Calcineurin A (cid:1) but Not A (cid:2) Augments I Cl(Ca) in Rabbit Pulmonary Artery Smooth Muscle Cells*

Activation of Ca 2 (cid:3) -dependent Cl (cid:4) currents (I Cl(Ca) ) increases membrane excitability in vascular smooth mus- cle cells. Previous studies showed that Ca 2 (cid:3) -dependent phosphorylation suppresses I Cl(Ca) in pulmonary artery myocytes, and the aim of the present study was to de-termine the role of the Ca 2 (cid:3) -dependent phosphatase calcineurin on chloride channel activity. Immunocyto- chemical and Western blot studies with isoform-specific antibodies revealed that the (cid:1) and (cid:2) forms of the CaN catalytic subunit are expressed in PA cells but that only the (cid:1) variant translocated to the cell periphery upon a rise in intracellular [Ca 2 (cid:3) ]. I Cl(Ca) evoked by pipette solutions containing a [Ca 2 (cid:3) ] set at 500 n M was considerably larger when the pipette solution included constitu- tively active CaN containing the (cid:1) catalytic isoform. This stimulatory effect was lost by boiling the enzyme or by the inclusion of a specific CaN inhibitory peptide and was not shared by the inclusion of the (cid:2) of cation ATP-sensitive K (cid:3) channels in cells on Cl (cid:2) we of shows the Cl(Ca) Ca 2 channels or evoked by pipette n M free Ca 2 (cid:3) . Moreover, CsA was to be ineffective on currents recorded from cells where the control currents were relatively small. These data suggest that the activity of CaN may vary among of but that the level of activity appears to be a crucial determinant of Cl (cid:2) current amplitude. Taken together with our previous findings with CaMKII inhibitors in PA cells, implicate a contemporaneous of the channel protein or a closely associated regulatory subunit by phosphorylation dephosphorylation A similar contra-parallel regulation also exists in coronary artery smooth muscle cells, be in Ca 2 (cid:3) sensitivity

Smooth muscle cells actively accumulate chloride resulting in an equilibrium potential ϳ30 mV less negative than the resting membrane potential (1). Consequently, the activation of Cl Ϫ channels leads to Cl Ϫ efflux and membrane depolarization. Ca 2ϩ -dependent Cl Ϫ currents (I Cl(Ca) ) 1 have been recorded from a wide number of smooth muscle cells where they have been implicated in agonist-induced and spontaneous contractions (2,3). In all of the smooth muscle cells, the generation of I Cl(Ca) has an obligatory requirement for increased intracellular [Ca 2ϩ ] with a threshold of ϳ200 nM (4,5). However, in tracheal and arterial smooth muscle cells, the activation of I Cl(Ca) was augmented by inhibitors of Ca 2ϩ /calmodulin-dependent kinase, (5,6) and internal dialysis with constitutively active CaMKII suppressed I Cl(Ca) in pulmonary artery myocytes (6). These data revealed an inhibition of channel activity contemporaneously with the generation of the current.
However, these studies did not take into account other possible Ca 2ϩ -dependent pathways. The aim of the present study was to assess whether I Cl(Ca) activity in rabbit pulmonary artery (PA) myocytes is also influenced by Ca 2ϩ -dependent dephosphorylation. Calcineurin (CaN) is a heterodimeric serine/ threonine protein phosphatase that is involved in a number of cellular responses (7)(8)(9). CaN is composed of a catalytic subunit (CaNA) that is activated by Ca 2ϩ binding to its regulatory subunit (CaNB) and by the binding of the Ca 2ϩ /calmodulin complex (Ca 2ϩ /CaM). The catalytic subunit of calcineurin (CaNA) exists in three distinct isoforms (␣, ␤, and ␥), each encoded by a separate gene, and isoform-specific substrates have been identified (10). Western blot analysis indicated that the ␣ and ␤ isoforms are expressed in pulmonary arteries but that only the CaNA-␣ isoform translocates from the cytosol to the membrane following an elevation of intracellular Ca 2ϩ concentration. Intracellular dialysis with constitutively active CaNA-␣ produced a large enhancement of I Cl(Ca) elicited by 500 nM Ca 2ϩ -containing pipette solutions that was attenuated by the co-dialysis with a peptide inhibitor of CaN. In comparison, CaNA-␤ had no stimulatory effect on I Cl(Ca) in PA myocytes. Consequently, through the use of recombinant calcineurin, we show that the regulation of I Cl(Ca) by this phosphatase is highly dependent on the calcineurin isoform. This study, in association with our earlier work with CaN inhibitors in coronary artery smooth muscle cells (11), reveals the crucial influence of calcineurin on calcium-activated chloride channels and highlights the complex pathways that govern Ca 2ϩ -dependent Cl Ϫ activity in vascular myocytes. * The work was supported in part by grants from the CIHR (Grant MOP-10863), the Western Affiliate of the American Heart Association (Grant 0355060Y), and COBRE (Grant NCRR 5 P20 RR15581) of the University of Nevada School of Medicine (Reno, NV) (to N. L.). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18

MATERIALS AND METHODS
Single Cell Electrophysiology-Cells were prepared from the main and second branch pulmonary arteries isolated from New Zealand White rabbits (2-3 kg) as described previously (6,(11)(12)(13). After isolation, cell were stored in a low Ca 2ϩ -physiological salt solution at 4°C and used within 6 h. The composition of the modified physiological salt solution was as follows: NaCl (120 mM); NaHCO 3 (25 mM) (pH 7.4 after bubbling with 95% O 2 , 5% CO 2 gas); KCl (4.2 mM); KH 2 PO 4 (1.2 mM); MgCl 2 (1.2 mM); CaCl 2 (0.05 mM); and glucose (11 mM). I Cl(Ca) were recorded predominantly in the whole-cell voltage clamp mode and were evoked directly by pipette solutions containing free Ca 2ϩ set at 500 nM. For experiments in which I Cl(Ca) was elicited by 500 nM Ca 2ϩ , the external solution contained: NaCl (126 mM); Hepes-NaOH (10 mM), pH 7.4; glucose (20 mM); CaCl 2 (1.8 mM); MgCl 2 (1.2 mM); and tetraethylammonium chloride (10 mM). The pipette solution contained: tetraethylammonium chloride (20 mM); CsCl (106 mM); Hepes (5 mM); BAPTA (10 mM); MgATP (3 mM); GTP (0.2 mM); and MgCl 2 (0.42 mM) and pH was set to 7.2 by the addition of CsOH. Free [Ca 2ϩ ] was set by adding an appropriate amount of CaCl 2 (7.3 mM) determined by the EQCAL buffer program (Biosoft, Ferguson, MO) and was independently verified using a Ca 2ϩ -sensitive electrode (Thermo Orion, Model 93-20, Beverly, MA) and calibrated Ca 2ϩ solutions available from a commercial source (World Precision Instruments, Inc., CALBUF-2, Sarasota, FL). Sustained Ca 2ϩ -activated Cl Ϫ currents generated by this technique have been characterized fully in previous studies (6,11,12). The main voltage step protocol used in these experiments was the same as that used in earlier studies to characterize I Cl(Ca) . In the perforated patch experiments, the pipette solution contained CsCl (126 mM), Hepes (5 mM), EGTA (5 mM), amphotericin (300 g ml Ϫ1 from a 60 mg ml Ϫ1 stock in dimethyl sulfoxide), and pH set to 7.2 by the addition of CsOH. For those experiments, the external solution was identical to that used in experiments in which I Cl(Ca) was evoked by 500 nM Ca 2ϩ as described above. All of the enzymes, MgATP, BAPTA, ionomycin, and ML-7, were purchased from Sigma. Cyclosporin A was purchased from Calbiochem (EMD Biosciences, San Diego, CA), and calcineurin inhibitory peptide (CaN-AIP) was from Biomol Research Laboratories (Plymouth Meeting, PA).
Synthesis of Constitutively Active Calcineurin-Constitutively active calcineurin isoforms were created by introducing stop codons into the cDNA for the catalytic subunit, CaNA, causing the translated CaNA subunits to truncate immediately C-terminal of the CaM-binding domain and delete the auto-inhibitory domain. All methodologies for cDNA manipulation, baculovirus screening, and purification of CaN using monolayer cultures of Sf21 cells have been described previously (7,10,14).
Immunofluorescence and Confocal Imaging-The immunocytochemical detection of CaN isoforms in single PA myocytes was performed as described for coronary artery myocytes by Ledoux et al. (11) using polyclonal goat anti-calcineurin A-␣ and A-␤ antibodies (Santa Cruz Biotechnology) both at a dilution of 1:50. These antibodies are reported by Santa Cruz Biotechnology to recognize a C-terminal epitope in human, rat, and mouse CaNA-␣ or CaNA-␤, respectively. However, the amino acid sequences of rabbit CaNA-␣ (GenBank TM accession number AAN23152) and CaNA-␤ (GenBank TM accession number AAN23153) are identical to the human CaNA-␣ and CaNA-␤ amino acid sequences, so it is very possible that these antibodies specifically recognized rabbit CaNA-␣ and CaNA-␤ in our experiments. Cells were imaged at rest and after stimulation with ionomycin and 500 nM Ca 2ϩ to raise intracellular [Ca 2ϩ ] and create an internal environment similar to the conditions of the electrophysiological experiments. Contraction of the myocytes was prevented by incubation with the myosin light chain kinase inhibitor, ML-7 (3 M). The primary antibodies were diluted in phosphate-buffered saline containing 1% normal donkey serum and 0.04% Triton X-100. Negative control experiments were performed by repeating the above steps in the absence of primary antibodies. Coverslips containing the cells were washed three times in phosphate-buffered saline and exposed to a Cy5-coupled anti-mouse antibody (Alexa 647; Molecular Probes) and a TRITC-coupled anti-rabbit antibody (Alexa 546; Molecular Probes) at a dilution of 1:400 for 1 h in the dark at room temperature. Solutions of both secondary antibodies were prepared in goat and diluted in 1% normal donkey serum and 0.04% Triton X-100 (Jackson Immunoresearch Laboratories, Inc.). The bar graphs shown in ]. Quantification of data from the experiments shown in A and B is shown in panels C and D. Panel C represents the mean results of experiments where CaNA-␣ and CaNA-␤ antibodies were applied individually (CaNA-␣:Control (n ϭ 4), 500 nM Ca 2ϩ (n ϭ 6) and CaNA-␤:Control (n ϭ 5), 500 nM Ca 2ϩ (n ϭ 5)). Panel D shows the mean Ϯ S.D. data from dual labeling experiments (Control (n ϭ 4), 500 nM Ca 2ϩ (n ϭ 5)). For panels C and D, the open and solid bars refer, respectively, to myocytes exposed to normal external solution (1.8 mM Ca 2ϩ ) labeled with the CaNA-␣ or CaNA-␤ antibodies, whereas the right-slanted and left-slanted hatched bars correspond to myocytes exposed to 500 nM of free Ca 2ϩ solution containing 80 nM ionomycin and 3 M ML-7, which were dually labeled with CaNA-␣ and CaNA-␤ antibodies.
of an arbitrary region of the cell located outside the nuclear region. Fluorescence intensity of ϳ5-10 pixels wide on the two sides of the membrane spanned by the line scan was averaged and normalized to averaged fluorescence intensity in the cytoplasm.
Western Blot Analysis of Rabbit Pulmonary Artery Smooth Muscle-50 g of protein from the 1000 ϫ g supernatant of homogenized rabbit PA smooth muscle was separated by 10% SDS-PAGE and electroblotted onto nitrocellulose sheets. Protein concentrations were determined by the Bradford assay using bovine ␥-globulin as standard. The blots were probed with Santa Cruz Biotechnology goat anti-CaNA-␣ or goat anti-CaNA-␤ antibodies (1:10,000 dilution) followed by horseradish peroxidase-conjugated rabbit anti-goat IgG (1:50,000 dilution) antibodies. Immunodetection was carried using ECL Advance from Amersham Biosciences, and the TIFF images were collected with a CCD camera imaging system (Labworks, UVP Inc.). Densitometry was carried out using Un-Scan-It (Silk Scientific).
Statistics-All of the data are the mean Ϯ S.E. of n cells from at least two different animals. Significance was taken with p values below 0.05.

Endogenous Expression of Distinct Isoforms of Calcineurin
A-Western blot analysis was performed using tissue lysates derived from endothelium-denuded pulmonary arteries. Antibodies selective for CaNA-␣ and CaNA-␤ revealed that both isoforms were expressed, and a comparison with purified CaN standards showed that ϳ0.56 g of CaNA-␣ and 0.3 g of CaNA-␤ were present (Fig. 1). Consequently, PA smooth muscle expresses both the ␣ and ␤ variants of CaNA but the ␣ appears to be slightly more abundant. The use of a CaN antibody that was nonspecific for these isoforms revealed that this enzyme translocated to the membrane of coronary artery myocytes upon a rise of intracellular [Ca 2ϩ ] (11). We used antibodies that were specific for the ␣ and ␤ isoforms of CaNA to examine the cellular distribution of the individual isoforms in freshly dissociated PA myocytes under resting conditions and after raising intracellular [Ca 2ϩ ]. Under control conditions, both isoforms appeared evenly distributed between the cytosol and membrane ( Fig. 2A). After the cells were exposed to a medium containing 80 nM ionomycin, 3 M ML-7 (to inhibit myosin light chain kinase and thus contraction; see Ref. 11) and 500 nM Ca 2ϩ designed to raise [Ca 2ϩ ] i , CaNA-␣ translocated toward the plasma membrane, whereas CaNA-␤ remained for the most part distributed homogenously throughout the cytoplasm and membrane ( Fig. 2A). Consequently, the ratio of optical density of immunofluorescence intensity for CaNA-␣ labeling at the membrane over cytosol was ϳ2.5 regardless of whether the immunodetection involved single (Fig. 2C) or dual (Fig. 2D) antibody experiments. The preferential localization of CaNA-␣ was not observed when the extracellular solution did not contain Ca 2ϩ . These results reveal that in PA cells CaN translocation in response to an elevation of intracellular Ca 2ϩ concentration is isoform-specific.
Effect of CaN Inhibition on Ca 2ϩ -activated Cl Ϫ Channels in PA Myocytes-The influence of endogenous CaN-mediated dephosphorylation on I Cl(Ca) was investigated using the CaNspecific blocker, cyclosporin A (CsA), which is a neutral lipophilic cyclic undecapeptide that binds to cyclophilin A to form a complex that suppresses CaN activity by interfering with the active site of the catalytic domain (8,9). Initial experiments were performed using the perforated patch variant of the whole-cell recording configuration, and I Cl(Ca) was elicited by promoting Ca 2ϩ influx through voltage-dependent calcium channels as characterized in previous studies on various types of smooth muscle cell (e.g. [15][16][17]. Calcium channels were opened by depolarization from the holding potential of Ϫ60 mV, and I Cl(Ca) was manifest as an outward current at test potentials positive to the theoretical chloride equilibrium (ϳ0 mV) and as a slowly declining inward current upon repolarization to the holding potential (Fig. 4A). The application of 2 M CsA for 180 s had a small insignificant effect (p ϭ 0.09) on the voltagegated calcium current at ϩ10 mV (Fig. 3, A and B) but pro- duced a 39 Ϯ 10% inhibition of the inward Cl Ϫ tail current at Ϫ60 mV (n ϭ 5; p ϭ 0.006, Fig. 3C) that was associated with an increase in its rate of decay (Fig. 3D).
To isolate an effect of CsA on CaN-mediated dephosphorylation of the underlying Cl Ϫ channel as opposed to possible effects on the voltage-dependent calcium channels or Ca 2ϩhomeostatic mechanisms, experiments were performed where I Cl(Ca) was evoked directly by pipette solutions containing 500 nM Ca 2ϩ and 3 mM ATP. With this pipette solution, the rupture of the cell membrane in PA myocytes to achieve whole-cell mode resulted in a large inward current at the holding potential of Ϫ50 mV and the generation of time-dependent outward relaxations following depolarization to ϩ70 mV (Fig. 4). Under the ionic conditions used, these currents represent sustained I Cl(Ca) that have been characterized extensively in vascular smooth muscle cells including PA myocytes (6,(11)(12)(13). In accord with previous observations (12), the amplitude of I Cl(Ca) at Ϫ50 mV declined progressively to a steady-state level over ϳ2 min following rupture that was associated with a reduction in the amplitude of the outward relaxation at ϩ70 mV (Fig. 4, A  and B). Following the initial period of rundown, the amplitude of I Cl(Ca) remained at a constant level ϳ20% of the initial amplitude for the remainder of the experiment. Currents were elicited in cells bathed in normal external solution and alternated with cells incubated in 2 M CsA for 10 min before rupturing the membrane seal to gain whole-cell access. Incubation of cells in CsA significantly attenuated the amplitude of I Cl(Ca) generated by 6 min of cell dialysis with 500 nM Ca 2ϩ in seven cells from three animals (Fig. 4D), and the mean late current at ϩ70 mV was 501 Ϯ 70 and 263 Ϯ 51 pA in the absence and presence of CsA (p ϭ 0.043). In an additional study, CsA had no apparent effect on the amplitude of I Cl(Ca) (Fig. 4E). However, in these cells, the control current was significantly smaller (see Fig. 4F) that was associated with a more prominent rundown of I Cl(Ca) upon membrane rupture. These data show that CsA was able to inhibit sustained I Cl(Ca) , and the effectiveness of this agent was proportional to the amplitude of the control currents. Overall, these experiments show that the suppression of endogenous CaN by CsA diminishes the amplitude of I Cl(Ca) in PA smooth muscle cells but also suggest that CaN is particularly labile in the whole-cell configuration resulting in variable effects of the phosphatase inhibitor.
Effect of Dialysis with Constitutively Active Forms of CaN-To circumvent any variable influence due to endogenous CaN and to shift the cellular status in favor of dephosphorylation, we undertook experiments using pipette solutions enriched with constitutively active recombinant CaN isoforms. The inclusion of CaNA-␣ in a pipette solution containing 500 nM Ca 2ϩ attenuated the initial rundown observed upon rupture of the cell membrane (Fig. 5A), and this was followed by a progressive enhancement of current amplitude over the next 20 min. Consequently, intracellular dialysis with CaNA-␣ augmented considerably the amplitude of I Cl(Ca) (Fig. 5, B and C). After a 6-min recording, the mean current at the end of a step to ϩ90 mV was 17 Ϯ 5 and 47 Ϯ 8 pA pF Ϫ1 in the absence and presence of 500 nM CaNA-␣, respectively (n ϭ 8 and 6). The increase in I Cl(Ca) amplitude was associated with an increase in the rate of current activation at positive potentials and a slowing of current decline at Ϫ80 mV (Fig. 5D). The inclusion of CaNA-␣ that had been boiled for 15 min to reduce enzyme activity failed to increase I Cl(Ca) (mean current at ϩ90 mV after a 2-min recording time was 16 Ϯ 6 pA pF Ϫ1 compared with 7.4 Ϯ 2 pA pF Ϫ1 under control conditions; Fig. 5C).
Similar to previous studies (6,12), currents evoked by 500 nM Ca 2ϩ in the absence and presence of CaNA-␣ reversed (E rev ) close to the theoretical chloride equilibrium potential (ϩ2 mV) at ϩ6 Ϯ 2 and ϩ6 Ϯ 1 mV (n ϭ 4), respectively. Replacement of external NaCl with sodium thiocyanate shifted E rev to Ϫ40 Ϯ 1 and Ϫ42 Ϯ 5 mV, respectively (Fig. 6). These data show that the large current recorded in the presence of CaNA-␣ was not due to the de novo activation of a contaminating current but was due to an enhanced activation of I Cl(Ca) .
The stimulatory effects of CaNA-␣ were abolished by codialysis with CaN-AIP (Fig. 7A), a peptide inhibitor of CaN that mimics the auto-inhibitory domain (6,10). In these experiments, I Cl(Ca) at ϩ90 mV with CaNA-␣ alone was 68 Ϯ 14 pA pF Ϫ1 (n ϭ 6), but in combination with 100 M CaN-AIP, the mean current at ϩ90 mV was 19 Ϯ 4 pA pF Ϫ1 (n ϭ 5). In contrast to the striking data with CaNA-␣, the ␤ isoform failed to enhance I Cl(Ca) . In these experiments, a pipette solution containing 500 nM CaNA-␣ was alternated with one containing 500 nM CaNA-␤ and the mean I Cl(Ca) values at ϩ90 mV were 56 Ϯ 11 and 15 Ϯ 4 pA pF Ϫ1 , respectively (n ϭ 9 both groups, Fig. 7B). These data not only reveal a high degree of isoform specificity but also show that the stimulation produced by CaNA-␣ was not the result of a nonspecific effect due to the cell dialysis of a foreign protein. This point was supported by the observation that CaNA-␣ did not enhance currents evoked by pipette solutions containing 10 mM BAPTA with no added Ca 2ϩ (effectively zero Ca 2ϩ , Fig. 7D). These experiments also reveal that CaN-dependent dephosphorylation "alone" cannot stimulate I Cl(Ca) and establish that CaN is a crucial regulator of Cl Ϫ channel activity but is "not" the impetus for channel activation. DISCUSSION The findings of the present study reveal CaN to be an important regulator of I Cl(Ca) in PA smooth muscle cells and demonstrate that the modulation of the underlying channels is mediated solely by CaNA-␣. Our data show that the inhibition of endogenous CaN with the specific agent, cyclosporin A, reduced the amplitude of I Cl(Ca) evoked either directly by pipette solutions containing 500 nM Ca 2ϩ or as a consequence of Ca 2ϩ influx through voltage-dependent Ca 2ϩ channels. These data are consistent with our previous observation in coronary cells that endogenous CaN is an important regulator of I Cl(Ca) in vascular smooth muscle cells (11). The enrichment of the pipette solution with a constitutively active form of CaNA-␣ augmented markedly the amplitude of I Cl(Ca) allied with a considerable change in the voltage-dependent kinetics. Remarkably, this effect was not shared by the ␤ isoform of CaNA, although Western blot analysis revealed that both ␣ and ␤ isoforms are expressed at significant levels in the PA. Consistent with a specific role of CaNA-␣ for regulating I Cl(Ca) in PA cells was the observation that this isoform translocated toward the plasma membrane following elevation of intracellular Ca 2ϩ levels, a property not shared by CaNA-␤.

Specific Regulation of I Cl(Ca) by CaNA-␣ in Pulmonary Artery
Smooth Muscle Cells-CaN-dependent dephosphorylation influences a number of cellular processes including the regulation of transcription factors, synaptic vesicle recycling, and cardiac muscle hypertrophy (8,9). CaN has also been shown to modulate a number of cation channels including voltage-dependent and ATP-sensitive K ϩ channels in smooth muscle cells (18,19), but data on Cl Ϫ channels are sparse. Recently, we reported that inhibitors of CaN reduced the amplitude of I Cl(Ca) in coronary artery myocytes in a Ca 2ϩ -dependent manner and that these agents decreased the apparent binding affinity of the Cl Ϫ channel for Ca 2ϩ (11). This study shows that the CaN inhibitor CsA reduces the amplitude of I Cl(Ca) in PA cells activated as a consequence of Ca 2ϩ influx through voltage-dependent calcium channels or evoked directly by pipette solutions containing 500 nM free Ca 2ϩ . Moreover, CsA was observed to be relatively ineffective on currents recorded from cells where the control currents were relatively small. These data suggest that the activity of CaN may vary among populations of myocytes but that the level of activity appears to be a crucial determinant of Cl Ϫ current amplitude. Taken together with our previous findings with CaMKII inhibitors (6) in PA cells, these results implicate a contemporaneous modulation of the channel protein or a closely associated regulatory subunit by phosphorylation and dephosphorylation mechanisms. A similar contraparallel regulation also exists in coronary artery smooth muscle cells, but there seems to be differences in Ca 2ϩ sensitivity that probably reflect a tissue-specific pattern of expression of the kinases and phosphatases regulating I Cl(Ca) . Consistent with this idea, CaMKII inhibition stimulates I Cl(Ca) evoked by 500 nM Ca 2ϩ in PA cells but has no effect on this current in coronary artery cells (6), although effects are observed when the channel is stimulated by 1 M free Ca 2ϩ (11). However, peptide and inorganic inhibitors of CaN consistently inhibited I Cl(Ca) elicited by 500 nM Ca 2ϩ in coronary artery myocytes (11). Consequently, CaN-mediated dephosphorylation in PA myocytes may be subjugated by an overwhelming activity of CaMKII (and perhaps other kinases). The rundown of I Cl(Ca) activity seen in PA cells following activation is likely to reflect a change in the kinase/phosphatase balance in the vicinity of the channel. This hypothesis was corroborated by the use of constitutively active CaN. It is worth stressing that in perforated patch experiments CsA consistently suppressed inward tail I Cl(Ca) and altered deactivation kinetics without influencing peak I Ca(L) significantly. These data lend support to the notion that I Cl(Ca) may be physiologically regulated by CaN in conditions minimizing intracellular dialysis and infer that CaN may have a relatively greater impact on I Cl(Ca) regulation when the stimulating rise in [Ca 2ϩ ] i is transient. Under these conditions, the activation of both CaMKII and CaN by Ca 2ϩ /CaM would be less than that observed with a sustained rise in [Ca 2ϩ ] i ; however, as CaN has a greater Ca 2ϩ sensitivity than CaMKII (8,20), the influence of the phosphatase is likely to dominate.
Unique Regulation of I Cl(Ca) by Calcineurin A-␣-One novel finding of our study was the isoform-specific enhancement of I Cl(Ca) produced by the inclusion of a constitutively active form of CaN. Whereas both CaNA-␣ and CaNA-␤ were shown to be expressed in pulmonary arteries, only intracellular dialysis with CaNA-␣ modulated I Cl(Ca) , although the ␣ and ␤ forms of the catalytic A domain are 81% identical at the amino acid level (21) and have a similar Ca 2ϩ dependence. Co-application of CaNA-␣ with a peptide fragment analogous to the auto-inhibitory domain confirmed that the effects of CaNA-␣ were due to a specific phosphatase action. Moreover, immunocytochemical experiments revealed that only endogenous CaNA-␣, but not CaNA-␤, translocated toward the membrane under conditions mimicking our patch clamp experiments with internal Ca 2ϩ clamped at 500 nM. CaN heterodimers containing CaNA-␣ and CaNA-␤ catalytic subunits exhibit a similar Ca 2ϩ dependence, as this is conferred by the B subunit, but display different substrate affinities and catalytic activities in vitro (10). Moreover, trans-genetic approaches have revealed specific functions for the CaNA isoforms. For example, CaNA-␤ is required for T-cell proliferation as well as cardiac hypertrophy (22,23), whereas CaNA-␣(Ϫ/Ϫ) mice display hyperphosphorylated -proteins in the brain and altered post-synaptic de-potentiation in the hippocampus (24,25). This study is the first to show isoform-selective effects on a native ion channel and suggest that the expression of the ␣ isoform of CaNA is the crucial determinant of the phosphorylation status of the Ca 2ϩ -activated Cl Ϫ channel. Consequently, the relative expression of CaNA-␣ or differences in the ␣:␤ ratio of CaN heterodimers between different vascular smooth muscle tissues would be expected to alter the level of I Cl(Ca) regulation by this phosphatase. This exquisite property allows the regulation of I Cl(Ca) by CaN to be fine-tuned by the cell through alterations in the composition of CaN heterodimers. Moreover, the inability of CsA to suppress I Cl(Ca) when the control amplitude was relatively small suggests strongly that the generation of I Cl(Ca) was reliant upon the level of CaN activity. The precise mechanism conferring specificity of CaNA-␣ on I Cl(Ca) cannot be deduced from our data and will require further investigation. Fig. 8A shows the basic features of the structural domains of CaN and the percentage of amino acid sequence identity of the rat brain ␣ and ␤ isoforms (26). The greatest sequence divergence between these two isoforms is observed at the N and C termini (Ͻ30%) and to a lesser extent at the Linker I region (60%). An interesting characteristic of the ␤ isoform is the presence of 10 proline residues at the N terminus, a sequence not shared by the ␣ isoform (Fig. 8B). Such a sequence could make this domain a target of interacting proteins, which could obstruct or limit access of the phosphatase to the target protein. Future studies will be undertaken to compare the effects of various chimeric constructs of the two isoforms on the anion current.
As yet, the molecular nature of the protein underlying I Cl(Ca) remains undefined (27) and therefore precise information as to how dephosphorylation heightens I Cl(Ca) activity can only be speculated. Our data show unequivocally that Ca 2ϩ -dependent dephosphorylation does not gate the opening of the Cl Ϫ channel by a rise in [Ca 2ϩ ] i as CaNA-␣ failed to generate I Cl(Ca) when the internal solution contained 10 mM BAPTA only (i.e. pipette [Ca 2ϩ ] was in the low nanomolar range). The channel properties underlying the kinetics of I Cl(Ca) elicited by the technique used in this study have been characterized in non-muscle cells FIG. 7. Specificity of the effects of CaNA-␣ on I Cl(Ca) . Panel A shows that co-application of a CaN inhibitory peptide (CaN-AIP, E) suppressed the stimulatory effect of CaNA-␣ (q) on I Cl(Ca) . Each point is the mean Ϯ S.E. of 6 -7 cells. B shows the mean amplitude of I Cl(Ca) activated by 500 nM Ca 2ϩ (n ϭ 9) recorded at the end of a 1.5-s test step to different voltages ranging between Ϫ90 and ϩ130 mV in the presence of constitutively active CaNA-␣ (q) or CaNA-␤ (E). Panel C shows the amplitude of I Cl(Ca) activated by 500 nM Ca 2ϩ at the end of a test pulse to ϩ90 mV normalized to cell size under different recording conditions. The control and CaNA-␣ data were pooled from a number of experiments looking at different comparisons. Each bar is the mean Ϯ S.E. of at least four cells. Panel D shows that CaNA-␣ (q) had no effect on currents generated with pipette solutions containing 10 mM BAPTA and no added Ca 2ϩ . For all of the panels, data were consistently obtained after 6 min of cell dialysis. Each point is the mean Ϯ S.E. of five cells. (28,29). The voltage-dependent outward relaxation reflects an increase in open probability due to an increase in the binding affinity for Ca 2ϩ and a slower rate of channel closure, whereas the exponentially declining inward current at negative potentials is a simple approximation of the channel deactivation. As CaNA-␣ increased the rate of activation at positive potentials and slowed that rate of decay at negative potentials, these observations suggest that the removal of phosphate groups either increases the apparent binding affinity or slows the rate of channel closure. These questions will be addressed in future experiments.
This study shows that CaN is a crucial regulator of I Cl(Ca) in PA cells and that this regulation exhibits a high degree of isoform selectivity. However, the effects of CaN rest in a delicate balance with the suppressive effects of CaMKII and probably other kinases and it is the relative contribution of these enzymes that dictates the amplitude of I Cl(Ca) . A corollary to this point is that the relative dominance of CaMKII and CaN probably differs between smooth muscles and will alter with different [Ca 2ϩ ] i . In view of its greater sensitivity to Ca 2ϩ -CaM, CaN would be expected to dominate I Cl(Ca) regulation at lower [Ca 2ϩ ] i , whereas CaMKII will predominate when [Ca 2ϩ ] i is raised (i.e. during agonist stimulation). A necessary caveat to this generalized hypothesis is that other Ca 2ϩ -independent phosphatases that have not been tested in this study may also regulate I Cl(Ca) .