration of DIDS and A-9-C (10³ M) used to inhibit contraction is only slightly higher than the IC₅₀ for inhibition of agonist-induced current. However, we obtained inconsistent inhibition of NE-induced contraction with 10⁵ M niflumic acid. We therefore carried out our experiments at a concentration of 10⁴ M. This result is somewhat in contrast to data from Criddle et al. (12) showing that 10⁵ M niflumic acid produced 38% inhibition of a maximal response to NE (10⁶ M) and 55% inhibition of rat aortic ring contractions produced by brief exposure (30 s) to this same dose of NE. It seems unlikely that any of these agents interfere with -adrenergic activation of the tissue because NE-induced I_K,Ca is not blocked (19, 20). Of the agents used, DIDS appears to be both highly effective and specific for NE-induced contraction over KCl-induced contraction. The effectiveness of DIDS in these experiments may be enhanced by its ability to block not only Cl channels but also Cl transport (13). The mechanism by which A-9-C and niflumic acid inhibit KCl-induced contraction is not clear. We have shown previously (27) that neither DIDS nor niflumic acid inhibit voltage-dependent Ca²⁺ current directly. Tamoxifen, which has been shown to inhibit swelling-induced Cl currents nearly completely at 10⁵ M (41), had no effect on the response to NE. On the basis of the lack of effect of this drug in our system, these channels do not appear to play a role in NE-induced VSM activation.

If agonist-induced SR Ca²⁺ release is the primary trigger for activation of Cl current, we hypothesized that compounds that interfere with SR function such as ryanodine and CPA should prevent the potentiation produced by low-Cl buffer. Indeed, we saw a reduction in the potentiation of the initial peak of contraction. However, the increase in contraction at the 20-min time point was actually more dramatic in the presence of these drugs. The most likely explanation for this is that, in addition to altering the contractile response of the VSM, these compounds may also induce SR dysfunction in the endothelium, resulting in reduced nitric oxide (NO') production in response to NE. This issue is explored further in our companion paper (26), which defines the important effect of endothelial NO' on the response to NE in low-Cl buffer. Indeed, NO' appears to have more influence on agonist-induced Cl current during the maintained phase of contraction.

Although it is well established that a low-conductance I_Cl,Ca is activated by the release of intracellular Ca²⁺ from the SR (23, 32, 37), it remains unclear over what period of time these channels are active. In the rat portal vein, the channels remain open as long as intracellular Ca²⁺ is elevated in the range over which the channels are active (170-600 nM) (31). The lack of effect of low-Cl buffer on KCl-induced contraction would suggest that these channels are not activated by Ca²⁺ entering through voltage-gated channels. This is consistent with our previous patch-clamp results in the rabbit coronary artery (27), in which I_Cl,Ca was only activated by depolarizing pulses when SR function was intact and Ca²⁺-induced Ca²⁺ release could occur. In the presence of caffeine, inward Ca²⁺ currents failed to activate a significant amount of Cl current (27). This may represent an insufficient local Ca²⁺ level or may reflect a clustering of Cl channels in the vicinity of the SR.

In summary, these data demonstrate that Cl currents contribute significantly to VSM activation. This represents the first systematic documentation of the critical functional importance of Cl current to agonist-induced VSM contraction. The essential features of our findings are summarized in Fig. 11. Inward transport of Cl via Na⁺-K⁺-2Cl cotransport (bumetanide sensitive) and Cl/HCO₃ exchange (inhibited in HEPES buffer) produce an E_Cl that is more positive than the V_m. Activation of -receptors by NE causes the release of intracellular stores of SR Ca²⁺. This Ca²⁺ activates a Cl current, producing Cl efflux, depolarization, and activation of voltage-dependent Ca²⁺ channels. It is not clear whether these Cl channels can be activated by any other mechanism; however, our data suggest that the channels remain active long after the initial Ca²⁺ transient (Fig. 2). One might speculate that the high local levels of Ca²⁺ achieved by the release of the SR pool may be required to activate the current; however, lower levels may be adequate to maintain channel activity. Alternatively, a Cl conductance other than I_Cl,Ca may be active during the sustained phase of contraction. These findings have important implications for our understanding of how vasoconstriction is initiated and maintained.