Differential regulation of Ca²⁺-activated Cl⁻ currents in rabbit arterial and portal vein smooth muscle cells by Ca²⁺-calmodulin-dependent kinase

Abstract

1. Ca²⁺-activated chloride currents (I_Cl(Ca)) were recorded from smooth muscle cells isolated from rabbit pulmonary (PA) and coronary artery (CA) as well as rabbit portal vein (PV). The characteristics and regulation by Ca²⁺-calmodulin-dependent kinase II (CaMKII) were compared between the three cell types.

2. In PA and CA myocytes dialysed and superfused with K⁺-free media, pipette solutions containing fixed levels of free Ca²⁺ in the range of 250 nm to 1 μm evoked well sustained, outwardly rectifying I_Cl(Ca) currents in about 90 % of cells. The CaMKII inhibitor KN-93 (5 μm) increased the amplitude of I_Cl(Ca) in PA and CA myocytes. However, the threshold intracellular Ca²⁺ concentration for detecting this effect was different in the two arterial cell types. KN-93 also enhanced the rate of activation of the time-dependent current during depolarising steps, slowed the kinetics of the tail current following repolarisation, and induced a negative shift of the steady-state activation curve.

3. In PA myocytes, the effects of KN-93 were not mirrored by its inactive analogue KN-92 but were reproduced by the inclusion of autocamtide-2-related CaMKII inhibitory peptide (ARIP) in the pipette solution. Cell dialysis with constitutively active CaMKII (30 nm) significantly reduced I_Cl(Ca) evoked by 500 nm Ca²⁺.

4. In PV myocytes, I_Cl(Ca) was evoked by pipette solutions containing up to 1 μm free Ca²⁺ in less than 40 % of cells. Application of KN-93 to cells where I_Cl(Ca) was sustained produced a small inhibition (∼25 %) of the current in 70 % of the cells.

5. The present study shows that regulation of Ca²⁺-dependent Cl⁻ channels by CaMKII differs between arterial and portal vein myocytes.

Ca²⁺-dependent Cl⁻ channels (I_Cl(Ca)) are thought to play an important role in the depolarisation and vasoconstriction mediated by several hormones and neurotransmitters regulating peripheral vascular resistance and blood flow (for a review, consult Large & Wang, 1996). In vascular smooth muscle cells, low-conductance (∼1–3 pS) I_Cl(Ca) channels (Klöckner, 1993; van Renterghem & Lazdunski, 1993; Hirakawa et al. 1999) are thought to underlie a macroscopic current that exhibits the following biophysical properties: (1) activation by elevation of intracellular Ca²⁺ levels in the range of ∼180–600 nm (Pacaud et al. 1992); this can be evoked by Ca²⁺ entry through voltage-dependent Ca²⁺ channels (Pacaud et al. 1989; Lamb et al. 1994; Greenwood & Large, 1996) and reverse-mode Na⁺-Ca²⁺ exchange (Leblanc & Leung, 1995) or by Ca²⁺ released from the sarcoplasmic reticulum (SR) either in response to membrane receptor agonists (Byrne & Large, 1988; Amédée et al. 1990; Klöckner & Isenberg, 1991; van Renterghem & Lazdunski, 1993) or spontaneously (Wang et al. 1992); (2) voltage-dependent deactivation kinetics (Pacaud et al. 1989; Hogg et al. 1993; Lamb et al. 1994; Leblanc & Leung, 1995; Greenwood & Large, 1996); and (3) sensitivity to fenamate compounds such as niflumic acid (Pacaud et al. 1989; Lamb et al. 1994; Leblanc & Leung, 1995; Greenwood & Large, 1995).

Whether I_Cl(Ca) channels are uniquely sensitive to changes in [Ca²⁺]_i or are subject to regulation by enzymatic reactions has remained elusive. Ca²⁺-calmodulin-dependent kinase II (CaMKII) has been implicated, to various degrees, in the regulation of I_Cl(Ca) in different cell types. In Jurkat T lymphocytes I_Cl(Ca) evoked by the application of a Ca²⁺ ionophore was inhibited by a CaMKII inhibitor peptide and was stimulated by application of activated CaMKII alone (Nishimoto et al. 1991). Arreola et al. (1998) showed that CaMKII inhibitors suppressed the development of I_Cl(Ca) in T84 human tumour cells. However, in the same study, inhibition of CaMKII had no effect on I_Cl(Ca) evoked in parotid acinar cells (Arreola et al. 1998). In comparison only one study has investigated the putative influence of CaMKII on I_Cl(Ca) in smooth muscle preparations. Wang & Kotlikoff (1997) showed that in equine tracheal myocytes, the duration of caffeine-activated I_Cl(Ca) was briefer than the underlying Ca²⁺ transient in control conditions but had a similar time course to the Ca²⁺ transient in the presence of the CaMKII inhibitor KN-93 (Sumi et al. 1991). These data led the authors to propose that CaMKII-dependent phosphorylation attenuated the activation of the Cl⁻ channel. Thus, the aim of the present study was to investigate the possible role of CaMKII in the regulation of I_Cl(Ca) recorded from single cells isolated from two main arteries (pulmonary artery and coronary artery) and also portal vein, all vessels known to exhibit I_Cl(Ca) (see Lamb et al. 1994; Large & Wang, 1996; Yuan, 1997). In the present study we have evoked I_Cl(Ca) by using K⁺-free pipette solutions of known free [Ca²⁺]. As the activation of the Cl⁻ channel with this technique is not reliant upon either Ca²⁺ influx or Ca²⁺ release, any effect of CaMKII modulators on I_Cl(Ca) represents a direct effect on the channel protein or a closely associated regulator. Moreover, as the mechanism of current activation is standardised by this technique the characteristics of I_Cl(Ca) in the different smooth muscle cells can be compared. The results suggest that CaMKII may attenuate the activation of I_Cl(Ca) in arterial smooth muscle cells but seems to have less of an effect on I_Cl(Ca) in portal vein myocytes. Preliminary results of this study have been communicated previously (Leblanc & Lupien, 1997; Greenwood et al. 2000).

METHODS

Isolation of vascular myocytes

Cells were prepared from veins and arteries isolated from New Zealand White rabbits (2–3 kg) that had been killed by an overdose of the anaesthetic pentobarbitone injected into the ear vein (2 mg kg⁻¹) in accordance with British and Canadian regulations. The preparation of portal vein (PV) myocytes has been described in full previously (Greenwood & Large, 1996). Arterial myocytes were isolated from either the main pulmonary artery (PA) or the left descending and circumflex coronary arteries (CA). After dissection and removal of connective tissue, the PA and CA were cut into small strips and placed in a physiological salt solution (PSS) containing either 50 μm CaCl₂ (PA) or 10 μm CaCl₂ (CA) at 37 ^oC for 10 min. The PA was incubated with 1 mg ml⁻¹ papain, 0.15 mg ml⁻¹ dithiothreitol and 2 mg ml⁻¹ bovine serum albumin for 20 min and segments of CA were incubated with 1 mg ml⁻¹ collagenase type 1A and 0.05 mg ml⁻¹ protease type XXVII for 20-25 min at 35 ^oC. In all cases, cells were released by gentle agitation with a wide-bore Pasteur pipette. Cells were stored at 4 ^oC and used within 6 h.

Electrophysiology

All currents were recorded in the whole-cell voltage clamp mode using CED software and List amplifier or pCLAMP software and Axopatch-1D amplifier (Axon Instruments Inc., Foster City, CA, USA). Analysis was performed using the applicable software as well as Origin (Microcal, Northampton, MA, USA). In initial experiments, I_Cl(Ca) was recorded in PA and PV cells with the perforated patch configuration and was elicited as a consequence of Ca²⁺ entry through voltage-dependent Ca²⁺ channels (VDCCs). Cells were held at −50 mV (PA) or −70 mV (PV) and then depolarised to +10 mV (PV) or +20 mV (PA) for 200 ms (PA) or 500 ms (PV) to open VDCCs. I_Cl(Ca) was apparent as an outward current during the test pulse and a slowly declining inward current upon repolarisation to the holding potential (see Greenwood & Large, 1996 and Yuan, 1997 for full description). In the majority of experiments with PA cells the concentration of external Ca²⁺ was increased from 1.8 to 10 mm and 1 μm Bay K 8644 was included in the bathing solution to enhance the amplitude of Ca²⁺ current. For PV cells, external Ca²⁺ was 1.8 mm and 1 μm Bay K 8644 was also included to enhance the magnitude of VDCCs.

In the second series of experiments, I_Cl(Ca) was recorded using the conventional whole-cell configuration and was activated by pipette solutions containing free Ca²⁺ of known concentrations (see below). Cells were held at −50 mV and voltage-dependent characteristics were studied by initially depolarising to +70 mV for 1000 ms followed by repolarisation to −80 mV for 750 ms. Current-voltage relationships were constructed by stepping the cell from the holding potential of −50 mV to test potentials between −90 and +150 mV for 1500 ms. Current amplitudes were measured immediately after the initiation of the step following settling of the capacitative transient and at the end of the test pulse before repolarisation.

Solutions

For all perforated patch clamp experiments (Fig. 1 and Fig. 4C), the external solution (solution A) contained (mm): NaCl 126, Hepes 10, glucose 20, CaCl₂ 1.8, MgCl₂ 1.2; pH was set to 7.2 with NaOH. The pipette solution (solution B) for these experiments was (mm): CsCl 126, Hepes 10, glucose 20, MgCl₂ 1.2 and amphotericin B or nystatin (300 μg ml⁻¹); pH was adjusted to 7.2 with CsOH. In experiments in which the basic properties of I_Cl(Ca) in the three preparations of interest were investigated and compared (Figs 2–9), the external solution (solution C) contained (mm): NaCl 126, Hepes-HCl 10 (pH 7.4), glucose 20, CaCl₂ 1.8, MgCl₂ 1.2, TEA-Cl 10 and 4-aminopyridine 5. The pipette solution (solution D) contained (mm): TEA-Cl 20, CsCl 106, Hepes 5, BAPTA 10, MgATP 3, GTP disodium salt 0.2, MgCl₂ 0.42; pH was set to 7.2 by addition of CsOH. Free [Ca²⁺] was set by adding the appropriate amount of CaCl₂ determined by the EQCAL buffer program (Biosoft, Ferguson, MO, USA). In CA cells, we tested the effects of the CaMKII inhibitor KN-93 using either the above sets of solutions, or the following external (solution E) solution (mm): NaCl 130, NaHCO₃ 10, TEA-Cl 5.4, MgCl₂ 0.5, glucose 5.5, Hepes-NaOH 10 (pH 7.35) and CaCl₂ 1.8; and pipette (solution F) solution (mm): caesium aspartate 100, CsCl 20, TEA-Cl 20, Hepes-CsOH 5, EGTA 10, MgATP 5, GTP disodium salt 0.2; pH was set to 7.2 with CsOH. Free [Ca²⁺]_i was adjusted to fixed levels by adding the appropriate amounts of CaCl₂ and MgCl₂. Since the properties of I_Cl(Ca) and KN-93-induced effects on this current were quantitatively similar, all CA data from experiments with KN-93 were pooled together and expressed as the mean percentage change (Fig. 6B).

Figure 1. Effect of KN-93 on currents evoked by depolarisation recorded using the perforated patch configuration.

A, sample experiment performed in a portal vein myocyte (PV) illustrating the effects of 5 μm KN-93 on I_Cl(Ca) (I_Cl(Ca)tail) elicited by Ca²⁺ entry through voltage-gated Ca²⁺ channels (I_Ca). Currents were recorded using the nystatin perforated patch method (solutions A and B). Bay K 8644 (1 μm) was included in the external solution to enhance I_Ca. The top three traces are representative currents recorded before (a), during (b) and following (c) the application of 5 μm KN-93. Currents were elicited by repetitive steps to +10 mV applied every 30 s from a holding potential (V_h) of −70 mV. The lower graph plots the amplitude of peak inward I_Ca at +10 mV (•) and slow I_Cl(Ca) tail current at −70 mV (^) as a function of time for this particular experiment; measurements from the above three current tracings are correspondingly identified. B, representative currents evoked in a pulmonary myocyte (PA) by stepping the cell from −50 mV to +20 mV for 200 ms in control conditions and after application of 5 μm KN-93 for 5 min (amphotericin B perforated patch method). Notice that KN-93 inhibited both currents in the two preparations.

Figure 4. Characteristics of currents evoked in PV cells by 500 nm free Ca²⁺.

A, representative trace showing the voltage- and time-dependent current produced by depolarisation to +70 mV from a holding potential of −50 mV followed by repolarisation to −80 mV. Time-dependent currents at +70 and −80 mV were well fitted by a single exponential (shown by overlying line) with τ values of 221 and 86 ms, respectively, as indicated. Horizontal arrow indicates 0 pA amplitude level. B, ensemble of currents recorded in cells bathed in external solution containing NaCl (Ba) or NaSCN (Bb). Cells were initially stepped to +70 mV followed by a step to voltages between −100 and +100 mV in 20 mV increments. Dotted line shows the zero current level. C, in pulmonary (PA), coronary (CA) or portal vein (PV) myocytes, bar graph reporting the percentage of cells displaying a Ca²⁺-activated Cl⁻ current triggered by a fixed elevated intracellular level of Ca²⁺ (500 nm Ca_i²⁺ (solutions C and D), Ca²⁺ entry though L-type Ca²⁺ channels (Ca²⁺ current) (solutions A and B), or SR Ca²⁺ release mediated by cell exposure to 5 mm caffeine (solutions A and B; Caffeine flush). Total number of cells studied: PA, n = 200; CA, 500 nm Ca_i²⁺, n = 33; Ca²⁺ current, n = 17; Caffeine flush, n = 18; PV, 500 nm Ca_i²⁺, n = 403; Ca²⁺ current, n = 55; Caffeine flush, n = 22.

Figure 2. Characteristics of current evoked by 500 nm free intracellular Ca²⁺ in pulmonary artery smooth muscle cells.

A, representative current tracings recorded from a pulmonary artery (PA) myocyte in the absence (Control) and presence of 100 μm niflumic acid (NFA). From a V_h of −50 mV, voltage-dependent currents were evoked by 1 s step depolarisations to +70 mV, followed by 1 s return steps to −80 mV. The time-dependent activation and deactivation of the current at +70 and −80 mV, respectively, were well fitted by a single exponential (represented by a continuous line) with respective time constants of 255 and 123 ms. Arrow indicates the zero current level. B, ensemble of currents evoked by 500 nm free Ca²⁺ in PA myocytes bathed in an external solution containing either NaCl (a) or NaSCN (b), respectively. Cells were initially stepped to +70 mV followed by a step to voltages between −100 and +60 mV in 20 mV increments. Dotted line shows the zero current level. All currents were recorded using solutions C and D.

Figure 9. Inhibition of I_Cl(Ca) evoked by 500 nm free Ca²⁺ in PV myocytes by KN-93.

A, time course of the effect of 5 μm KN-93 on I_Cl(Ca) evoked by 500 nm free Ca²⁺. ^, instantaneous current upon stepping the cell from −50 to +70 mV. •, amplitude of the current at the end of the test step. KN-93 was applied for the period denoted by the horizontal bar. Insets: two families of I_Cl(Ca) currents recorded prior to (left) and after steady-state inhibition by KN-93 (right). These currents were elicited by 1.5 s steps from −90 to +130 mV applied in 20 mV increments from a holding potential of −50 mV. B, bar graphs reporting mean percentage change by 5 μm KN-93 (□) relative to the control (100 %; ▪) for the instantaneous (a) and late (b) currents recorded at +70 mV in PV myocytes dialysed with 500 nm(n = 7) or 1 μm(n = 5) free [Ca²⁺] as indicated. In a and b, * and † indicate a significant difference relative to control with P < 0.05 and P < 0.01, respectively. All currents were recorded using solutions C and D.

Figure 6. Effects of KN-93 on I_Cl(Ca) in CA myocytes.

A, typical experiment showing the effect of 5 μm KN-93 on I_Cl(Ca) currents recorded in a CA cell dialysed with 1 μm free Ca_i²⁺ (solutions E and F). From a V_h of −50 mV, currents were elicited by 1 s steps ranging from −90 to +130 mV applied in 20 mV increments (0.2 Hz). Arrows indicate the zero current level. B, bar graphs reporting mean percentage change relative to the control (100 %; ▪) of the instantaneous (a) and late (b) currents at +70 mV induced by 5 μm KN-93 (□) in CA myocytes dialysed with 500 nm(n = 11) or 1 μm(n = 14) free [Ca²⁺] as indicated (solutions C and D: n = 6 and n = 10 for 500 nm and 1 μm Ca_i²⁺, respectively; solutions E and F: n = 5 and n = 4 for 500 nm and 1 μm Ca_i²⁺, respectively). In a and b, * indicates a significant difference relative to control with P < 0.05 and P < 0.01, respectively.

In experiments where the reversal potential of the evoked current was determined, the theoretical chloride equilibrium potential was shifted to more negative potentials by replacement of the extracellular NaCl with NaSCN. Junction potentials were minimised by the use of a NaCl-containing agar bridge between bath and reference electrode. All enzymes and niflumic acid were purchased from Sigma Chemical Company (St Louis, MO, USA). Bay K 8644 and KN-93 were purchased from Calbiochem (La Jolla, CA, USA) and dissolved in dimethylsulphoxide (DMSO). At the final concentration used (0.1 %), DMSO had no detectable effect on membrane current. Autocamtide-2-related inhibitory peptide was purchased from Bachem Ltd (Norfolk, UK) and was added in powder form in the normal pipette solution at a concentration of 1 μm. Experiments were performed with control pipette solutions alternated with the same pipette solution containing the peptide. Activated CaMKII was kindly provided by Dr Brian Perrino (University of Nevada, Reno, NV, USA). The α subunit of rat brain CaMKII was expressed in baculovirus-infected Sf9 insect cells as a truncated, monomeric enzyme. The kinase was converted to a constitutively active form by autothiophosphorylation in the presence of 6 μm calmodulin and 0.4 mm ATPγS according to the method of Lledo et al. (1995). Autophosphorylation was confirmed by Western blot analysis using an antibody against autothiophosphorylated CaMKII. In this series of experiments, currents were recorded over three separate days using control pipette solutions (500 nm free Ca²⁺ only) alternated with pipette solutions containing 500 nm free Ca²⁺ and 30 nm autothiophosphorylated CaMKII.

Statistics

All data are the means of n cells ±s.e.m. The software Statistica for Windows 99 (version 5.5, Tulsa, OK, USA) was used to determine statistical significance between individual means using paired or unpaired Student's t test when two groups were compared, or one-way ANOVA test with a Duncan's post hoc multiple range test for repeated measure when more than two groups were analysed. P < 0.05 was considered to be statistically significant.

RESULTS

In initial experiments I_Cl(Ca) evoked by Ca²⁺ influx through voltage-dependent Ca²⁺ channels were recorded in portal vein (PV) and pulmonary artery (PA) smooth muscle cells using the perforated patch technique. From negative holding potentials (−70 mV to −50 mV), step depolarisation to +10 or +20 mV elicited a rapidly activating, nifedipine-sensitive inward Ca²⁺ current (see Fig. 1). Repolarisation to the holding potential resulted in a large, slowly declining inward tail current representing I_Cl(Ca) (see Greenwood & Large, 1996; Yuan, 1997). Using this method, both I_Ca and I_Cl(Ca) in PV and PA myocytes were stable for the period of the experiment under control conditions. Figure 1A shows the time course of changes of peak I_Ca and I_Cl(Ca) tail currents recorded in a PV cell before, during exposure and after washout of KN-93. Three selected membrane current recordings are shown at the top and are correspondingly labelled a, b and c in the lower graph. KN-93 inhibited both I_Ca and I_Cl(Ca)tail. The relative amount of block produced by KN-93 in this particular experiment was 68 and 45 % for I_Cl(Ca)tail and I_Ca, respectively. While I_Ca(L) continued to decline following washout of the drug, I_Cl(Ca)tail partially recovered. In five experiments, KN-93 respectively inhibited I_Ca and I_Cl(Ca)tail by 38 ± 6 and 51 ± 10 %. Similar effects were observed in PA cells (Fig. 1B). KN-93 (5 μm) inhibited I_Ca and I_Cl(Ca) by 69 ± 9 and 54 ± 2 %, respectively (n = 5). These effects were partially reversible upon washout. As the amplitude of I_Ca was crucial for activation of I_Cl(Ca), the inhibitory effect of KN-93 on I_Ca precluded any investigation into the possible modulation of Cl⁻ channels by CaMKII. Consequently, an alternative method was employed to activate I_Cl(Ca) in vascular myocytes that was less reliant upon Ca²⁺ homeostatic mechanisms.

Characteristics of I_Cl(Ca) activated by pipette solutions of known free Ca²⁺ in arterial myocytes

In PA smooth muscle cells, rupturing the cell membrane to achieve whole-cell recording mode with a pipette solution of 500 nm free Ca²⁺ evoked an inward current at the holding potential of −50 mV in 180 out of 200 cells studied. After an initial run-up that stabilised after 3–5 min of cell dialysis, this current was well maintained for at least 20 min and had distinct voltage-dependent characteristics that were similar to the characteristics of I_Cl(Ca) evoked by this technique in other non-smooth muscle cell types (e.g. Arreola et al. 1996; Nilius et al. 1997). Thus, depolarisation to potentials greater than +10 mV resulted in an instantaneous current that represents the channels open at −50 mV that was followed by the development of a time-dependent outward current presumably as more channels open at the depolarised potential. The development of the outward current followed an exponential time course as shown by Fig. 2 and the mean time constant (τ) for the development of the outward current at +70 mV was 229 ± 6 ms (n = 92). Upon repolarisation to -80 mV, the instantaneous current was followed by a slowly declining inward current that represents the closure of the channels opened by the previous depolarisation (see Arreola et al. 1996). The time course of the decaying current at −80 mV was also well fitted by a single exponential which represents the rate of channel deactivation; the mean τ at −80 mV was 99 ± 3 ms (n = 92). In comparison to the τ for the exponential fit of the outward current at depolarised potentials that showed no significant voltage dependence, the τ for deactivation of the channels at negative potentials was voltage dependent with τ changing e-fold for a 102 mV shift in voltage. Both the instantaneous current and current recorded at the end of the test step had a reversal potential (E_rev) close to the theoretical Cl⁻ equilibrium potential (E_Cl) under these conditions (−2 mV). Moreover, E_rev for both the instantaneous and late currents were shifted to more negative potentials when external NaCl was replaced by the more permeant anion thiocyanate (NaSCN; see Fig. 2B). For example the E_rev for the steady-state current was −5 ± 2 mV (n = 6) in normal NaCl-containing PSS and −43 ± 2 mV (n = 6) when the external solution contained NaSCN (an example of an experiment is shown in Fig. 2B). Application of niflumic acid, a relatively selective blocker of Ca²⁺-activated Cl⁻ channels in smooth muscle cells (Large & Wang, 1996), inhibited the current evoked by 500 nm free Ca²⁺ (Fig. 2A). These data suggest that in PA smooth muscle cells, pipette solutions containing Ca²⁺ clamped at 500 nm could evoke I_Cl(Ca) that had distinctive kinetics.

In CA cells dialysed with 500 or 1000 nm Ca²⁺, well-maintained currents with similar voltage-dependent characteristics to those observed in PA myocytes were also observed. As in PA cells, depolarisation from the holding potential of −50 mV elicited an instantaneous outward current followed by a slowly developed outward current. However, currents recorded in CA cells were significantly smaller than those registered in PA cells. As in PA cells, the current was sensitive to niflumic acid and relaxation during depolarising steps and upon repolarisation could be well fitted by a single exponential (the mean time constant (τ) at +70 and −80 mV was 302 ± 45 and 82 ± 6 ms, respectively, n = 4). Although not analysed in detail, the voltage dependence of the time constants of activation and deactivation of the current exhibited a similar pattern to that described in PA cells. In cells dialysed with 500 nm free Ca²⁺, the steady-state current measured at the end of 1.5 s pulses reversed near E_Cl with a mean E_rev of −6 ± 1 mV (n = 8) that was shifted to −41 ± 6 mV (n = 6) in external NaSCN. Thus, Cl⁻ currents with characteristic voltage-dependent properties could also be evoked in CA smooth muscle cells.

Current-voltage relationship and ca²⁺ dependence of I_Cl(Ca) in arterial cells

The current-voltage relationship of anion currents in PA and CA cells was determined by stepping the cell from the holding potential of −50 mV for 1.5 s to test potentials between −90 and +130 mV. Examples of typical currents in PA and CA cells dialysed with 500 nm Ca²⁺ are shown in Fig. 3Aa and Ba. Panels Ab and c and Bb and c show mean current-voltage relationships of the instantaneous (panels Ab and Bb) and late (panels Ac and Bc) currents, recorded with 250 or 500 nm free Ca²⁺ in PA cells, and 500 nm or 1 μm free Ca²⁺ in CA cells. In both cell types, the evoked current exhibited similar voltage-dependent characteristics. The current-voltage relationship of the instantaneous current displayed slight outward rectification while the late current exhibited more pronounced outward rectification. Consistent with the existence of Ca²⁺-activated Cl⁻ channels in these cells, the current was sensitive to changes in intracellular Ca²⁺ levels. For example, in PA cells late current amplitude at +70 mV evoked by 250 and 500 nm free Ca²⁺ was 4.6 ± 0.7 and 9.0 ± 0.2 pA pF⁻¹, respectively. In coronary myocytes, while the instantaneous current was similar in cells dialysed with 500 nm or 1 μm Ca²⁺, late current was significantly larger in cells dialysed with 1 μm relative to 500 nm Ca²⁺ for steps ≥+90 mV (Fig. 3Bb and c). Furthermore, examination of the currents in PA and CA cells exposed to solutions of identical composition (500 nm free Ca²⁺) indicates that PA myocytes generate larger currents than CA myocytes: this was largely due to the development of a larger time-dependent current during the depolarising step (Fig. 3Aa and Ba). Such a difference is also apparent when comparing the mean I–V relationships of the late current displayed in panels c of Fig. 3A and B; at +90 mV late current in PA and CA cells was 14.0 ± 1.7 and 7.0 ± 0.8 pA pF⁻¹, respectively.

Figure 3. Voltage and Ca²⁺ dependence of anion currents evoked in pulmonary and coronary arterial myocytes.

A and B show the results of experiments designed to examine the voltage dependence of Cl⁻ currents elicited in pulmonary (PA) and coronary (CA) arterial smooth muscle cells. Aa, representative family of currents evoked in a PA cell dialysed with 500 nm Ca²⁺. From a V_h of −50 mV, 1.5 s steps from −90 to +130 mV were applied in 20 mV increments. Ab and c, respectively, show the mean I–V relationships for the instantaneous (measured immediately after the capacitative current transient indicated by * in A) and late currents (indicated as ** in A) expressed as current density in pA pF⁻¹ evoked by an identical protocol to that shown in Aa from PA cells dialysed with either 250 nm Ca²⁺ (^) or 500 nm Ca²⁺ (•). For 250 nm Ca²⁺, n = 11; for 500 nm Ca²⁺, n = 10. B, same nomenclature to that of A except that all data were obtained in CA cells which were dialysed with either 500 nm (•; n = 8) or 1 μm free Ca²⁺ (□; n = 11). All currents were recorded using solutions C and D.

These results indicate that well sustained membrane anion currents exhibiting instantaneous and time-dependent characteristics, outward rectification, anion permeability, sensitivity to niflumic acid and activation by internal Ca²⁺ can be evoked in the majority of pulmonary and coronary arterial myocytes dialysed with 250, 500 or 1000 nm Ca²⁺. These currents share many properties with Ca²⁺-activated Cl⁻ currents recorded in smooth muscle (Large & Wang, 1996) and other cell types (Arreola et al. 1996; Nilius et al. 1997; Kuruma & Hartzell, 2000).

characteristics of I_Cl(Ca) activated by pipette solutions of known free Ca²⁺ in portal vein myocytes

In PV myocytes, pipette solutions containing 500 nm or 1 μm free Ca²⁺ evoked an inward current at the holding potential of −50 mV with similar voltage-dependent characteristics to the currents evoked in arterial smooth muscle cells. Depolarisation to +70 mV produced a time-dependent outward relaxation that could be fitted by a single exponential with a mean τ of 225 ± 12 ms (n = 47; Fig. 4a). Repolarisation to −80 mV produced an inward current that declined exponentially with a mean τ of 111 ± 8 ms. The evoked current was inhibited by niflumic acid, and similar to the typical experiment shown in Fig. 4B, had a reversal potential (E_rev) of −1 ± 2 mV (n = 5) with NaCl-containing external solution and E_rev was shifted to −40 ± 6 mV (n = 5) upon replacement of NaCl with NaSCN.

Although the currents elicited in PV cells had very similar characteristics to I_Cl(Ca) evoked in arterial cells the availability of I_Cl(Ca) in PV cells was considerably less than in the arterial cells. In PV cells, pipette solutions containing 500 nm free Ca²⁺ evoked a Cl⁻ current with time-dependent kinetics at positive potentials in only 131 out of 403 cells (33 % of total cells) compared with PA cells where pipette solutions containing 500 nm free Ca²⁺ elicited a Ca²⁺-activated Cl⁻ current in 180 out of 200 cells (Fig. 4C). Similarly when the perforated patch recording configuration was used I_Cl(Ca) evoked by Ca²⁺ influx through VDCCs or Ca²⁺ release from internal Ca²⁺ stores (caffeine flush) was only recorded in 36 % (20 out of 55 cells) and 50 % (11 out of 22 cells) of PV myocytes, respectively (see Fig. 4C). Thus, in PV myocytes I_Cl(Ca) has qualitatively similar voltage-dependent characteristics to I_Cl(Ca) elicited in arterial myocytes. However, I_Cl(Ca) in PV cells was less readily available than I_Cl(Ca) in arterial myocytes.

Effects of KN-93 in PA and CA smooth muscle cells

Application of 5 μm KN-93, a selective inhibitor of CaMKII (Sumi et al. 1991), to PA myocytes produced a marked enhancement of currents evoked by 500 nm free Ca²⁺ in 18 out of 22 cells (Fig. 5) and had no effect in the remaining cells. Of these 18 cells, KN-93 increased I_Cl(Ca) at −50 mV from −1.16 ± 0.14 to −2.35 ± 0.3 pA pF⁻¹ and enhanced the current recorded after 1.5 s depolarisation to +70 mV from 4.9 ± 0.7 to 11 ± 1 pA pF⁻¹ (Fig. 5a). However, application of KN-93 for 10 min had no effect on the current evoked by 250 nm free Ca²⁺ in six PA cells. The increase in amplitude produced by KN-93 in PA cells did not reverse after 5 min washout and was associated with a change in the kinetics of the current. In KN-93, τ at +70 mV decreased from 237 ± 10 to 186 ± 11 ms and at −80 mV τ increased from 98 ± 5 to 134 ± 6 ms (Fig. 5B, n = 16). KN-92, an inactive analogue of KN-93, had no effect on either the current amplitude or kinetics of I_Cl(Ca) evoked in PA or CA cells. In PA cells the mean late current at +70 mV was 11.0 ± 2.0 and 10.0 ± 1.5 pA pF⁻¹ in the absence and presence of 5 μm KN-92 (n = 7). The τ for the outward and inward time-dependent currents in the presence of KN-92 was 235 ± 7 and 90 ± 10 ms, respectively, which was not significantly different from the control τ (248 ± 5 and 99 ± 9 ms, respectively). Application of 5 μm KN-93 increased the instantaneous and late components of I_Cl(Ca) when CA cells were dialysed with 1 μm[Ca²⁺]_i (Fig. 6A and B) but not with 500 nm[Ca²⁺]_i (Fig. 6B).

Figure 5. Effects of KN-93 on I_Cl(Ca) in PA myocytes.

A, this panel shows representative current tracings from a PA myocyte dialysed with 500 nm Ca²⁺ which were recorded before (a), during (b) and following (c) the application of 5 μm KN-93. Currents were elicited by 1 s steps to +70 mV followed by a 0.75 s repolarising step to −80 mV; V_h= −50 mV. Superimposed on the evoked current in Ac is the current recorded after exposure to 100 μm niflumic acid (NFA). Values next to the current are the time constants for the exponential fit to the outward and inward currents. The lower graph plots the amplitude of the late current as a function of time for this particular experiment; measurements from the above three current tracings are correspondingly identified. B, bar graph reporting the mean time constants of activation at +70 mV and deactivation at -80 mV measured in the absence (Control) and presence of KN-93 (n = 17). * Significantly different from Control with P < 0.01. All currents were recorded using solutions C and D.

Figure 7 shows the effect of KN-93 on the voltage dependence of the chord conductance (G) of I_Cl(Ca) in PA myocytes dialysed with 500 nm Ca²⁺. This parameter was calculated using the following equation: G = I/(V − E_rev), where I is the late current, V the applied voltage and E_rev the reversal potential of the current. KN-93 increased the maximal conductance of I_Cl(Ca) and induced a leftward shift of the steady-state activation curve; the apparent half-maximum voltage (V_0.5) was 96 ± 11 mV in control and 66 ± 3 mV (n = 8) in KN-93. There was no significant change in the slope factor. KN-93 induced a similar negative shift of V_0.5 in seven CA myocytes dialysed with 1 μm Ca²⁺ (111 ± 6 mV in control and 73 ± 3 mV in KN-93, P < 0.05).

Figure 7. Effects of KN-93 on the steady-state activation curve of I_Cl(Ca) in PA cells.

The graph shows the chord conductance (calculated as I/(V − E_rev)) of the late current plotted against test potential for currents recorded with 500 nm free Ca²⁺ (•) and 500 nm free Ca²⁺ in the presence of 5 μm KN-93 (^). Each point is the mean ±s.e.m. of 9 cells. The continuous and dashed lines are least-squares Boltzmann fits to each set of data points. Currents were recorded using solutions C and D.

Effect of autocamtide-2-related inhibitory peptide in PA cells

Autocamtide-2-related inhibitory peptide (ARIP) is a 13 amino acid synthetic peptide that is highly selective for CaMKII with a K_i of 40 nm (Ishida et al. 1995). Experiments were performed comparing I_Cl(Ca) evoked with pipettes containing 500 nm Ca²⁺ with or without the inclusion of ARIP (1 μm). Inclusion of ARIP in the pipette solution produced a rapid and marked enhancement of current amplitude that was more pronounced at negative potentials. The instantaneous currents recorded with ARIP in the pipette solution 2 min after cell rupture were −5.9 ± 0.8 and 6.7 ± 0.6 pA pF⁻¹ at −90 and +90 mV, respectively (n = 9). These are compared with currents recorded with control internal solutions of −2.3 ± 0.4 and 3.9 ± 0.6 pA pF⁻¹, at −90 and +90 mV, respectively (n = 7). Figure 8A shows the mean I–V curves for instantaneous (Aa) and late (Ab) currents evoked by pipette solutions containing ARIP compared with control pipette solutions. Thus, a specific inhibitor peptide of CaMKII also increases I_Cl(Ca) in PA smooth muscle cells.

Figure 8. Effects of a peptide inhibitor and an autophosphorylated form of CaMKII on I_Cl(Ca) evoked in PA cells.

A, effects of cell dialysis with 1 μm autocamtide-2-related inhibitory peptide (ARIP), a specific inhibitor of CaMKII, on the voltage dependence of the instantaneous (a) and late (b) I_Cl(Ca) currents recorded from PA myocytes dialysed with 500 nm Ca²⁺, with (n = 9) and without (n = 7) ARIP. Differences between control and ARIP were significant for the instantaneous current (a) with P < 0.01 for voltages positive to +10 mV and negative to −10 mV, and for the late current (b) with P < 0.01 for voltages negative to −10 mV and with P < 0.05 for voltages between +10 and +50 mV. B, effects of a constitutively active autophosphorylated form of CaMKII on I_Cl(Ca) in PA myocytes exposed to 500 nm free internal Ca²⁺. The peptide was dialysed into the cell at a concentration of 30 nm. Currents in the absence or presence of the activated kinase were recorded after 3 min of cell dialysis in separate cells but from the same day of isolation. Shown are two families of I_Cl(Ca) currents recorded in control conditions (a) or following cell dialysis with the kinase (b;+AutoCaMKII). The voltage clamp protocol consisted of 1.5 s steps ranging from −90 to +150 mV applied in 20 mV increments from V_h= −50 mV. Bc and d, respectively, depict the mean I–V relationships for the instantaneous and late I_Cl(Ca) currents recorded in the absence (•, n = 7) or presence (^, n = 7) of the kinase. Differences between control and +AutoCaMKII were significant for the instantaneous current (c) with P < 0.01 for voltages positive to +30 mV and negative to −10 mV, and for the late current (d) with P < 0.01 for voltages positive to +10 mV and with P < 0.05 for voltages negative to −10 mV. In Ba and b, arrows indicate the zero current level. All currents were recorded using solutions C and D.

Effect of autothiophosphorylated camkii on I_Cl(Ca) in PA myocytes

The previous experiments have shown that in arterial myocytes, inhibition of endogenous CaMKII by KN-93 enhanced the magnitude of I_Cl(Ca). Consequently, internal perfusion of PA smooth muscle cells by constitutively active CaMKII should reduce the amplitude of I_Cl(Ca). Figure 8B shows a family of currents evoked by 500 nm free Ca²⁺ 3 min after achieving the whole-cell configuration in the absence (Fig. 8Ba) or presence of activated CaMKII (Fig. 8Bb). Both the instantaneous (Fig. 8Bc) and late (Fig. 8Bd) I_Cl(Ca) currents were significantly reduced in cells dialysed with the constitutively activated enzyme. The late current at −90 and +90 mV in the presence of activated CaMKII was −0.83 ± 0.1 and 5.7 ± 0.6 pA pF⁻¹(n = 7) compared with −3.7 ± 1.2 and 12.3 ± 2.2 pA pF⁻¹ under control conditions (n = 7). The kinetics of the time-dependent current were also modulated by activated CaMKII. The time constant of inward tail current relaxation at -50 mV following depolarisation to +110 mV was 124 ± 9 and 83 ± 4 ms in the absence and presence of autothiophosphorylated CaMKII, respectively. Consequently, constitutively active CaMKII modulated I_Cl(Ca) in PA myocytes in a manner opposite to that observed with inhibitors of endogenous CaMKII.

Effect of kn-93 on I_Cl(Ca) evoked in PV smooth muscle cells

In comparison with the data obtained from arterial myocytes, application of KN-93 for 10 min to sustained I_Cl(Ca) in PV cells dialysed with 500 nm free Ca²⁺ failed to enhance the amplitude of either the instantaneous or late current (n = 10). In all cells the instantaneous and late currents at +70 mV in control conditions were 1.6 ± 0.2 and 3.7 ± 0.6 pA pF⁻¹, respectively. In the presence of 5 μm KN-93 the instantaneous and late currents at +70 mV were 1.4 ± 0.2 and 3.05 ± 0.5 pA pF⁻¹, respectively. KN-93 had no significant effect on the time course of either the outward or inward time-dependent current components. The mean τ at −80 mV in the absence and presence of KN-93 was 105 ± 11 and 87 ± 9 ms, respectively (n = 10). Moreover, in 7 out of the 10 cells where KN-93 was applied, the evoked current was inhibited by KN-93 (an example is shown in Fig. 9A). A similar inhibitory effect of KN-93 was observed in 4 out of 5 PV cells (the other cell displayed no response) when I_Cl(Ca) was evoked by 1 μm free Ca²⁺. Under control conditions the mean amplitude of the current at the end of a 1000 ms step to +70 mV was 6.9 ± 0.6 and 4.8 ± 0.86 pA pF⁻¹ after a 4 min application of 5 μm KN-93 (n = 5). Mean percentage change induced by KN-93 on the instantaneous and late currents for these experiments are reported in Fig. 9B. In addition, I_Cl(Ca) evoked by 500 nm free Ca²⁺ in the presence of 1 μm ARIP in the pipette solution (3 min of cell dialysis) was not significantly different than I_Cl(Ca) evoked by 500 nm free Ca²⁺ alone. In four experiments, the instantaneous current elicited at +90 mV was 2.4 ± 0.7 and 3.1 ± 1.3 pA pF⁻¹ in cells dialysed with control and ARIP, respectively, and the late current was 4.1 ± 1.2 and 6.5 ± 3.8 pA pF⁻¹ in the absence and presence of ARIP, respectively. These data show that in comparison with arterial myocytes, I_Cl(Ca) in PV myocytes is not increased by inhibition of CaMKII.

DISCUSSION

In the present study we have investigated the possible modulation of I_Cl(Ca) in arterial and venous smooth muscle cells by CaMKII. In initial experiments I_Cl(Ca) was recorded using the perforated patch technique and was evoked by Ca²⁺ influx through voltage-dependent Ca²⁺ channels (I_Ca), as described in previous studies (Lamb et al. 1994; Greenwood & Large, 1996; Yuan, 1997). However, continued use of this method was precluded because application of the CaMKII inhibitor KN-93 rapidly inhibited I_Ca. As this conductance is essential for activation of I_Cl(Ca) any inhibition will markedly affect the activation of I_Cl(Ca). Consequently, we performed experiments where I_Cl(Ca) in arterial and venous myocytes was evoked by pipette solutions of known free [Ca²⁺]. With this technique the channel was activated directly without any reliance upon secondary mechanisms such as spontaneous or triggered Ca²⁺ release from the sarcoplasmic reticulum (SR), or Ca²⁺ influx through voltage-dependent Ca²⁺ channels or reverse-mode Na⁺-Ca²⁺ exchange (Large & Wang, 1996). Moreover, the effects of modulators on the Cl⁻ channel could be investigated without any equivocal effects on Ca²⁺ homeostatic mechanisms. Using this technique it was observed that inhibition of CaMKII by KN-93 or a specific inhibitor peptide (autocamtide-2-related inhibitory peptide, ARIP) enhanced the amplitude and steady-state availability of I_Cl(Ca) in the majority of arterial smooth muscle cells but had opposite effects on I_Cl(Ca) in PV myocytes. In addition, I_Cl(Ca) was undetectable in more than 50 % of the PV cells whereas the majority of arterial myocytes expressed this current. These data suggest that Ca²⁺-activated Cl⁻ channels in vascular smooth muscle cells share common voltage-dependent characteristics but are regulated to different degrees by CaMKII in PV cells compared with CA and PA myocytes.

Ca²⁺-activated Cl⁻ currents evoked by pipette solutions of known free [Ca²⁺]

Cl⁻ currents evoked by pipette solutions containing clamped [Ca²⁺]_i have been recorded in various non-smooth muscle cell types and exhibit distinct voltage-dependent characteristics (Ishikawa & Cook, 1993; Arreola et al. 1996, 1998; Nilius et al. 1997). In the present study, I_Cl(Ca) in all three cell types was evoked at the holding potential (−50 mV) and depolarisation to positive potentials elicited an instantaneous current followed by a slowly developing time-dependent outward current during the depolarising step similar to that seen in acinar cells and endothelial cells (Arreola et al. 1996, 1998). The amplitude of the instantaneous current upon depolarisation is a reflection of the chloride conductance at the holding potential. The exponential time course of I_Cl(Ca) upon depolarisation to positive potentials has been argued to represent an increase in open probability due to an increase in binding of Ca²⁺ ions to the channel and a slower rate of channel closure at depolarised potentials (see Arreola et al. 1996; Nilius et al. 1997; Kuruma & Hartzell, 2000). Repolarisation to a negative potential (e.g. −80 mV in the present study) produced an instantaneous current followed by an inward tail current reflecting closure of the channels opened during the depolarising step. This current also follows an exponential time course that is determined predominantly by the rate of channel closure (Arreola et al. 1996; Nilius et al. 1997).

Although the characteristics of I_Cl(Ca) evoked by free Ca²⁺ were similar in the three types of myocytes, Ca²⁺-activated Cl⁻ currents with typical time- and voltage-dependent properties were recorded less frequently in PV cells compared with arterial myocytes. Furthermore, in experiments using the perforated patch technique we failed to elicit I_Cl(Ca) in a considerable number of PV cells even though large voltage-dependent Ca²⁺ currents were activated, or the cells contracted promptly following the rapid mobilization of SR Ca²⁺ stores by caffeine. In comparison, application of caffeine or activation of Ca²⁺ currents evoked sizeable I_Cl(Ca) in the majority of arterial myocytes. These data suggest that the availability of Ca²⁺-dependent Cl⁻ channels for activation by Ca²⁺ is severely limited in PV myocytes compared with PA and CA myocytes. This may be due to a difference in expression of the channel especially as the current density is higher in the arterial cells compared with the PV or may be due to an unknown regulator. Another possibility is that the portal vein is a heterogeneous muscle tissue composed of longitudinal and circular layers of smooth muscle cells that could exhibit distinct phenotypes as seen in the gastrointestinal tract. Overall the data show that the direct activation of Cl⁻ channels by free Ca²⁺ in the pipette solution activates qualitatively similar currents that have markedly different availability.

Modulation of cl⁻ currents by CaMKII

CaMKII is a multimeric Ca²⁺- and calmodulin-dependent kinase of 300–600 kDa that phosphorylates serine and threonine residues and undergoes complex regulation by other kinases as well as autophosphorylation (Singer et al. 1996). The kinase is composed of one or more primary subunits (α, β, λ, δ) of which the λ and δ isoforms predominate in smooth muscle (Singer et al. 1996). CaMKII has been implicated in various regulatory roles including the modulation of myosin light chain kinase (Singer et al. 1996), Ca²⁺-dependent facilitation of Ca²⁺ channels (McCarron et al. 1992; Ledoux et al. 1999) and regulation of A-type K⁺ currents (Koh et al. 1999). In the present study, the amplitude of I_Cl(Ca) evoked by 500 nm free Ca²⁺ in PA or 1 μm free Ca²⁺ in CA smooth muscle cells was enhanced by application of the selective CaMKII inhibitor KN-93. However, KN-93 had no effect on I_Cl(Ca) evoked by pipette solutions containing 250 nm Ca²⁺ in PA cells, which is consistent with the notion that in smooth muscle, Ca²⁺ stimulates this enzyme with an IC₅₀ of approximately 690 nm (Abraham et al. 1996). The lower [Ca²⁺]_i sensitivity for modulation by CaMKII in CA relative to PA cells is interesting and might reflect intrinsic differences between the two cell types whereby the delicate balance between kinase and phosphatase activities would be shifted in favour of the latter. Moreover in the present study the structurally similar but inactive analogue KN-92 had no effect on I_Cl(Ca). The effects of KN-93 in PA smooth muscle cells were mirrored by inclusion of a highly specific inhibitor peptide of CaMKII, autocamtide-2-related inhibitory peptide (ARIP; Ishida et al. 1995), in the pipette solution. Thus, currents evoked by 500 nm Ca²⁺ after 2 min activation were significantly larger in the presence of ARIP compared with controls. Moreover, intracellular application of a low concentration of an autophosphorylated variant of CaMKII, which produces Ca²⁺-independent phosphorylation of serine/threonine residues on target proteins, resulted in the quick reduction of I_Cl(Ca) in PA cells. The rapid effects of the inhibitory peptide or the constitutively active form of CaMKII suggest that activation of this enzyme is exceedingly quick, which is supported by recent work by Rokolya & Singer (2000), who have shown that activation of CaMKII by histamine in swine carotid artery reaches a maximum within 10 s of agonist application. In addition, it is possible that CAMKII and the Ca²⁺-activated Cl⁻ channel are co-localised allowing precise regulation of the Cl⁻ channel by the kinase.

KN-93, ARIP and autoCaMKII influenced predominantly the instantaneous currents and late current in the negative range of potentials. KN-93 also produced a leftward shift in the steady-state activation curve in the two arterial cell types, and prolonged the decay of the inward tail current at negative potentials in PA myocytes. A similar trend was also apparent in CA cells. These data suggest that in PA and CA smooth muscle cells a rise in intracellular [Ca²⁺] activates Cl⁻ channels concomitant with the stimulation of CaMKII that leads to a suppression of the current amplitude although the threshold [Ca²⁺]_i for such an effect was apparently higher in CA than PA cells. The mechanism by which phosphorylation inhibits Cl⁻ channel activation is unknown but as the instantaneous currents are markedly affected by KN-93, ARIP and AutoCaMKII, then the presence of phosphate groups on the channel protein may occlude ion flux. Alternatively, Wang & Kotlikoff (1997) proposed that CaMKII-dependent phosphorylation mediated a rapid inactivation of Ca²⁺-activated Cl⁻ channels in tracheal myocytes. As CaMKII inhibition in the present study slowed channel deactivation and promoted currents in the negative range of potentials, then CaMKII-dependent phosphorylation may speed up the rate of channel closure analogous to channel inactivation or deactivation. Inhibition of this process by KN-93 consequently stabilises the channel in an open state conformation. The present study shows that in PA and CA myocytes, activation of I_Cl(Ca) is attenuated by CaMKII-dependent phosphorylation. The inability of KN-93 to modulate I_Cl(Ca) in a small number of PA and CA cells suggests that the degree of channel activation and therefore effectiveness of KN-93 may be dictated by additional regulators to CaMKII, such as phosphatases. Interestingly, Roeper et al. (1997) reported that the inactivation of Kv1.4 K⁺ channels was determined by a reciprocal regulation by CaMKII and Ca²⁺-dependent phosphatases. Thus in smooth muscle cells, the activity of Ca²⁺-activated Cl⁻ channels may be dependent on the underlying precise balance of kinases and phosphatases, which may vary among different smooth muscle cell types.

In comparison with the arterial cells, neither application of KN-93 nor intracellular perfusion with ARIP enhanced I_Cl(Ca) in PV cells regardless of whether the current was activated by 500 nm or 1 μm free Ca²⁺. Moreover in most PV cells KN-93 inhibited the current. Thus, in PV cells CaMKII-dependent phosphorylation appears to accentuate channel activation similar to Cl⁻ currents recorded from T lymphocytes (Nishimoto et al. 1991) and T84 colonocytes (Arreola et al. 1998). It remains to be determined if the differential effects of KN-93 in the different smooth muscle cell types investigated in this study are due to the regulation of the channel by different isoforms of CaMKII or whether CaMKII phosphorylates another unidentified regulatory element.

The present study shows for the first time that in PA, CA and PV smooth muscle cells, Ca²⁺-activated Cl⁻ currents evoked by a constant level of free Ca²⁺ have similar voltage-dependent characteristics. However, the availability of the channel and its regulation by CaMKII are markedly different between the arterial and venous smooth muscle cells studied in the present investigation. In arterial smooth muscle cells, activation of CaMKII could serve as an important negative feedback mechanism, especially at elevated [Ca²⁺]_i, to reduce the depolarising influence of I_Cl(Ca) channels on membrane potential and tone during stimulation by vasoconstricting agents. Indeed, niflumic acid-sensitive vasorelaxation was found to be inversely related to agonist concentration in phenylephrine-preconstricted pressurised rabbit mesenteric arterioles (Remillard et al. 2000). In contrast, portal vein smooth muscle cells fire action potentials spontaneously (Holman et al. 1968; Hume & Leblanc, 1989). Activation of I_Cl(Ca) channels by CaMKII through spontaneous Ca²⁺ discharge of the SR or during sympathetic stimulation could promote action potential firing by stimulation of spontaneous transient inward currents (STICs; Hogg et al. 1993), which are thought to play a role in the rythmic contractions of trachea (Janssen & Sims, 1994), or by modulation of the action potential waveform (Pacaud et al. 1989).