Swelling-activated cation channels mediate depolarization of rat cerebrovascular smooth muscle by hyposmolarity and intravascular pressure

Abstract

1. Increases in intravascular pressure depolarize vascular smooth muscle cells. Based on the attenuating effects of Cl⁻ channel antagonists, it has been suggested that swelling-activated Cl⁻ channels may be integral to this response. Consequently, this study tested for the presence of a swelling-activated Cl⁻ conductance in both intact rat cerebral arteries and isolated rat smooth muscle cells.

2. A 50 mosmol l⁻¹ hyposmotic challenge (300 to 250 mosmol l⁻¹) constricted rat cerebral arteries. This constriction contained all the salient features of a pressure-induced response including smooth muscle cell depolarization and a rise in intracellular Ca²⁺ that was blocked by voltage-operated Ca²⁺ channel antagonists. The hyposmotically induced depolarization was attenuated by DIDS (300 μm) and tamoxifen (1 μm), a response consistent with the presence of a swelling-activated Cl⁻ conductance.

3. A swelling-activated current was identified in cerebral vascular smooth muscle cells. This current was sensitive to Cl⁻ channel antagonists including DIDS (300 μm), tamoxifen (1 μm) and IAA-94 (100 μm). However, contrary to expectations, the reversal potential of this swelling-activated current shifted with the Na⁺ equilibrium potential and not the Cl⁻ equilibrium potential, indicating that the swelling-activated current was carried by cations and not anions. The swelling-activated cation current was blocked by Gd³⁺, a cation channel antagonist.

4. Gd³⁺ also blocked both swelling- and pressure-induced depolarization of smooth muscle cells in intact cerebral arteries.

5. These findings suggest that swelling- and pressure-induced depolarization arise from the activation of a cation conductance. This current is inhibited by DIDS, tamoxifen, IAA-94 and gadolinium.

The magnitude and distribution of tissue blood flow are controlled by an integrated network of resistance arteries (Segal & Duling, 1986). Under dynamic conditions, arterial tone is controlled by several factors including tissue metabolism, perivascular nerve activity, blood flow and intravascular pressure (Segal, 1994; Kurjiaka & Segal, 1995). Bayliss (1902) provided the first evidence that arteries constrict in response to elevation of intravascular pressure. This pressure-induced vasoconstriction (i.e. the myogenic response) depends in part on membrane potential depolarization of vascular smooth muscle and the associated Ca²⁺ influx through voltage-operated Ca²⁺ channels (Harder et al. 1987; Brayden & Wellman, 1989; Knot & Nelson, 1995, 1998). The identity of the ion channels that underlie myogenic depolarization has not yet been elucidated.

Pressure-induced depolarization could in theory arise from an augmentation of inward current (cation, Ca²⁺ or Cl⁻ channels) and/or inhibition of outward K⁺ current. It has been proposed that mechanically sensitive ion channels are key to the initiation of smooth muscle cell depolarization (Meininger & Davis, 1992; Setoguchi et al. 1997). Indeed, smooth muscle cation channels that are sensitive to stretch and positive pressure application have been identified (Davis et al. 1992; Setoguchi et al. 1997) and thus by default have been implicated as an integral component of the depolarizing response. Recent reports have noted, however, that certain Cl⁻ channels in smooth muscle are also sensitive to mechanical stimuli including cell swelling (Yamazaki et al. 1998; Greenwood & Large, 1998). Based on these patch clamp findings, and in combination with functional data showing that Cl⁻ channel antagonists block myogenic depolarization (Nelson et al. 1997), Nelson (1998) subsequently suggested that pressure-induced depolarization may in part depend on the activation of a swelling-activated Cl⁻ conductance.

The present study, therefore, tested whether or not a swelling-activated Cl⁻ current was functionally present in intact cerebral resistance arteries and in isolated smooth muscle cells. In keeping with a role for a swelling-activated Cl⁻ conductance, a hyposmotic challenge depolarized intact cerebral arteries and activated a whole-cell current, both of which were blocked by Cl⁻ channel antagonists. However, these electrical responses did not arise from an augmented Cl⁻ conductance but instead arose from activation of cation channels that were sensitive to Cl⁻ channel antagonists. Collectively, our findings suggest swelling- and pressure-induced depolarization in intact arteries result from the activation of cation and not Cl⁻ channels.

METHODS

Animals and tissues

Sprague-Dawley rats (12–16 weeks of age) were used in this study. Rats were killed with an intraperitoneal injection (2 ml) of pentobarbitone (65 mg ml⁻¹). The brain was carefully removed and placed in cold bicarbonate-buffered isosmotic (300 mosmol l⁻¹) physiological salt solution (PSS) containing (mm): NaCl, 120; KCl, 5; Hepes, 2.5; Na-Hepes, 2.5; CaCl₂, 2; MgCl₂, 1; glucose, 5; and mannitol, 50; pH 7.4. The osmolarity of all solutions used in this study was measured with a vapour osmometer (WesCor Inc). Cerebellar and basilar arteries were used for cell isolation whereas segments of the cerebellar and posterior cerebral arteries (∼150 μm diameter, 2–3 mm in length) were used for diameter and membrane potential measurements.

Intact resistance arteries

Intact cerebral arteries were studied using an arteriograph chamber (Living Systems Inc., Burlington, VT, USA) as previously described (Nelson et al. 1997). Endothelial cells were removed from all arteries by passing an air bubble through the lumen of the cannulated artery for 2 min; successful removal was confirmed by the loss of acetylcholine-induced dilatations.

Arterial diameter was monitored using an automated video dimension analyser system (Living Systems Inc.). Membrane potential (V_m) was measured by inserting a glass microelectrode filled with 3 m KCl (tip resistance, 60–80 MΩ) into the vessel wall. The criteria for a successful cell penetration were: (1) a sharp negative V_m deflection on entry; (2) a stable potential for at least 1 min after entry; and (3) a sharp positive V_m deflection upon removal. To monitor changes in cytosolic Ca²⁺, cerebral arteries were incubated in PSS with the membrane-permeable acetyoxymethylester analogue of fura-2 (2 μm) and Pluronic acid (0.5%). Following a 60 min loading period, cerebral arteries were mounted in an arteriograph. The relative change in cytosolic Ca²⁺ was monitored using Ion Optix microfluorimetry equipment (Milton, MA, USA) and was calculated from ratio imaging (excitation at 340 and 380 nm; emission at 510 nm), as previously described (Knot & Nelson, 1998).

Following a 30 min equilibration in warm (37°C) isosmotic PSS, cerebral arteries (pressurized to 15 mmHg) were exposed to a PSS that was rendered hyposmotic (250 mosmol l⁻¹) by removing mannitol. Once a peak response was attained, arteries were exposed to inhibitors of Cl⁻ channels (DIDS, 300 μm; tamoxifen, 1 μm), voltage-operated Ca²⁺ channels (nisoldipine, 1 μm) or cation channels (Gd³⁺, 30 μm). Arterial diameter and cytosolic Ca²⁺ responses were continuously monitored whereas V_m recordings were typically limited to 5–10 min. Due to this limited duration, multiple impalements were required to assess the V_m responses to hyposmotic challenge, to Cl⁻ channel antagonism, and to cation channel blockade. All V_m measurements were conducted in the presence of 1 μm nisoldipine.

The effects of Gd³⁺ (30 μm) and tamoxifen (10 μm) on myogenic depolarization were also examined in cerebral arteries pressurized to 60 mmHg. These experiments were performed in isosmotic (300 mosmol l⁻¹) bicarbonate-buffered saline containing (mm): NaCl, 119; KCl, 3; KH₂PO₄, 1.7; MgSO₄, 1.2; NaHCO₃, 25; EDTA, 0.02; CaCl₂, 1.6; and glucose, 11; pH 7.4 when equilibrated with 95% O₂ and 5% CO₂ (37°C).

Single smooth muscle cells

Single smooth muscle cells were enzymatically isolated from cerebellar and basilar arteries as previously described (Porter et al. 1998). Briefly, arteries were cut into 2 mm segments and placed in isolation solution of the following composition (mm): NaCl, 60; sodium glutamate, 85; KCl, 5.6; MgCl₂, 2; glucose, 10; CaCl₂, 0.1; and Hepes, 10 (pH 7.4). At the end of a 10 min equilibration (37°C), artery segments were placed in isolation solution (37°C) containing 1 mg ml⁻¹ albumin, 0.5 mg ml⁻¹ papain and 1 mg ml⁻¹ dithioerythritol. After 20 min in isolation solution, artery segments were placed for 8–10 min in a second isolation solution containing 0.1 mM CaCl₂ and 1 mg ml⁻¹ of collagenase type F and hyaluronidase. The tissue was subsequently washed twice (10 min each) in isolation solution and triturated with a polished wide-bore pipette. Cells were stored on ice and used the same day.

Conventional whole-cell patch clamp was used to measure ionic currents in cerebrovascular smooth muscle cells (Robertson et al. 1996). Recording electrodes (resistance, 4–7 MΩ) were pulled from borosilicate glass. Cells were voltage clamped and held at −50 mV; whole-cell currents were monitored every 2 min while voltage was slowly ramped (−100 to +100 mV, 0.167 mV ms⁻¹). A 1 m NaCl-agar salt bridge between the bath and the Ag-AgCl reference electrode was used to minimize offset potentials. Liquid junction potentials were measured as previously described (Neher, 1992) and were less than 2 mV with any solution change. Membrane currents were filtered at 1 kHz, digitized at 5 kHz and stored in a personal computer system for subsequent analysis. pCLAMP 7.0 and Clampfit 6.0 (Axon Instruments) were used to record and analyse membrane currents. Quantification of the whole-cell current characteristics was typically done at +100 mV since currents at negative potentials (−100 mV) were small. Cell capacitance averaged 17.2 ± 1.4 pF and was measured with the cancellation circuitry in the voltage-clamp amplifier (Axopatch 200A amplifier, Axon Instruments). Cells were excluded from analysis if cell capacitance shifted more than 0.5 pF during an experiment. All electrical recordings were performed at room temperature (22°C).

To examine the effects of osmolarity, cells were equilibrated in isosmotic solution (300 mosmol l⁻¹) for 6 min and then exposed to a hyposmotic bath solution for 6 min. To explore the ion selectivity of the currents, the composition of the bath was altered by lowering the Na⁺ and/or Cl⁻ concentration (see below) or by adding DIDS (300 μm), tamoxifen (1 μm), IAA-94 (100 μm) or Gd³⁺ (30 μm). The pipette solution contained (mm): NaCl, 110; Hepes, 5; EGTA, 5; ATP (Mg²⁺ salt), 5; and D-mannitol, 70 (300 mosmol l⁻¹). Isosmotic bath solution was potassium free and contained (mm): NaCl, 110; MgCl₂, 1.5; Hepes, 10; glucose, 10; and D-mannitol, 70. The bath was made hyposmotic (250 mosmol l⁻¹) or hyper-osmotic (350 mosmol l⁻¹) by altering the mannitol concentration. To reduce the bath Cl⁻ or Na⁺ concentration by 80 mm, sodium aspartate or N-methyl-D-glucamine-Cl (NMDG-Cl) was substituted for NaCl. In experiments where the NaCl concentration was reduced by 90 mm, the osmolarity of the hyposmotic bath was maintained with mannitol. Cell width was monitored using a video micrometer in a subset of experiments as an index of cell swelling.

Chemicals, drugs and enzymes

Chemical reagents, collagenase F, hyaluronidase, dithioerythritol, GdCl₃ and tamoxifen were obtained from Sigma Chemical Co. Papain, 4,4′-diisothiocyanatostilbene-2,2′-disulphonic acid (DIDS) and indanyloxyacetic acid (IAA-94) were purchased from Worthington Biochemical (Lakewood, NJ, USA), Calbiochem (La Jolla, CA, USA) and Research Biochemicals Inc. (Natick, MA, USA), respectively. DIDS, tamoxifen, IAA-94 and amphotericin B were dissolved in DMSO (final solvent concentration, 0.05–0.1%).

Statistical analysis

Data are expressed as means ±s.e.m. and n indicates the number of animals. Comparisons of data were made using Student's paired t tests. Data were considered to be significantly different at P < 0.05.

RESULTS

Hyposmotic challenge in intact cerebral arteries

To explore the possible role of osmotically sensitive Cl⁻ channels in pressure-induced depolarization, the effect of osmolarity on V_m and the diameter of isolated cerebral arteries (without endothelium) was investigated. A 50 mosmol l⁻¹ hyposmotic challenge (300 to 250 mosmol l⁻¹) elicited a sustained constriction (173 ± 9 to 124 ± 10 μm, n = 6, P < 0.05; Fig. 1A) in rat cerebral arteries pressurized to 15 mmHg. This vasoconstriction exhibited features that were identical to those seen when intravascular pressure is elevated, including smooth muscle cell depolarization (−51 ± 2 to −36 ± 2 mV, n = 5, P < 0.05; Fig. 1B) and a rise in cytosolic Ca²⁺ (340 nm/380 nm ratio, 0.68 ± 0.05 to 0.79 ± 0.05, n = 5, P < 0.05; Fig. 1C) that was blocked by nisoldipine (1 μm), a voltage-operated Ca²⁺ channel antagonist. Nisoldipine also blocked swelling-induced vasoconstriction (swelling response was 29 ± 8 and 2 ± 3 μm in the absence and presence of nisoldipine, respectively, n = 5, P < 0.05). Consistent with smooth muscle cell depolarization possibly arising from the activation of a swelling-activated Cl⁻ conductance, the addition of DIDS (300 μm, n = 3) or tamoxifen (1 μm, n = 3), two Cl⁻ channel antagonists, to the hyposmotic medium reversed the depolarization (Fig. 1D). In contrast, under isosmotic conditions tamoxifen (1 μm, n = 3) had no effect on V_m (300 mosmol l⁻¹, −50 ± 2 mV; 300 mosmol l⁻¹+ tamoxifen, −50 ± 1 mV).

Figure 1. Responses of intact cerebral arteries to hyposmotic challenge.

Representative diameter (A), V_m (B) and cytosolic Ca²⁺ (C) responses to a 50 mosmol l⁻¹ hyposmotic challenge (300 to 250 mosmol l⁻¹). Nisoldipine (1 μm), a voltage-operated Ca²⁺ channel antagonist, blocked the rise in cytosolic Ca²⁺. D, effects of Cl⁻ channel antagonists (DIDS, 300 μm; tamoxifen, 1 μm) on hyposmotically induced depolarization. Data are means +s.e.m.* Significant difference from 250 mosmol l⁻¹.

Hyposmotic challenge in isolated cerebral artery myocytes

We next used conventional whole-cell patch clamp recording techniques to determine whether swelling-activated Cl⁻ currents could be detected in single cerebral artery smooth muscle cells. In isotonic bath solution and in the absence of K⁺, a small outwardly rectifying current was observed (Fig. 2A). Reductions in bath osmolarity (300 to 250 mosmol l⁻¹) swelled smooth muscle cells and increased cell width (from 6.6 ± 0.5 to 10.9 ± 0.8 μm in 6 min, n = 5). The outwardly rectifying current (Fig. 2A and B) was augmented by a hyposmotic challenge and this change was sustained over time (Fig. 2C). The outward and inward currents (difference currents) activated by cell swelling were 51.6 ± 3.3 pA (+100 mV; n = 16) and −2.9 ± 0.6 pA (−100 mV; n = 16), respectively. Swelling responses were unaffected by the addition of K⁺ channel antagonists (tetraethylammonium (10 mm, n = 2), 4-aminopyridine (2 mm, n = 2) or glibenclamide (1 μm, n = 2)) to the bath or when CsCl (n = 2) was substituted for NaCl in the pipette. However, putative Cl⁻ channel antagonists (DIDS (300 μm), tamoxifen (1 μm) and IAA-94 (100 μm)) blocked the swelling-activated currents (Fig. 3) and a substantial portion of the basal current. Outward currents at +100 mV were reduced by 72.0 ± 8.2, 78.0 ± 4.3 and 65.4 ± 3.9 pA by DIDS (n = 6), tamoxifen (n = 7) and IAA-94 (n = 7), respectively. Inward currents (−100 mV) were also reduced by the Cl⁻ channel blockers (DIDS, 1.9 ± 0.6 pA, n = 6; tamoxifen, 2.8 ± 1.3 pA, n = 7; IAA-94, 2.0 ± 1.5 pA, n = 7). Further experiments revealed that pretreatment with DIDS (n = 3) reduced the isosmotic basal current and eliminated the ability of cells to respond to a hyposmotic challenge (data not shown). Such responses are consistent with the presence of a swelling-activated Cl⁻ conductance in these cells. Note that the hyposmotic challenge augmented the whole-cell current in 83% of all cells tested.

Figure 2. Hyposmotic challenge enhances an outwardly rectifying whole-cell current in isolated cerebrovascular smooth muscle cells.

A, ramp pulses (−100 to +100 mV, 0.167 mV ms⁻¹, holding potential =−50 mV, conventional whole-cell) were applied before and during a 50 mosmol l⁻¹ hyposmotic challenge. B, the swelling-activated difference current obtained by subtraction of the currents shown in A. C, whole-cell currents (+100 mV) in isosmotic (300 mosmol l⁻¹) and hyposmotic (250 mosmol l⁻¹) media are maintained over time. Following a 6 min period in isosmotic medium, cells were either maintained in this bathing solution or exposed to a 50 mosmol l⁻¹ hyposmotic challenge. Data are means ±s.e.m.* Significant difference from the isosmotic current.

Figure 3. Chloride channel antagonists inhibit swelling-activated currents.

Ramp pulses (see Fig. 2 legend) were applied before and during a hyposmotic challenge and in the presence of DIDS (300 μm; A), tamoxifen (1 μm; B), or IAA-94 (100 μm; C). D, peak outward (DIDS, n = 6; tamoxifen, n = 6; IAA-94, n = 7) current measured at +100 mV. Data are means ±s.e.m.* Significant difference from 300 mosmol l⁻¹; ** significant difference from 250 mosmol l⁻¹.

We next examined the ionic basis of the swelling-activated current in isolated smooth muscle cells by altering the ionic composition of the bath. Contrary to expectations, an 80 mM reduction in bath Cl⁻, which shifted the Cl⁻ equilibrium potential (E_Cl) from +1 to +31 mV, did not affect the reversal potential (V_rev) of the whole-cell current (Fig. 4A, n = 3). In contrast, when the NaCl concentration of the hyposmotic bath was reduced from 110 to 20 mM (Fig. 4B), V_rev shifted from −1 ± 1 to −38 ± 2 mV (n = 3) in close correspondence with the change in the Na⁺ equilibrium potential (E_Na, 0 to −42 mV). Such findings suggest that swelling activates a cation conductance that is sensitive to Cl⁻ channel antagonists.

Figure 4. The V_rev of swelling-activated current shifts with E_Na but not E_Cl.

Ramp pulses (see Fig. 2 legend) were applied before and during a hyposmotic challenge. E_Cl was shifted from +1 to +31 mV by lowering the Cl⁻ concentration from 113 to 33 mM in the hyposmotic bath. B, E_Na and E_Cl were shifted (E_Na from 0 to −42 mV; E_Cl from +1 to +40 mV) by lowering the NaCl concentration from 110 to 20 mM in the hyposmotic bath.

Basal current properties and comparison to the swelling-activated current

Our initial experiments (Fig. 3) demonstrated that Cl⁻ channel blockers reduce swelling-activated currents below basal levels, thereby suggesting that the channels underlying the swelling response are active under basal conditions. If this is correct, then the basal current should have properties that are similar to those of the swelling-activated current. Indeed, the next series of experiments revealed an outwardly rectifying basal current with a V_rev that shifted (from 0.4 ± 0.8 to −31.5 ± 0.3 mV, n = 4) in close correspondence with the changes in E_Na (Fig. 5A). Analogous to the swelling-activated current, the basal current displayed osmotic sensitivity, with hyper-osmotic solutions reducing peak outward (+100 mV) and inward (−100 mV) current by 39 ± 3% (n = 4) and 19 ± 4% (n = 4), respectively (Fig. 5B). Furthermore, we observed identical fractional block of the basal and hyposmotically activated currents by DIDS, tamoxifen and IAA-94 (Fig. 5C). Given this broad range of similarities (outward rectification, ion selectivity, osmotic sensitivity, fractional block), it appears that the cation channels underlying the swelling response also contribute to a sizable proportion of the basal current.

Figure 5. Basal current properties.

A, the basal current V_rev shifts with changes in E_Na. E_Na was altered (from 0 to −33 mV) by lowering the Na⁺ concentration from 110 to 30 mM in the isosmotic bath (300 mosmol l⁻¹). B, hyper-osmotic (350 mosmol l⁻¹) bath solution reduces the basal current. Inset, summary data (n = 4) showing peak outward currents (+100 mV) in isosmotic (□) and hyper-osmotic (▪) solutions. * Significant difference from 300 mosmol l⁻¹. C, fractional block of the basal and swelling-activated current by DIDS (300 μm), IAA-94 (100 μm) and tamoxifen (1 μm). Fractional block was calculated by dividing the leak-subtracted current (+100 mV) in the presence of the channel blocker by that in the absence of the blocker. Leak was estimated by linearly extrapolating the residual current at −100 mV (in the presence of blocker) and over the voltage range. Basal and swelling-activated currents were monitored in isosmotic (300 mosmol l⁻¹) and hyposmotic (250 mosmol l⁻¹) baths, respectively (n = 4–7 per group). Data are means ±s.e.m.

Gd³⁺ effects on responses to hyposmotic PSS

Following the identification of a swelling-activated cation current, we tested whether a known cation channel antagonist (Gd³⁺) would attenuate swelling-induced currents in isolated smooth muscle cells or alter the membrane depolarization induced by hyposmotic solutions. Indeed, the outwardly rectifying current activated by a hyposmotic challenge was blocked by Gd³⁺ (30 μm) as was a substantive proportion of the basal current (Fig. 6A and B). Correspondingly, pretreatment with Gd³⁺ (n = 3) reduced the isosmotic basal current and eliminated the ability of cells to respond to a hyposmotic challenge (data not shown). Finally, the depolarization of vascular smooth muscle cells in the intact artery in response to hyposmotic solution was reversed by Gd³⁺ (30 μm; Fig. 6C).

Figure 6. Gd³⁺ inhibits swelling-activated responses.

A, ramp pulses (see Fig. 2 legend) were applied before and after a hyposmotic challenge and in the presence of Gd³⁺ (30 μm). B, peak outward current at +100 mV (n = 7). C, effects of Gd³⁺ (30 μm) on hyposmotically induced depolarization (250 mosmol l⁻¹, n = 4) in intact cerebral arteries. Data are means ±s.e.m.* Significant difference from 300 mosmol l⁻¹; ** significant difference from 250 mosmol l⁻¹.

Effects of Gd³⁺ and tamoxifen on pressure-induced depolarization

Based on the evidence obtained in the present study supporting the presence of a mechanosensitive cation channel, a final series of experiments assessed the effects of cation channel inhibitors on the depolarizing response of intact arteries to increased intravascular pressure. As predicted if the swelling-activated channels participate in this response, we found that extracellular Gd³⁺ (30 μm) hyperpolarized intact cerebral arteries pressurized to 60 mmHg but not those maintained at 15 mmHg (Fig. 7). Similar to Gd³⁺ and in correspondence with the previously published effects of DIDS and IAA-94 (Nelson et al. 1997), 10 μm tamoxifen reversed pressure-induced depolarization at 60 mmHg (Fig. 7).

Figure 7. Gadolinium and tamoxifen block pressure-induced depolarization.

Cerebellar arteries were pressurized to 15 or 60 mmHg and V_m responses to Gd³⁺ (30 μm; 15 mmHg, n = 3; 60 mmHg, n = 6) and tamoxifen (10 μm, n = 5) were examined. Data are means ±s.e.m.* Significant difference from 60 mmHg.

DISCUSSION

Novel findings of this study

In the present study, we have provided evidence that a hyposmotic challenge constricts rat cerebral arteries by depolarizing the vascular smooth muscle cells and activating voltage-operated Ca²⁺ channels. The depolarization is due to the activation of gadolinium-sensitive, cation currents, which we have observed in single cerebral artery myocytes using the patch clamp technique. We have also provided the first demonstration of the inhibitory effect of gadolinium on depolarizations of intact cerebral arteries in response to both hyposmotic challenge and increased intravascular pressure. Gadolinium had no effect on the smooth muscle membrane potential of intact arteries held at low pressure, indicating that Gd³⁺-sensitive channels have very low activity in situ in the absence of the hyposmotic or pressure stimuli. These observations suggest a direct role for non-selective cation channels in myogenic depolarization and tone in cerebral resistance arteries.

Hyposmotic challenge

Elevated intravascular pressure constricts resistance arteries through an unidentified mechanotransduction pathway. The mechanism of vasoconstriction does, however, include a pressure-induced depolarization of the vascular smooth muscle cells (Harder et al. 1987; Knot & Nelson, 1998). We have previously shown that Cl⁻ channel blockers (i.e. DIDS and IAA-94) reverse this depolarization (Nelson et al. 1997) and consequently the constriction (Nelson et al. 1997; Robertson, 1998) associated with the intravascular pressure change. Several investigators have subsequently observed swelling-activated chloride channels in vascular smooth muscle (Yamazaki et al. 1998; Greenwood & Large, 1998). Together, these findings led us to hypothesize that pressure-induced, myogenic depolarization may depend on the activation of osmotically sensitive Cl⁻ channels (Nelson, 1998). Initial experiments in the present study were indeed consistent with the view that a swelling-activated Cl⁻ conductance is present. In intact cerebral arteries, a 50 mosmol l⁻¹ hyposmotic challenge elicited a depolarization that was reversed by Cl⁻ channel antagonists (e.g. DIDS and tamoxifen). Corresponding patch clamp experiments revealed the presence of basal and swelling-activated currents that were blocked by DIDS, tamoxifen and IAA-94. However, when the ionic composition of the bath was varied, it became evident that Cl⁻ channel antagonists were actually blocking a swelling-activated cation conductance. The inherent ability of DIDS, IAA-94 and tamoxifen to block the cation conductance was surprising. It is not, however, without precedent. For example, DIDS (∼100–300 μm) has been reported to block sodium currents in guinea-pig ventricular myocytes (Liu et al. 1998). IAA-94 (∼100–200 μm) has been noted by some (Doughty et al. 1998), although not all (Nelson et al. 1997) investigators, to inhibit voltage-operated Ca²⁺ channels. Similarly, low micromolar concentrations of tamoxifen (∼1–10 μm) can block ligand-gated cation channels (Allen et al. 1998). In the light of the preceding results, and given that a swelling-activated Cl⁻ conductance was not observed in cerebrovascular myocytes, we conclude that a swelling-activated Cl⁻ conductance is unlikely to be associated with swelling- or pressure-induced depolarization. Instead, our results indicate that the depolarizing responses to cell swelling and increased intravascular pressure are due to the activation of cation channels.

Gadolinium has been used extensively to block cation channels in a broad range of excitable cells (Yang & Sachs, 1989; Wellner & Isenberg, 1993; Setoguchi et al. 1997; Nakazawa et al. 1997). In support of a link between the cation current activation and depolarization, Gd³⁺ reversed the depolarization induced by hyposmolarity in intact arteries and inhibited swelling-activated cation currents in isolated arterial smooth muscle cells (present study). Since Gd³⁺ does not appear to block swelling-activated Cl⁻ currents in vascular smooth muscle (Yamazaki et al. 1998), these findings further dissociate hyposmotically induced depolarization from the activation of a mechanically sensitive Cl⁻ current.

Role for cation channels in pressure-induced depolarization

Although mechanically sensitive smooth muscle cation currents have been described in previous patch clamp investigations (Davis et al. 1992; Setoguchi et al. 1997; Bae et al. 1999), a link between the activation of these channels and depolarization of a pressurized resistance artery has not yet been established. To assess the functional importance of the swelling-activated cation current, we examined the effects of Gd³⁺, a potent blocker of swelling-activated cation currents, on the depolarizations induced by increased intravascular pressure. Consistent with a prominent role for swelling-activated cation channels, pressure-induced depolarizations were reversed by Gd³⁺. A broader comparison of the pharmacological fingerprint which characterizes the swelling-activated cation current and the V_m responses further substantiates a link between channel activation and pressure-induced depolarization. Specifically, swelling-induced cation currents are blocked by Gd³⁺ as well as by Cl⁻ channel antagonists like DIDS, tamoxifen and IAA-94. Results within this and our previous study (Nelson et al. 1997) also demonstrate that these same antagonists reverse pressure-induced depolarizations.

It is evident from the present findings that swelling-induced inward currents are small at negative potentials, particularly within the physiological V_m range (−30 to −60 mV). Consequently, in order for this inward current to depolarize cerebral arteries, a high input resistance is required. This appears to be the case, with studies reporting that smooth muscle input resistance ranges from 5 to 20 GΩ (Nelson & Quayle, 1995). Past studies have verified that high input resistance enables small conductances to regulate smooth muscle V_m. For example, outward current through inwardly rectifying K⁺ (Kir) channels is, in general, so small that it is rarely documented (Quayle et al. 1993; Robertson et al. 1996; Knot et al. 1996). Despite its small magnitude, this outward current has been reported to hyperpolarize intact cerebral arteries by as much as 10–15 mV (Knot et al. 1996).

It is important to recognize that pressure-induced depolarization might arise from an augmentation of inward current and/or an inhibition of outward K⁺ conductance. Although the present study supports a role for cation channels, it does not address whether changes in intravascular pressure influence K⁺ channel conductance. Given that inward rectifying, ATP-sensitive, voltage-dependent and Ca²⁺-activated K⁺ channels are all to some degree involved in regulating the membrane potential of pressurized arteries (Nelson et al. 1995; Knot & Nelson, 1995; Knot et al. 1996; Knot & Nelson, 1998), it is conceivable that pressure-induced depolarization in part arises from the inhibition of one or several of these K⁺ channels as well as from the activation of non-selective cation channels.

Conclusions

This study tested whether swelling-induced Cl⁻ currents are present in the cerebral arteries and whether such currents might be involved in myogenic depolarization. Our findings revealed that a hyposmotic challenge elicited a functional response that (1) was similar to the myogenic response, and (2) was blocked by Cl⁻ channel antagonists. Contrary to expectations, patch clamp experiments identified a swelling-activated cation conductance that was inhibited by Cl⁻ channel antagonists. In the light of these findings as well as subsequent experiments demonstrating the inhibitory effects of Gd³⁺ in cerebellar and cerebral arteries, we conclude that cation channel activation is likely to be essential for swelling- and pressure-induced depolarization.