Regulation of Intracellular Cl- Concentration through Volume-regulated ClC-3 Chloride Channels in A10 Vascular Smooth Muscle Cells*

We previously found that antisense oligonucleotide specific to ClC-3 (ClC-3 antisense) prevented rat aortic smooth muscle cell proliferation, which was related to cell volume regulation. In the present study, we further characterized the regulation of intracellular Cl- concentrations ([Cl-]i) via volume-regulated ClC-3 Cl- channels in an embryo rat aortic vascular smooth muscle cell line (A10 cell) and ClC-3 cDNA-transfected A10 cells (ClC-3-A10) using multiple approaches including [Cl-]i measurement, whole cell patch clamp, and application of ClC-3 antisense and intracellular dialysis of an anti-ClC-3 antibody. We found that hypotonic solution decreased [Cl-]i and evoked a native ICl.vol in A10 cells. The responses of [Cl-]i and ICl.vol to hypotonic challenge were enhanced by expression of ClC-3, and inhibited by ClC-3 antisense. The currents in A10 (ICl.vol) and in ClC-3-A10 cells (ICl.ClC-3) were remarkably inhibited by intracellular dialysis of anti-ClC-3 antibody. Reduction in [Cl-]i and activation of ICl.vol and ICl.ClC-3 in A10 and ClC-3-A10 cells, respectively, were significantly inhibited by activation of protein kinase C (PKC) by phorbol-12,13-dibutyrate (PDBu) and inhibition of tyrosine protein kinase by genistein. Sodium orthovanadate (vanadate), a protein-tyrosine phosphatase inhibitor, however, enhanced the cell swelling-induced reduction in [Cl-]i, accompanied by the activation of ICl.vol and ICl.ClC-3 in a voltage-independent manner. Our results suggest that the volume-regulated ClC-3 Cl- channels play important role in the regulation of [Cl-]i and cell proliferation of vascular smooth muscle cells.

of K ϩ , Cl Ϫ , and H 2 O, and returns the cell volume to normal size. It is generally understood that RVD is mainly mediated by Cl Ϫ efflux through a volume-regulated Cl Ϫ channel (VRC). 1 Therefore, VRC may play an essential role in cell proliferation through regulation of cell volume. In our previous study, we determined effects of different kinds of Cl Ϫ channel blockers on endothelin-1-induced proliferation in cultured rat aortic vascular smooth muscle cells. It was found that the aortic vascular smooth muscle cell proliferation was only inhibited by DIDS (2). Furthermore, ClC-3 antisense inhibited the functional expression of ClC-3 and endothelin-1-induced proliferation in cultured rat aortic vascular smooth muscle cells (3). These results provide evidence that ClC-3 may be the gene responsible for I Cl.vol and mediate volume regulation in vascular smooth muscle cells.
Although it has been suggested that the ClC-3 gene may encode VRC and mediate the volume regulation process in guinea pig ventricular myocytes (4), canine pulmonary smooth muscle cells (5), bovine non-pigmented ciliary epithelial cells (6), HeLa cells, and Xenopus laevis oocytes (7), others have presented results against the ClC-3 hypothesis (8, 9, 20 -22). Therefore, molecular identification of VRC has not been determined. In the present study, we further determined the relationship between volume-regulated Cl Ϫ and ClC-3 channels in A10 and ClC-3-A10 vascular smooth muscle cells by continuously monitoring the change in [Cl Ϫ ] i , whole cell patch clamp, ClC-3 antisense, and intracellular dialysis of an anti-ClC-3 antibody techniques. Our results strongly suggest that the ClC-3 channel is responsible for swelling-induced Cl Ϫ current and Cl Ϫ movement and mediates volume regulation in A10 vascular smooth muscle cell.

MATERIALS AND METHODS
Cell Culture-A10 vascular smooth muscle cells from the American Type Culture Collection. A10 cells were grown in Dulbecco's modified Eagle's medium/F12 medium with 10% fetal calf serum, 100 g/ml streptomycin, and 100 units/ml penicillin. Cultures were maintained at 37°C in a humidified incubator in a 95% O 2 plus 5% CO 2 atmosphere. For electrophysiological experiments, the cells were subcultured in coverslips for 1-2 days.
Transfection was performed with Lipofectamine 2000 reagent according to the manufacturer (Invitrogen, Life Technologies, Inc.). Stably transfected clonal cell lines were selected using geneticin (G418) at 400 g/ml for 2 weeks following the transfection. The surviving G418-resistant cells were further plated and passed in the presence of 200 g/ml G418. The expression of ClC-3 protein was detected by Western blot analysis.
Transfections of Antisense, Sense, or Missense Oligonucleotides-The antisense and sense oligonucleotides corresponding specifically to the initiation codon region of the human ClC-3 mRNA were synthesized (Sangon, Shanghai, China) as reported previously (6). The antisense sequence was 5Ј-TCC ATT TGT CAT TGT-3Ј. The sense oligonucleotide had the sequence 5Ј-ACA ATG ACA AAT GGA-3Ј. We designed missense oligonucleotide 5Ј-TCT ATT CCT GTA TTG-3Ј, which consisted of the same bases employed in the antisense probe, but in a "random" order, and did not recognize any known sequence available in Gen-Bank TM . The first three bases at either end in all oligonucleotides were phosphorothioated. To examine the uptake of oligonucleotide by A10 cells, the oligonucleotides were labeled with fluorescence. For transient transfection, the cells in the quiescent state were transfected with oligonucleotides by incubation for 48 h with Lipofectamine 2000 (5 l/ml).
Measurement of [Cl Ϫ ] i -6-Methoxy-N-ethylquinolinium iodide (MEQ) was reduced to its cell-permeable derivative 6-methoxy-N-ethyl-1,2-dihydroquinoline (dihydro-MEQ) as described previously (10 Electrophysiological Experiments-Cells in a chamber of 500 l in volume were continuously superfused at the rate of 2 ml per min. The Cl Ϫ currents were recorded with an Axopatch 200B Amplifier (Axon Instrument, Foster City, CA) using conventional whole cell recording technique. Patch pipettes were made from borosilicate glass using a two-stage puller (pp-83, Narishige, Tokyo, Japan) and had the resistances of 3-5 ⍀ when the pipettes were filled with the pipette solution. A 3 mM KCl-agar salt bridge between the bath and the Ag-AgCl reference electrode was used to minimize the changes of liquid junction potentials. To determine the whole cell current-voltage curve, the cell was held at Ϫ40 mV, and test potentials were applied from Ϫ100 mV to ϩ120 mV for 400 ms in ϩ20 mV increments at an interval of 5 s. To obtain time-dependent changes in current amplitude, cells were clamped from a holding potential of Ϫ40 mV to a hyperpolarizing potential of Ϫ80 mV for 250 ms, then back to Ϫ40 mV for 25 ms, and then to a depolarizing potential of ϩ100 mV for 250 ms. This protocol was repeated every 30 s. Currents were filtered at a frequency of 2 kHz and digitized at 5 kHz using pCLAMP8.0 software (Axon Instruments). The data were directly entered into the hard drive of a PC-compatible computer. All experiments were performed at room temperature (25°C).
The hypotonic bath solution contained (mM): 107 N-methyl-D-glucamine chloride (NMDG-Cl), 1.5 MgCl 2 , 2.5 MnCl 2 , 0.5 CdCl 2 , 0.05 GdCl 3 , 10 glucose, and 10 Hepes, pH 7.4 adjusted with NMDG. This solution osmolarity measured by a freezing point depression osmometer (OS-MOMAT030, Germany) was 230 mosmol/kg⅐H 2 O. A 300 mosmol/kg⅐H 2 O isotonic bath solution was made by adding 70 mM D-mannitol to the hypotonic solution. A 370 mosmol/kg⅐H 2 O hypertonic bath solution was prepared by adding 140 mM D-mannitol to the hypotonic solution. In the experiment for Cl Ϫ dependence, Cl Ϫ in the hypotonic medium was replaced by equimolar aspartate Ϫ to obtain the hypotonic solution containing 39 mM [Cl Ϫ ] o . In the experiment for anion selectivity, the hypotonic bath solution contained (mM): 115 NaX, 10 glucose, and 10 Hepes, pH 7.4 adjusted by NaOH. X Ϫ denotes I Ϫ , Cl Ϫ or aspartate Ϫ . The pipette solution contained (mM): 95 CsCl, 20 TEACl, 5 ATP-Mg, 5 EGTA, 5 Hepes, and 80 D-mannitol, pH 7.2 adjusted by CsOH. The osmolarity of this solution was 300 mosmol/kg⅐H 2 O. In the intracellular dialysis experiments, anti-ClC-3 antibody was diluted to 300 g/ml by distilled water, and added to the pipette solution (final concentration was 5 g/ml.). For preabsorbed anti-ClC-3 antibody, antibody and antigen were dissolved by distilled water separately, mixed in a ratio of 1:10, then stored at 4°C overnight. The mixed solution was added to the pipette solution. Finally, the pipette solution contained 5 g/ml antibody and 50 g/ml antigen. The osmolarity of the pipette solutions was not significantly altered by adding antibody alone or antigen-presorbed antibody.
Based on the shift of the reversal potential (⌬E rev ), the permeability ratios (P X /P Cl ) were calculated by the modified Goldman-Hodgkin-Katz equation shown in Equation 1, use of a polyclonal antibody directed against ClC-3. The anti-ClC-3 antibody recognized a major band at 80 -90 kDa. The ClC-3 protein expression was significantly increased in ClC-3-A10 cells (Fig. 1).
Effect of ClC-3 Antisense on Expression of ClC-3 Protein-To determine the uptake of oligonucleotide by A10 cell, the oligonucleotides were labeled with fluorescence. As shown in Fig.  2A, under resting conditions the fluorescence in the cells was negligible, but the fluorescence in cells treated with antisense, sense, or missense were greatly increased, which confirmed the uptake of oligonucleotides by these cells. Fig. 2B shows that 100 g/ml ClC-3 antisense oligonucleotide decreased ClC-3 protein expression in a time-dependent manner. After 48 h of ClC-3 antisense treatment, the decrease in ClC-3 protein expression reached a maximum. Fig. 2C shows that the cells in the quiescent state were transfected with oligonucleotides by incubation for 48 h with Lipofectamine 2000 , the ClC-3 protein was decreased by 12.7 Ϯ 1.6%, 26.8 Ϯ 2.8%, 57.8 Ϯ 2.4%, and 58.2 Ϯ 2.3% (data from six different experiments; p Ͻ 0.01) in 5 l/ml Lipofectamine 2000 plus 25, 50, 100, and 200 g/ml ClC-3 antisense, respectively. 100 g/ml ClC-3 antisense reached a maximal effect. Sense, missense oligonucleotides, and Lipofectamine 2000 did not alter ClC-3 protein expression (n ϭ 6 for each group; p Ͼ 0.05 versus control).
The Characteristics of Cl Ϫ Currents-In A10 cells, the patch clamp whole cell currents were very small when the cells were exposed in the isotonic solution. If the bath solution was changed from isotonic solution to hypotonic solution, it evoked the large outward rectifying currents (I Cl.vol ) with a reverse potential of Ϫ2.0 Ϯ 1.5 mV. In ClC-3-A10 cells, hypotonic solution also evoked an outward rectifying current (I Cl,ClC-3 ) with a reverse potential of Ϫ2.6 Ϯ 1.2 mV. Compared with I Cl.vol , I Cl.ClC-3 was larger. The current densities of I Cl.vol (n ϭ 12) and I Cl.ClC-3 (n ϭ 19) were Ϫ23.5 Ϯ 1.5 and Ϫ55.9 Ϯ 2.7 pA/pF (p Ͻ 0.05 versus I Cl.vol ) at Ϫ80 mV, respectively, and 41.8 Ϯ 2.8 and 100.6 Ϯ 5.6 pA/pF (p Ͻ 0.01 versus I Cl.vol ) at ϩ80 mV, respectively (Fig. 3, A and B). When the [Cl Ϫ ] o in the bath solution was changed from 116 to 39 mM, the current was significantly decreased, with a change in the reverse potential from Ϫ1.6 Ϯ 2.0 to 27.9 Ϯ 2.6 mV in A10 cells (n ϭ 6) and from Ϫ2.3 Ϯ 1.7 to 26.6 Ϯ 1.9 mV in ClC-3-A10 cells (Fig. 3, C and  D). Both I Cl.vol and I Cl.ClC-3 were inhibited by DIDS. After exposure to DIDS, the current densities of I Cl.vol (n ϭ 8) and I Cl.ClC-3 (n ϭ 7) were reduced from Ϫ21.8 Ϯ 2.4 to Ϫ14.3 Ϯ 1.6 pA/pF and Ϫ56.0 Ϯ 4.1 to Ϫ34.1 Ϯ 2.5 pA/pF at Ϫ80 mV, respectively, and reduced from 36.4 Ϯ 3.0 to 16.4 Ϯ 1.6 pA/pF and 101.1 Ϯ 9.1 to 41.1 Ϯ 4.9 pA/pF at ϩ80 mV, respectively. DIDS had the same inhibition in I Cl.vol and I Cl.ClC-3 . However, DIDS had more inhibition on currents at ϩ80 mV than that at Ϫ80 mV (Fig. 3, E and F). This indicates that the effects of DIDS on I Cl.vol and I Cl.ClC-3 were voltage-dependent.
100 g/ml ClC-3 antisense remarkably decreased I Cl.vol by 54.3 Ϯ 4.6% at ϩ80 mV and 56.0 Ϯ 5.8% at Ϫ80 mV, which were similar to the decreased magnitude of ClC-3 protein expression induced by 100 g/ml ClC-3 antisense, whereas sense, missense, and Lipofectamine 2000 failed to inhibit this current. Based on results from the ClC-3 antisense experiments, we further examined the inhibitory effects of intracellular dialysis of an anti-ClC-3 antibody on I Cl,vol and I Cl,ClC-3 . Fig. 4 shows that intracellular dialysis of an anti-ClC-3 antibody could completely block the I Cl,vol (Fig. 4, A and B) and I Cl,ClC-3 (Fig. 4, C  and D). To exclude nonspecificity of the anti-ClC-3 antibody, FIG. 4. Effects of anti-ClC-3 antibody (Ab) and preabsorbed anti-ClC-3-antibody dialysis on Cl ؊ currents at ؎80 mV membrane potential. A, intracellular dialysis an anti-ClC-3 antibody completely blocked I Cl.vol , whereas intracellular dialysis preabsorbed anti-ClC-3 antibody had no significant effect on I Cl.vol. This indicates that anti-ClC-3 antibody blocked I Cl.vol through interference with ClC-3 rather than the protein of anti-ClC-3 antibody. B, mean current densities measured at Ϫ80 mV (downward bars) and ϩ80 mV (upward bars) in A10 cells (n ϭ 6). C, intracellular dialysis an anti-ClC-3 antibody completely blocked I Cl.ClC-3 , whereas preabsorbed anti-ClC-3 antibody had no significant effect on I Cl.ClC-3 . D, mean current densities measured at Ϫ80 mV (downward bars) and ϩ80 mV (upward bars) in ClC-3-A10 cells (n ϭ 6). *, p Ͻ 0.05; **, p Ͻ 0.01 versus preabsorbed anti-ClC-3-antibody.
the effect of preabsorbed anti-ClC-3 antibody on the current was determined. It was found that intracellular dialysis of preabsorbed anti-ClC-3 antibody did not inhibit I Cl,vol and I Cl,ClC-3 (Fig. 4, A and B). Table I illustrates the anion selectivity of both the I Cl,vol channel in A10 cell and I Cl,ClC-3 channel in ClC-3-A10 cell. The reverse potential induced by I Ϫ was more minus than that by Cl Ϫ , whereas aspartate Ϫ produced a positive reverse potential. The anion selectivity order of these channels was as follows: I Ϫ Ͼ Cl Ϫ Ͼ aspartate Ϫ . There was no significant difference in reverse potentials induced by anions and the anion permeability between I Cl,vol and I Cl,ClC-3 .  (Fig. 6D). Fig. 6B shows that 30 M genistein completely inhibited I Cl,vol (top trace) and I Cl,ClC-3 (bottom trace). The levels of [Cl Ϫ ] i induced by hypotonic solution in A10 and ClC-3-A10 cells were almost elevated to the normal level by genistein (Fig. 6E). Whereas, the protein-tyrosine phosphatase inhibitor, sodium orthovanadate, potentiated I Cl.vol (Fig. 6C, top trace) and I Cl,ClC-3 (Fig. 6C, bottom trace) in a voltage-independent manner. The densities of currents for I Cl,vol before and after exposure to 500 M vanadate were Ϫ24.1 Ϯ 2.3 and Ϫ34.2 Ϯ 3.1 pA/pF at Ϫ80 mV (n ϭ 8), respectively, and 40.2 Ϯ 2.9 and 55.6 Ϯ 2.7 pA/pF at ϩ80 mV (n ϭ 8), respectively. For I Cl,ClC-3, the densities of currents were Ϫ63.5 Ϯ 5.  (Fig. 6F). DISCUSSION In the present study, we first demonstrated that hypotonic solution significantly reduced [Cl Ϫ ] i and simultaneously activated an outward rectifying I Cl,vol in A10 cells. Both ClC-3 antisense and intracellular dialysis of anti-ClC-3 antibody significantly inhibited the hypotonicity-induced decrease in [Cl Ϫ ] i and activation of I Cl,vol . We then found that overexpression of ClC-3 in A10 cells (ClC-3-A10 cells) yielded a larger I Cl,vol with the same reverse potential, sensitivity to voltage-dependent inhibition by DIDS, and anion selectivity (I Ϫ Ͼ Cl Ϫ Ͼ aspartate Ϫ ) as that of endogenous I Cl.vol. in A10 cells under hypotonic conditions. Furthermore, we showed that the hypotonic cell swelling induced more decrease in [Cl Ϫ ] i in ClC-3-A10 cells, and activation of I Cl.ClC-3 also was remarkably inhibited by intracellular dialysis of anti-ClC-3 antibody. Finally, we found that hypotonic cell swelling-induced changes in [Cl Ϫ ] i are not only intimately linked to the action of I Cl.vol and I Cl.ClC-3 but  also, as having been reported for I Cl.vol and I Cl.ClC-3 in other cells (16 -18), well controlled by endogenous PKC and proteintyrosine phosphorylation. These data provide compelling evidence that ClC-3 channels are responsible for the Cl Ϫ trans-portation and I Cl.vol during hypotonic perturbations in A10 cells.

The Change in [Cl Ϫ ] i Induced by Hypotonic Solution-In
The molecular identification of the protein responsible for I Cl.vol has been particularly difficult to resolve. The main rea- son for this difficulty is that endogenous I Cl.vol is expressed practically in all types of cells, which can be always superimposed with membrane currents because of transgenic expression of candidate genes. Several molecular candidates responsible for I Cl.vol were previously proposed: 1) the multidrug transporter, P-glycoprotein (P-gp), a member of the ATP-binding cassette (ABC) family of transporters (11), 2) pICln, which encodes a small 235 amino acid protein with little homology to any known anion channel structure (12), 3) ClC-2, a member of the large ClC superfamily of voltage-dependent Cl Ϫ channels (13), and 4) ClC-3, also a member of the ClC superfamily (4). Although both P-gp and pICln expression appeared to yield chloride currents with many of the properties of native I Cl.vol , it now seems clear that the currents observed in these studies were not due to expression of either P-gp or pICln but likely due to contamination by endogenous I Cl.vol (14). Although ClC-2 represents a bona fide anion channel regulated by cell volume, expressed ClC-2 currents differ significantly in voltage sensitivity, anion selectivity, and pharmacology from conventional, outwardly rectifying I Cl.vol found in most cells (15).
The ClC-3 hypothesis was based on the fact that stable or transient transfection of a full-length ClC-3 cDNA cloned from guinea pig ventricle (gpClC-3) into NIH/3T3 cells yielded a basally active chloride conductance that was strongly modulated by cell volume (4). Many properties of the expressed I gpClC-3 resemble those reported for native I Cl.vol in heart and other tissues, including an outwardly rectifying unitary slope conductance of 40 pS, an anion selectivity of I Ϫ Ͼ Cl Ϫ Ͼ Asp Ϫ , inactivation at positive potentials, increase by extracellular hypotonicity, and inhibition by hypertonicity, by extracellular nucleotides, by phorbol esters, by stilbene derivatives, and by tamoxifen. Furthermore, site-directed mutagenesis of an asparagine near the end of the transmembrane-spanning domains (N579K) altered rectification and anion selectivity of the expressed I gpClC-3 . These observations were initially confirmed by other independent studies of several groups (16 -18). Thus, ClC-3 represented a viable molecular candidate responsible for native I Cl.vol in heart or any other mammalian cell type (19).
Several reports (20 -22), however, failed to support a role for exogenously expressed ClC-3 as a viable candidate for native I Cl.vol . The magnitude of swelling-activated Cl Ϫ current is not significantly different between nontransfected and human ClC-3 cDNA-transfected HEK293 cells (22). Overexpression of ClC-3 in HEK293 cells (22) and X. oocytes (23,24) did not produce outward rectifying I Cl.vol . In ClC-3 cDNA-transfected CHO-K1 cells, the ClC-3 channel current was not identified in the endogenous swelling-activated channel current, and not activated by cell swelling (20). Some of these inconsistencies might be attributed to failure to successfully express functional ClC-3 protein and/or difficulties in effectively separating transgenic currents from endogenously expressed Cl Ϫ currents present in expression cell systems such as X. oocytes and some mammalian cells (25).
Furthermore, the controversy surrounding the actual physiological role of ClC-3 Cl Ϫ channels was additionally fueled by the report that disruption of the Clcn3 gene did not change the volume-regulated chloride current in hepatocytes, pancreatic acinar cells (26), and salivary acinar cells (27) from Clcn3 Ϫ/Ϫ mice. In these and another studies (28), ClC-3 was primarily localized to intracellular membranes where it was proposed to function primarily in vesicular acidification, although other studies have clearly demonstrated plasma membrane localization of heterologously expressed ClC-3 (22, 29 -31) and endogenous ClC-3 (32, 33) in various cell types.
Recently, it has been shown that volume-sensitive osmolyte and anion channels (VSOACs) currents activated by hypotonic medium in atrial myocytes and pulmonary arterial smooth muscle cells from Clcn3 Ϫ/Ϫ and Clcn3 ϩ/ϩ mice were remarkably similar in activation kinetics, steady-state current densities, slight outward rectification, and anion selectivity. However, there also are significant differences in sensitivity to PKC regulation, inhibition by intracellular dialysis with a new anti-ClC-3 antibody, [ATP] i depletion, and high free [Mg 2ϩ ] i between them (34). These authors have suggested that in response to Clcn3 gene deletion, there may be compensatory changes in expression of other proteins that alter VSOAC channel subunit composition or associated regulatory subunits that give rise to VSOACs with different properties. It appears that differential sensitivity of native VSOACs to anti-ClC3 antibody and phorbol esters in different cell types in normal animals is consistent with the possible expression of distinct VSOACs subtypes. In fact, the responses of volume-regulated chloride current to PKC regulation in vascular smooth muscle cells from different blood vessels are not the same. Activation of PKC by PDBu increased the amplitude of swelling-activated chloride current in rabbit portal vein (35), whereas PKC activation inhibited I Cl.vol in mouse pulmonary arterial smooth muscle cells (34). As shown in this study, PDBu significantly inhibited I Cl.vol and increased [Cl Ϫ ] i under hypotonic condition. Our results provide evidence to support that ClC-3 channels are responsible for volume-regulated Cl Ϫ transportation and Cl Ϫ currents and may play an important role in volume regulation in A10 vascular smooth muscle cells.
Alternatively, it is possible that contradictory data about ClC-3 in mediating cell volume regulation are caused by differences in [Cl Ϫ ] i and ClC-3 expression levels in different types of cells. In vascular smooth muscle cells, [Cl Ϫ ] i is generally higher than other tissues, and the Cl Ϫ equilibrium potential (E Cl ) is approximately Ϫ20 mV, which is more positive than the resting membrane potential. Therefore, activation of the Cl Ϫ channels will lead to a net Cl Ϫ efflux followed by the membrane depolarization (36,37). It is currently not known why [Cl Ϫ ] i is higher in vascular smooth muscle. In sharp contrast to vascular smooth muscle cells, some neurons have passively distributed intracellular Cl Ϫ , and it sets the resting membrane potential closer to E Cl (38). It is interesting that the ClC-3 expression level in vascular smooth muscle cells, including aorta smooth muscle cells is very high (39), whereas only a small fraction of ClC-3 is expressed on the surface membrane in hepatocytes CHO-K1 cells and human hepatoma cell line Huh-7 (28), Our results may, therefore, shed a new light on understanding the regulation of intracellular Cl Ϫ concentrations in different cell types.