HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 275, Issue 35, 26986-26993, September 1, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/35/26986    most recent M004301200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Tulk, B. M. Articles by Edwards, J. C. Search for Related Content PubMed PubMed Citation Articles by Tulk, B. M. Articles by Edwards, J. C. Social Bookmarking                      What's this?

CLIC-1 Functions as a Chloride Channel When Expressed and Purified from Bacteria^*

Barry M. Tulk, Paul H. Schlesinger^§, Shefalee A. Kapadia, and John C. Edwards^¶

From the  Department of Internal Medicine, St. Louis University, the ^§ Department of Cell Biology and Physiology, Washington University School of Medicine, and ^¶ St. Louis Veterans Affairs Medical Center, St. Louis, Missouri 63106

Received for publication, May 18, 2000, and in revised form, June 27, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CLIC-1 is a member of a family of proteins related to the bovine intracellular chloride channel p64 which has been proposed to function as a chloride channel. We expressed CLIC-1 as a glutathione S-transferase fusion protein in bacteria. The fusion protein was purified by glutathione affinity, and CLIC-1 was released from its fusion partner by digestion with thrombin. After further purification, CLIC-1 was reconstituted into phospholipid vesicles by detergent dialysis. Chloride permeability of reconstituted vesicles was assessed using a valinomycin dependent chloride efflux assay, demonstrating increased vesicular chloride permeability with CLIC-1 compared with control. CLIC-1-dependent chloride permeability was inhibited by indanyloxyacetic acid-94 with an apparent IC₅₀ of 8.6 µM. The single channel properties of CLIC-1 were determined using the planar lipid bilayer technique. We found that CLIC-1 forms a voltage-dependent, Cl-selective channel with a rectifying current-voltage relationship and single channel conductances of 161 ± 7.9 and 67.5 ± 6.9 picosiemens in symmetric 300 and 150 mM KCl, respectively. The anion selectivity of this activity is Br  Cl > I. The open probability of CLIC-1 channels in planar bilayers was decreased by indanyloxyacetic acid-94 with an apparent IC₅₀ of 86 µM at 50 mV. These data convincingly demonstrate that CLIC-1 is capable of forming a novel, chloride-selective channel in the absence of other subunits or proteins.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

A variety of distinct chloride channel activities have been described that carry out a host of essential roles in cellular physiology. Chloride channels in plasma membranes play important roles in defense of cell volume, transepithelial transport, setting the membrane potential, bone resorption, and the response to certain neurotransmitters (1, 2). Chloride channels are also present in intracellular membranes where they are known to play important roles in acidification of intracellular compartments and in exocytosis (3, 4).

Over the past several years significant progress has been made toward identifying and understanding the molecular basis for chloride channels. To date, four distinct, structurally unrelated types of chloride channels have been identified: the ClC family (5), the ligand-gated family (e.g. -aminobutyric acid and glycine receptors) (6), the CLIC family (7), and the cystic fibrosis transmembrane conductance regulator (CFTR)¹, a member of the ATP binding cassette family of proteins (8).

The CLIC family is a closely related group of proteins homologous to the renal chloride channel p64 (7). The role of a CLIC family member in chloride permeability is most firmly established for p64 itself. P64 co-purifies with channel activity solubilized from bovine kidney microsomes (9). Immunodepletion of p64 from solubilized microsomes removes reconstitutable chloride channel activity (10), and expression of exogenous p64 in HeLa cells results in the appearance of a novel, outwardly rectifying, dinitrostilbene disulfonic acid-inhibitable, anion-selective channel (11).

CLIC-1, or NCC27, is another member of the CLIC family. This protein resides in an intracellular vesicular compartment in cultured epithelial cells and in the apical domain of renal proximal tubule cells (12). Expression of CLIC-1 in Chinese hamster ovary cells has been reported to lead to increased chloride permeability of plasma and nuclear membranes (13). Expression of a third family member, CLIC-3, leads to increased whole-cell chloride conductance (14).

Thus, expression of three CLIC family members in mammalian cells has been associated with the appearance of novel chloride channel activities. However, until now direct evidence for the formation of active channels by CLIC proteins alone has not been reported. Based on the current evidence, it is possible that CLIC proteins, rather than acting as ion channels themselves, may complex with or activate endogenous proteins to generate chloride channels.

In this paper, we present compelling evidence that CLIC-1 alone is capable of forming an active chloride channel. CLIC-1 was expressed as a GST fusion protein in bacteria, purified to homogeneity, and reconstituted into lipid vesicles. The chloride permeability of reconstituted vesicles was detected using a potential-driven chloride efflux assay, and the single channel properties of the CLIC-1 channel were determined using the planar lipid bilayer technique. We find that purified CLIC-1 functions as a rectifying, chloride-selective channel that it is inhibited by IAA-94.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials and General Methods-- Unless otherwise noted, all materials were obtained from Sigma. Protein quantification was performed using the Bicinchoninic Acid (BCA) protein assay (Pierce). SDS-PAGE and molecular cloning procedures were carried out by standard methods (15, 16).

Cloning and Expression of GST-CLIC-1-- cDNA clone PG11 (12) containing the entire coding region of CLIC-1 was subcloned into the multiple cloning site of the GST fusion protein vector pGEX-KG (17), yielding a construct encoding a fusion protein consisting of GST and CLIC-1 separated by a thrombin cleavage site. Thrombin digestion of the GST-CLIC-1 fusion protein is predicted to yield the complete CLIC-1 protein with the sequence N terminus-GSACELGTPSNSARAT (encoded by linker sequences) appended to the initial methionine of the protein.

Escherichia coli XL1-blue (Stratagene, La Jolla, CA) harboring either pGEX-KG or pGEX-PG11 were grown to late log phase. Isopropylthiogalactoside was added to 1 mM, and growth was continued for 5 h. Cells were harvested by centrifugation and stored at 20 C until needed.

Purification of GST-CLIC-1-- Unless otherwise stated, all procedures were performed at 4 °C. Frozen cell pellets from 500-liter cultures were thawed on ice in 10 ml of 10 mM Tris (pH 8.0), 1 mM EDTA, 1 mM DTT, and 0.1 volume of Sigma bacterial protease inhibitor mixture. Lysozyme was added to 1 mg/ml, and cells were incubated on ice for 30 min. After the addition of MgCl₂ to 10 mM and DNase and RNase to 10 µg/ml, the suspension was incubated with mixing for 1 h. Unbroken cells and debris were removed by centrifugation at 10,000 rpm in a Beckman JA-20 rotor, and the resulting supernatant was further clarified by centrifugation at 40,000 rpm for 1 h in a Beckman 70.1 Ti rotor.

Supernatants were applied directly to a 5-ml column of glutathione-agarose that had been pre-equilibrated in 10 mM Tris (pH 8.0), 150 mM NaCl, 1 mM DTT (equilibration buffer). Columns were washed with equilibration buffer followed by the same buffer containing 1 M NaCl and, finally, were equilibrated in thrombin cleavage buffer (20 mM Tris (pH 8.0), 150 mM NaCl, 1.4% N-octylglucopyranoside, 1 mM DTT, 2.5 mM CaCl₂). The resin was then suspended in an equal volume of cleavage buffer, recombinant human thrombin (Novagen, Madison, WI) was added to a final concentration of 2 units/ml of resin, and suspensions were incubated overnight at room temperature with gentle agitation.

Thrombin-treated glutathione-agarose suspensions were loaded into columns and washed with cleavage buffer. Fractions recovered from the pGEX-CLIC-1 column with significant A₂₈₀ were pooled. Equivalent fractions were pooled from thrombin-treated, GST-bound columns for use as control.

CLIC-1 and control fractions were diluted 3-fold with 1.4% N-octylglucopyranoside, 1 mM DTT and applied to 1-ml High-Q anion exchange cartridges (Bio-Rad) equilibrated in 10 mM Tris (pH 8.0), 1.4% N-octylglucopyranoside, 1 mM DTT. Columns were washed with equilibration buffer, then eluted with buffer containing 0.5 M NaCl. Peak fractions were pooled, as were corresponding control fractions, and 1 ml of each pool was applied independently to a 1.5 × 50-cm column of Sephacryl S-100 equilibrated in 150 mM NaCl, 20 mM Tris (pH 8.0), 1.4% N-octylglucopyranoside, and 1 mM DTT. Fractions containing CLIC-1 were identified by SDS-PAGE analysis and pooled. Corresponding fractions from the identically handled control samples were also pooled.

Molecular Weight Determinations-- Size exclusion chromatography and quasi-elastic light scattering were used to determine the apparent molecular weight (M_r) of CLIC-1. A 1.5 × 50-cm Sephacryl S-100 column equilibrated as described previously was calibrated using M_r standards (Sigma MW-GF-70 kit). The manufacturer's recommended protocol was then used to calculate the M_r of CLIC-1.

Quasi-elastic light scattering measurements were made on CLIC-1 purified through High-Q anion exchange chromatography using a Dyna Pro-801TC Dynamic Light Scattering instrument (Protein Solutions, Inc., Charlottesville, VA). Measurements were made according to the manufacturer's protocol.

Reconstitution of CLIC-1 Chloride Channel Activity-- Purified, recombinant CLIC-1 protein was reconstituted into phospholipid vesicles essentially as described previously (10, 11). Briefly, 9 volumes of purified CLIC-1 or identically prepared control fractions were mixed with 1 volume of 100 mg/ml asolectin (Sigma P5638), 90 mg/ml N-octylglucopyranoside in water. Mixtures were dialyzed for 2 days at 4 °C against 3 changes of 200 mM KCl, 10 mM HEPES (pH 7.0), 1 mM DTT. The resultant vesicles were used within 3 days in either chloride efflux assays or for fusion with planar lipid bilayers. The vesicle suspension was passed through a mini-extruder (Avanti Polar Lipids, Birmingham, AL) fitted with a 0.2-µm pore filter to obtain a homogeneous, unilamellar vesicle population. For planar lipid bilayer experiments, vesicles were diluted 1:1 with a solution of 400 mM KCl, 600 mM glycerol, 10 mM HEPES (pH 7.0) before extrusion.

Chloride Efflux Assay-- The chloride permeability of reconstituted vesicles was assayed essentially as described previously (11, 18). The significance of differences in mean efflux rates was determined using analysis of variance (19).

Since the initial intravesicular [Cl] was 200 mM, the final [Cl] after addition of detergent could be used to determine the initial intravesicular volume. We found that the intravesicular volume composes 1 to 2% of the total volume of the undiluted reconstituted vesicles prepared by our methods.

The concentration dependence of IAA-94 inhibition was determined from efflux assays conducted over a range of inhibitor concentrations. IAA-94 was dissolved in 20 mM sodium bicarbonate, pH 9.5. The vesicle suspension was brought to the desired concentration of IAA-94 immediately before passage through the mini-extruder, ensuring equilibration between intra- and extravesicular solutions. The IC₅₀ for IAA-94 was determined using the formula r = IC₅₀ × R_o/(IC₅₀ + [I]), where R is the efflux rate, R_o is the efflux rate in the absence of inhibitor, and [I] is the inhibitor concentration (20).

Single Channel Characterization of CLIC-1 in Planar Lipid Bilayers-- Delrin cuvettes with 250-µm diameter apertures and bilayer chambers were obtained from Warner Instrument Co. (Hamden, CT). Bilayers were formed from 30 mg/ml asolectin in decane. Membranes were suspended between 50 mM KCl, 1 mM HEPES (pH 7.0) and allowed to thin. Cis and trans chambers were connected via 3 M KCl salt bridges to AgCl electrodes, which in turn were connected to a CV 201A-capacitance feedback headstage and an Axopatch 200A integrating voltage clamp amplifier (Axon Instruments, Inc.). Data were collected using Clampex 7 software (Axon Instruments, Inc.). The membrane capacitance subroutine in Clampex 7 was used to monitor thinning; all membranes used in this study thinned to at least 200-picofarad capacitances.

Analog data filtered at 1 KHz were collected and stored on videotape using a VHS videorecorder interfaced with a Neurodata Neurocorder. Data were digitized upon playback through the Neurocorder using Clampex 7 software and a Digidata 1200 A/D converter (Axon Instruments, Inc.) interfaced with a personal computer. Data were collected at 5000 Hz, filtered upon playback at 500 Hz, and analyzed using the pClamp 6 suite of programs (Axon Instruments, Inc). Current measurements were determined at each holding potential from amplitude histograms, which were generated from single channel traces. Reversal potentials were determined from current-voltage plots using a least squares-fitting protocol. To generate time constants from voltage-dependent channel closure, data from averaged, normalized, current traces were plotted versus time. Using Microcal Origin software, single exponential fits were generated from the equation y = y_o + A1 × e^x/ by a ² minimization protocol.

Ion selectivity ratios were determined using the Goldman Hodgkin Katz equation (21). An IAA-94 dose-response curve was determined from 5-min records of single channel data obtained at +50 mV. P_o was determined from each data set according to the equation P_o = (t_o/ t_i)/N, where t_o = open time, t_i = time interval, and N = number of channels. P_o was plotted versus log[IAA-94] and fit to the equation y = [(A1-A2)/(1 + (x/x_o)P] + A2, where A1 = P_o in the absence of IAA-94, A2 = P_o in the presence of saturating IAA-94, x_o = IC₅₀, and p = Hill coefficient.

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Expression and Purification of CLIC-1High level expression of both GST and GST-CLIC-1 was readily obtained. Typical protein profiles of bacterial lysates prepared from cells expressing GST or GST-CLIC-1 are depicted in Fig. 1A, lanes a and b. Control and CLIC-1 samples were prepared by thrombin cleavage of GST and GST-CLIC-1, respectively, while each was bound to glutathione-agarose. Thrombin cleavage liberated intact CLIC-1 from GST, which eluted from glutathione-agarose in the wash fractions (Fig. 1B, lane e). No bands of >14 kDa were detected in the wash fractions of thrombin treated GST (Fig. 1B, lane a). GST and uncleaved GST-CLIC-1 were subsequently eluted from glutathione-agarose with 5 mM reduced glutathione (Fig. 1B, panels b and f). Cleavage efficiency was approximately 50%; a typical cleavage liberated 13 mg of CLIC-1/liter of culture.

View larger version (18K): [in this window] [in a new window]   Fig. 1.   Expression and purification of CLIC-1. A, 20 µg of crude bacterial lysates prepared from cells overexpressing the parent GST construct (a) or GST-CLIC-1 (b) were separated by SDS-PAGE and stained with Coomassie Blue. B, material isolated from GST control and GST-CLIC-1-expressing cells was purified. Equivalent volumes of material from the control and CLIC-1 samples at various stages of purification were separated by SDS-PAGE and stained with Coomassie Blue. Lanes a-d and e-h represent material recovered from GST and GST-CLIC-1, respectively. Lanes a and e, material released from glutathione-agarose-bound samples with thrombin. Lanes b and f, material eluted by reduced glutathione after thrombin treatment. Lanes c and g, material recovered following anion exchange chromatography. Lanes d and h, material recovered following Sephacryl S-100 chromatography. Lanes b, e, f, g, and h contain 21, 3, 10, 54, and 3 µg of protein, respectively.

CLIC-1 and control fractions were further purified using anion exchange and size exclusion chromatography. The final products are shown in Fig. 1B, lanes d and h. Anion exchange chromatography typically yielded between 1 and 5 mg of purified CLIC-1/liter of culture. 50% of the protein applied to Sephacryl S-100 was routinely recovered from this column.

Molecular Weight Determinations for Purified CLIC-1-- Both size exclusion chromatography and quasi-elastic light scattering analysis were utilized to determine the apparent M_r of CLIC-1. The apparent M_r determined from chromatography on Sephacryl S-100 is 34,400 ± 5500 (n = 4). Quasi-elastic light-scattering measurements were used to determine the solution molecular weight of CLIC as described previously (22) using CLIC-1 purified through High-Q anion exchange chromatography. The apparent molecular weight obtained by this technique was 3200 ± 4000. This value is in excellent agreement both with the apparent molecular weight obtained from the sizing column and from SDS-PAGE. However, these values are somewhat higher than the molecular weight predicted from our CLIC-1 fusion protein, which is 28.4 (13). The reason for the discrepancy in the predicted versus apparent M_r values for CLIC-1 is unknown but appears to be a general characteristic of CLIC proteins, since other family members also migrate slower during SDS-PAGE than would be expected based on their predicted M_r values (4, 7). Nonetheless, these data clearly demonstrate that CLIC-1, when purified by these methods, exists in monomeric form.

Vesicular Chloride Efflux Mediated by CLIC-1-- Purified CLIC-1 was reconstituted into phospholipid vesicles at concentrations between 25 and 200 µg/ml and assayed alongside identically treated control samples for chloride channel activity using a potential-driven chloride efflux assay (11, 18). Vesicles were prepared by dialysis to contain 200 mM KCl. Extravesicular KCl was removed from an aliquot of the preparation by passage through a desalting spin column that had been equilibrated in 400 mM sucrose. The effluent from the spin column was added to 400 mM sucrose, and extravesicular [Cl] was monitored with a chloride-selective electrode. After a steady state was reached, valinomycin (final concentration = 1 µM) was added to initiate potential-driven chloride efflux. This concentration of valinomycin was previously found to support maximum efflux from the vesicles (data not shown). At the end of the experiment, detergent was added to release any remaining intravesicular chloride. The potential dependence of chloride efflux excludes ion exchange or non-selective leak mechanisms as possible explanations for our data.

Data from typical efflux experiments are shown in Fig. 2A. We determined the fraction of intravesicular chloride released per second after the addition of valinomycin as described previously (11). The valinomycin-dependent fractional chloride efflux rate was used as a measure of the chloride-selective permeability of the vesicles.

View larger version (22K): [in this window] [in a new window]   Fig. 2.   Chloride efflux from reconstituted vesicles. A, raw efflux data collected from control vesicles and from vesicles reconstituted with CLIC-1. A continuous recording of the electrode output is plotted against time. The arrows labeled val and TX mark the points of addition of valinomycin and Triton X-100, respectively. The upper trace is from control vesicles; the lower trace is from CLIC-1 vesicles. In this experiment, 0.68 and 5% of initial, intravesicular chloride had escaped from control and CLIC-1 vesicles, respectively, at the point of TX100 addition. B, protein dependence of CLIC-1-mediated Cl efflux. Valinomycin-dependent Cl efflux from a single reconstitution experiment comparing control vesicles and vesicles reconstituted with CLIC-1 at 25, 50, or 100 µg/ml is shown. Average rates of efflux and S.E. are plotted. Control rate: 0.14 ± 0.03%/s; CLIC-1 25 µg/ml rate: 0.33 ± 0.014%/s; CLIC-1 50 µg/ml rate: 0.47 ± 0.11%/s; CLIC-1 100 µg/ml rate: 0.62 ± 0.12%/s (n = 3 for each group). C, inhibition of CLIC-1-mediated Cl efflux by IAA-94. Efflux rates from vesicles from a single reconstitution were determined in the presence and absence of 50 µM IAA94. Mean efflux rates and S.E. are shown. Control rate: 0.553 ± 0.06%/s (n = 4); CLIC-1 rate: 1.64 ± 0.244%/s (n = 4); CLIC-1 + IAA rate: 0.67 ± 0.093%/s (n = 3). D, concentration Dependence of IAA inhibition. R/Ro, the ratio of the p34-dependent efflux rate at each concentration of IAA to the efflux rate in the absence of IAA is plotted against the log of the IAA concentration. The average of three data points with S.E. is shown. A least squares fit predicts the Ki to be 8.6 µM.

Twelve independent preparations of CLIC-1 and corresponding control fractions were reconstituted (protein concentration = 200 µg/ml reconstitution mixture) and analyzed for valinomycin-dependent chloride efflux. Data were averaged from at least three separate measurements to determine the chloride efflux rates of each preparation. Rates of chloride efflux were significantly greater than control vesicles for each preparation of reconstituted CLIC-1 tested. Rates varied somewhat from prep to prep, ranging from a 1.6- to a 5.03-fold increase in CLIC-1-mediated, valinomycin-dependent chloride efflux compared with control vesicles. On average, a 2.59-fold increase in rate of valinomycin-dependent chloride release was observed for CLIC-1 vesicles compared with control (S.E. = 0.299; 95% confidence interval 1.93-3.25-fold increase).

The dependence of chloride efflux upon CLIC-1 was determined using vesicles reconstituted at increasing CLIC-1:lipid ratios (Fig. 2B). As expected, the rate of chloride release increased with increasing CLIC-1:lipid ratios. Thus, CLIC-1 confers dose-dependent, chloride-selective permeability to reconstituted vesicles.

IAA-94 is a chloride channel inhibitor that was used as an affinity ligand for the purification of p64 (9, 23). To determine whether the CLIC-1 channel is sensitive to IAA-94, reconstituted vesicles were mixed with the appropriate concentration of IAA-94 before passage through the mini-extruder, ensuring equilibration of inhibitor between intra- and extravesicular spaces. After passage through spin columns, vesicles were diluted into sucrose containing the same concentration of IAA-94. Initial rates of chloride efflux from vesicles in the presence or absence of IAA-94 are presented in Fig. 2C. We found that 50 µM IAA-94 markedly inhibits CLIC-1-mediated chloride efflux. Vehicle alone (20 mM NaHCO₃ (pH 9.5)) had no effect on CLIC-1-mediated Cl efflux (not shown). The relationship between IAA-94 concentration and inhibition of CLIC-1 activity is shown in Fig. 2D. The IC₅₀ derived from this data is 8.6 µM.

Electrophysiological Characterization of CLIC-1-- The current-voltage relationship, single channel conductance, and voltage dependence of CLIC-1 were characterized using the planar lipid bilayer technique. Five-microliter aliquots of vesicles reconstituted with CLIC-1 or control preparations were added to the cis chamber at 5-min intervals in the presence of either a 300-50 or a 150-50 mM KCl gradient (cis to trans). Membranes were monitored continuously for fusion events, and if no fusion occurred within 30 min, the experiment was terminated. Using vesicles reconstituted with CLIC-1, active channels typically appeared in bilayers after the first or second addition of vesicles. Under these conditions, the ion selectivity (K⁺ versus Cl) of these channels was readily apparent.

Nine independent preparations of reconstituted, purified CLIC-1 were tested for chloride channel activity in the presence of 300-50 mM KCl and four in 150-50 mM KCl. In all preparations, a characteristic voltage-dependent, anion-selective channel was readily observed. Corresponding control vesicles were also tested and found to be silent, with the exception that K⁺ channels were observed in 3 of the preparations. These K⁺ currents were observed with the same frequency in CLIC-1 containing vesicles and, hence, appear to be unrelated to the presence of CLIC-1. The voltage-dependent chloride channels that were consistently observed following CLIC-1 vesicle fusions were never observed in control vesicles.

Fusion of CLIC-1-containing vesicles with planar membranes resulted in the appearance of a multichannel, steady-state current at 50-mV holding potentials. This large single current level resulted from all the channels in the membrane being in the open state, since total conductance did not increase above this level during the experiment, and all the channels could be closed at high membrane potentials (see below). Thus, it was possible to determine the number of channels fused into a membrane from the number of current levels observed as the channels sequentially closed. In a typical experiment, three active channels were observed in the bilayer following a single fusion event.

Single channel closures were occasionally observed at 50 mV but occurred more frequently as the membrane potential was gradually depolarized above +50 mV or hyperpolarized below 50 mV. Exposure to strong depolarizing or hyperpolarizing voltages of ±150 mV caused CLIC-1 to close with time (see Fig. 3A). After closure, reducing the membrane potential to between ±100 mV allowed CLIC-1 to re-open (see Fig. 3A). In most cases, all of the channels that were initially observed in the bilayer rapidly returned to the open state. Once re-opened, channels could subsequently be induced to close when holding potentials were returned to ±150 mV. Thus, CLIC-1 is a voltage-dependent channel.

View larger version (29K): [in this window] [in a new window]   Fig. 3.   Representative current traces from planar lipid bilayer experiments. Purified, reconstituted CLIC-1 was allowed to fuse to planar bilayers as described. Data represent CLIC-1 activity recorded in symmetric, 300 mM KCl. A, representative trace of channel closure following exposure to a +150-mV holding potential. Upon switching the voltage to 80 mV, all the channels re-opened, whereas switching back to +150 mV once again caused the channels to close with time. B and C, representative traces of channels entering into states where transitions from open to closed states occur more frequently following pulses at a ±150-mV holding potentials. The expanded recordings are each 1 s in duration.

Conditioning pulses of ±150 mV occasionally induced an intermediate state of channel activity where frequent transitions between open and closed states were observed when holding potentials were adjusted to between ±100 mV. Examples of this behavior are depicted in Fig. 3, B and C. We employed this maneuver to generate single channel data under various experimental conditions (see below).

To quantitate voltage-dependent channel closure, currents were recorded at several holding potentials in 10-mV increments between +100 and +150 mV. Between each voltage step, membranes were held at 0 mV until all the channels that had closed at the higher membrane potentials re-opened. Total current was normalized against the maximum current observed at each holding potential, and averaged data were plotted as a function of time. Time constants were determined from a single exponential fit of each plot as described under "Experimental Procedures." Data from representative voltages are shown in Fig. 4A. A plot of the time constants versus holding potential is shown in Fig. 4B. As is shown, the rate at which CLIC-1 closed increased with voltage. The same behavior was observed at potentials between 100 and 150 mV (not shown). Thus, CLIC-1 channels are very sensitive to strong hyperpolarizing (100 to 150 mV) and depolarizing (+100 to +150 mV) voltages but relatively insensitive to voltages between ±50 mV.

View larger version (20K): [in this window] [in a new window]   Fig. 4.   Initial characterization of voltage-dependent channel closure. Current traces recorded at holding potential pulses between +100 and +150 mV at 10 mV intervals were normalized to the maximum current and plotted versus time. A, averaged data from currents recorded at the indicated voltages. Single exponential fits of the data are superimposed on the plots. The predicted current at infinite time was 22.3, 0, 15, 10, 9.3, and 0% of initial current at 100, 110, 120, 130, 140, and 150 mV, respectively. B, time constants were determined from single exponential fits of the data represented in A and plotted (±S.E.) versus voltage. n = 2, 4, 4, 4, 5, and 2 for traces averaged at 100, 110, 120, 130, 140, and 150 mV, respectively.

Single channel data were obtained at holding potentials between ±100 mV following channel closure at ±150 mV. As noted above, this maneuver produced records with clearly resolved single channel transitions that were then analyzed to determine the single channel properties of CLIC-1. Representative traces of single channel activities recorded from CLIC-1 in asymmetric (300 mM cis, 50 mM trans) and symmetric (300 mM cis and trans) KCl are depicted in Fig. 5, panels A and B, respectively. The current-voltage relationships derived from such data are shown in panel C. Several features are apparent from this plot. The channel has a rectifying current-voltage relationship in symmetric KCl, which in this example is inward. In asymmetric KCl, the reversal potential is positive, indicating significant selectivity for Cl over K⁺. This activity was observed following 12 separate fusion events derived from 9 independent reconstitutions. Although the data in Fig. 5 represent inwardly rectifying channel activity, both inwardly and outwardly rectifying channels were observed following fusion events, indicating that CLIC-1 inserts in both orientations during reconstitution.

View larger version (23K): [in this window] [in a new window]   Fig. 5.   Single channel activity of CLIC-1. Current was recorded at a series of holding potentials in both asymmetric and symmetric KCl solutions following holding potential pulses of ±150 mV. Arrows represent closed levels. A, representative single channel traces from CLIC-1 recorded in 300-50 mM KCl (cis to trans). B, representative single channel traces from channels recorded in symmetric, 300 mM KCl. C, current-voltage relationships for the CLIC-1 channel represented in A (circles) and B (squares).

The data obtained at negative holding potentials (i.e. the linear region of current-voltage plots) recorded in symmetric 300 mM KCl were fit by linear regression to determine CLIC-1 conductance, yielding an average single channel conductance of 161 ± 7.9 pS (n = 12). The E_rev, obtained from recordings in 300-50 mM KCl gradients, was 23.7 ± 2.6 mV (n = 10), which corresponds to a Cl to K⁺ selectivity ratio of 4.2:1. The current-voltage relationship was also determined in physiologic buffer (i.e. symmetric 150 mM KCl) and found to be 67.5 ± 6.9 pS (n = 4; data not shown). In a 150-50 mM KCl gradient, E_rev was 19 mV, indicating a Cl:K⁺ selectivity of 5.8:1 (n = 3).

An anion selectivity sequence was determined from the E_rev of CLIC-1 channel activity recorded in 300 mM KBr or KI cis, 300 mM KCl trans. A plot of the current-voltage relationships obtained under these conditions and compared with the data obtained in symmetric 300 mM KCl is depicted in Fig. 6. E_rev for Br and I were found to be 0.27 and 20.1 mV, respectively (n = 3). These values correspond to an anion selectivity series of Br (1.011)  Cl (1.000) > I (0.451).

View larger version (13K): [in this window] [in a new window]   Fig. 6.   CLIC-1 ion selectivity. CLIC-1 channel activity was recorded at a series of holding potentials in the presence of 300 mM KBr (open squares; n = 3), KCl (closed squares; n = 3), or KI (closed circles; n = 2) cis; 300 mM KCl trans. Data represent the average current ±S.E.

The single channel activity of CLIC-1 was inhibited by IAA-94. Currents were recorded over 5-min intervals in the absence and presence of several IAA-94 concentrations in the trans chamber. Representative traces at 0 and 500 µM IAA-94 are depicted in Fig. 7, panels A and B, respectively. The P_o was determined from 5-min traces recorded at each concentration of inhibitor, and the results are plotted in Fig. 7C. P_o did not appear to vary significantly within each 5-min trace. In this experiment, the P_o in the absence of inhibitor was 0.699. The dose-response curve shown in Fig. 7C was generated using a sigmoidal fitting function as described under "Experimental Procedures," yielding an IC₅₀ of 86 µM and a Hill coefficient of 1.4. Analysis of channel kinetics at each inhibitor concentration revealed that the mean closed time did not vary with [IAA-94], but mean open time decreased with increasing [IAA-94] (see legend of Fig. 7 for values).

View larger version (23K): [in this window] [in a new window]   Fig. 7.   IAA-94 inhibition of CLIC-1 channel activity. CLIC-1 activity was recorded in symmetric 300 mM KCl in the presence of increasing [IAA-94]. Panels A and B represent activity recorded at +50 mV in the absence and presence of 500 µM IAA-94 in the trans chamber, respectively. Panel C depicts a plot of the Po versus inhibitor concentration. Fitting the data to a sigmoidal function generated a dose-response curve. An IC50 of 86 µM and a Hill coefficient of 1.4 were derived from the fit. Mean open times were 215, 214, 145, 124, and 81 ms at IAA-94 concentrations of 0, 10, 100, 200, and 500 µM, respectively.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

In this paper we have demonstrated that the protein CLIC-1, when expressed in bacteria and purified to apparent homogeneity, is capable of conferring chloride-selective permeability to reconstituted phospholipid vesicles and gives rise to a discrete, IAA-94 inhibitable anion channel with distinctive single channel properties. These observations represent the first evidence that CLIC family proteins themselves are capable of forming anion channels in the absence of ancillary subunits or accessory proteins.

CLIC Proteins and Chloride Channels-- The name CLIC (for chloride intracellular channel) has recently been proposed for the family of proteins related to the bovine microsomal chloride channel p64 (7, 23). In addition to p64, at least six other members of the CLIC family have been described (7).

CLIC-1 (formerly NCC27) is highly homologous to the C-terminal half of p64. CLIC-1 is expressed in intracellular membranes in many cell types and in the apical domain of renal proximal tubule cells (12). Expression of CLIC-1 in Chinese hamster ovary-1 cells has been reported to lead to an increase in the chloride permeability of the plasma membrane and in the appearance of two distinct, low conductance chloride channels (13).

CLIC-4 (or p64H1), like CLIC-1, is homologous to the C-terminal half of p64 (7, 24). We recently demonstrated that this protein colocalizes with a marker of trans-Golgi network and of caveolae. Expression of this protein in HEK293 cells is associated with the appearance of an anion-permeable channel (24). CLIC-4 is associated with large, dense-core vesicles in rat hippocampal neurons, where it is hypothesized to provide the shunt pathway for Cl that is necessary for acidification to occur (25). The large, dense-core vesicles are similar to those with which p64 associates (4) and are most likely exocytic vesicles.

CLIC-2 has been identified by a genomic sequencing approach (26). No information regarding subcellular distribution or function has been published. CLIC-3 was identified using a yeast 2-hybrid screening approach, with the mitogen-activated protein kinase ERK7 being used as bait (14). Novel Cl-selective currents were observed in LTK cells transfected with CLIC-3. The chloride channel of osteoclast ruffled membrane, p62, is antigenically related to the CLIC family (27). The cDNA sequence of this homologue has not yet been reported. Finally, parchorin, a newly described member of the family, appears to potentiate Cl efflux when expressed in LLC-PKI cells (28).

Thus, the pre-existing data regarding CLIC family members have consistently implicated these proteins in various chloride channel activities. However, even in those cases where expression of a CLIC family protein is correlated with the appearance of a novel current, the nature of the expression systems used make it impossible to determine whether the CLIC protein alone is responsible for the observed activity. Alternative possibilities include the necessity of endogenous cofactors for assembly of active channels or the activation of an endogenous channel by CLIC proteins.

The experiments presented here address this problem directly. CLIC-1 was expressed in E. coli and purified to apparent homogeneity, so the activity observed upon reconstitution into phospholipid vesicles must be attributable to CLIC-1 alone. Chloride channel activity was assessed using two complementary assays. A population-based chloride efflux assay was used to demonstrate the presence of macroscopic chloride permeability associated with reconstituted CLIC-1, and planar lipid bilayer studies were used to characterize the single channel properties of this activity. We conclude that CLIC-1 is capable of forming anion-selective channels in the absence of other subunits or proteins.

CLIC-1-mediated Chloride Efflux-- Using a vesicle efflux assay, we demonstrated reproducible chloride efflux activity that was easily detectable above the background anion leak of reconstituted vesicles. Three aspects of the data strongly support the conclusion that this activity is due to CLIC-1. First, in each case, the control was prepared identically from the same bacterial strain expressing the parent pGEX-KG expression vector. Therefore, the efflux rates we attribute to CLIC-1 are not due to some low level contaminant recovered from the bacteria (which would also be present in control). Second, the chloride channel inhibitor IAA-94, which is known to interact with one CLIC family member (p64), inhibits efflux activity at reasonable concentrations. Our observed IC₅₀ of 8.6 µM is very close to the K_i of 9 µM reported for IAA inhibition of an epithelial outwardly rectifying chloride channel in the HT₂₉ colon carcinoma cell line (29) and about an order of magnitude higher than the K_i (1 µM) observed for inhibition of chloride permeability of kidney microsomal membranes (22). Third, the efflux activity we attribute to CLIC-1 is dose-dependent and therefore cannot be due to a nonspecific leakiness of our reconstituted vesicles.

Single Channel Properties of CLIC-1-- In nine independent preparations of CLIC-1, a Cl selective channel with a rectifying current-voltage relationship and an apparent single channel conductance of 161 pS in symmetric 300 mM KCl was observed following 12 separate membrane fusion events with CLIC-1-containing vesicles. Of note, we found this activity present in both inwardly and outwardly rectifying orientations, suggesting that the polarity of reconstitution was not fixed. In symmetric 150 mM KCl, we found the CLIC-1 conductance to be 67.5 pS (n = 4). Using asymmetric salt conditions, the Cl to K⁺ permeability ratio of this channel was shown to be 4.2:1 in 300-50 mM KCl and 5.8:1 in 150-50 mM KCl. In 300 mM K⁺(anion) cis and KCl trans, the anion selectivity series is Br  Cl > I.

IAA-94 was found to inhibit CLIC-1 chloride channels reconstituted into planar bilayers. At a +50-mV holding potential, IAA-94 was found to decrease the open probability of CLIC-1 with an IC₅₀ of 86 µM. This value is an order of magnitude greater than the IC₅₀ of 8.6 µM determined from efflux experiments. The reason for this discrepancy is unknown but may be due to differences in membrane potential or ionic strength. The two assay systems used to characterize CLIC-1 measure different properties of the channel (conductance versus rate of Cl efflux) and are carried out under different conditions. KCl-loaded vesicles reconstituted with CLIC-1 would have a very large membrane potential when diluted into sucrose in the presence of valinomycin, which would drive IAA-94 into the channel more efficiently than a +50-mV holding potential if IAA inhibition is voltage-dependent. The inhibition should also be relieved upon a reversal of membrane polarity. Consistent with this hypothesis, channels studied in bilayers and inhibited by IAA-94 were consistently found to re-open when membrane polarity was reversed (data not shown).

Finally, the CLIC-1 channel is voltage-dependent. In most experiments, the channel rarely closed at voltages between ±50 mV. Channel closures occurred more frequently as membranes were hyperpolarized or depolarized to a greater extent. Voltages between +100 and +150 mV caused the current to decrease with time in a dose-dependent manner, and the steady state current dropped to 0 at +150 mV. Similar observations were made at voltages between 100 and 150 mV. The mechanism whereby voltage closes CLIC-1 is unknown at this time.

The chloride channel characterized in this paper was consistently observed in our CLIC-1 preparations and was never observed in control preparations. The occasional contaminants we did observe were present in control preparations and occurred with the same frequency.

With the exception of single channel conductance, the electrical properties we determined for purified CLIC-1 in planar lipid bilayers were consistent with those reported by Valenzuela et al. (13), who found that heterologous expression of CLIC-1 in Chinese hamster ovary cells leads to the appearance of an outwardly rectifying, whole cell Cl conductance that is more selective for Cl than I at the single channel level. These investigators report single channel conductances in physiologic buffers of 22 ± 5 and 33 ± 4 pS for CLIC-1 expressed either on the plasma or nuclear membrane, respectively. We found that CLIC-1 has a conductance of 67.5 ± 6.7 pS in symmetric, 150 mM KCl when measured using the planar lipid bilayer technique.

There are several plausible explanations for the discrepancy observed in CLIC-1 single channel conductance. First, in native membranes CLIC-1 may associate with other subunits or proteins that were absent from our experiments, and these interactions could alter CLIC-1 activity. Second, it is possible that heterologous expression of CLIC-1 in mammalian cells initiates a cascade of events resulting in activation of other channels, which would have been detected by Valenzuela et al. (13) as elevated whole cell conductance. Finally, the short amino acid sequence present at the N terminus of the CLIC-1 construct we used in these experiments could alter the electrical properties of the native protein. We are currently investigating this possibility.

The voltage inactivation phenomena we observed was not reported by Valenzuela et al. (13), so it is not known whether CLIC-1 responds similarly to strong hyperpolarizing or depolarizing voltages in vivo. However, similar patterns of voltage inactivation have been reported for other channels, such as the batrachotoxin-inhibited, voltage-dependent Na⁺ channel isolated from rat brain (30). Using a planar lipid bilayer approach, French et al. (30) found that this channel had a P_o close to 1.0 at 60 mV and that hyperpolarization to 110 mV caused the P_o to drop significantly below 1.0. The voltage-dependent anion channel also behaves in this fashion in planar lipid bilayers. Colombini (31) reports that voltage-dependent anion channel P_o increased from approximately 0.4 at 80 mV to 1.0 when membranes were depolarized to 0 mV, then fell progressively back to approximately 0.4 again upon further polarization to +80 mV (31).

At least one other rectifying chloride channel has been reported to exhibit inactivation at strong hyperpolarizing and depolarizing voltages. Welsh et al. (32) report that an outwardly rectifying chloride channel (ORCC) present in human airway epithelia is inactivated at 120-mV holding potentials and re-activated at positive holding potentials. In another report, these investigators demonstrated that the ORCC also inactivates with time at +120-mV holding potentials following pre-pulses at 80 mV holding potentials (33). Both whole cell conductance and average single channel current decreased with time at +120-mV holding potentials.

Protein kinase A-dependent activation of CFTR leads to activation of the ORCC observed by Welsh et al. (32) in normal airway epithelial cells. Several investigators report a slope conductance for this channel of approximately 40 pS in physiologic buffers (34), which is similar to the conductances reported for CLIC-1 channels by Valenzuela et al. (13) following expression in Chinese hamster ovary-1 cells (see above). Thus, based on voltage inactivation, conductance, and current-voltage characteristics, it is tempting to speculate that CLIC-1 is the ORCC activated by CFTR in epithelial cells. However, the ion selectivities of these two channels differ significantly (Br  Cl > I versus I > Br > Cl, respectively), so it seems unlikely that CLIC-1 itself is the CFTR-associated ORCC. Perhaps another member of the CLIC family of proteins is responsible for the CFTR-associated ORCC.

	ACKNOWLEDGEMENT

We thank Dr. Qais Al-Awqati, Columbia University, New York, NY, for generously providing the IAA-94.

	FOOTNOTES

^* This work was supported by National Institutes of Health Grants R29 DK46212 and RO1 AR44838 and grants from the Barnes-Jewish Hospital Foundation.The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed: Renal Division 657/111-JC, St. Louis VA Medical Center, 915 N. Grand Blvd., St. Louis, MO 63106. Tel.: 314-652-4100 (ext. 5302); Fax: 314-289-7012; E-mail: John.Edwards3@med.va.gov.

Published, JBC Papers in Press, June 28, 2000, DOI 10.1074/jbc.M004301200

	ABBREVIATIONS

The abbreviations used are: CFTR, cystic fibrosis transmembrane conductance regulator; GST, glutathione S-transferase; DTT, dithiothreitol; PAGE, polyacrylamide gel electrophoresis; P_o, open probability; E_rev, reversal potential; A₂₈₀, absorbance of 280-nm wavelength light; IAA-94, indanyloxyacetic acid-94; pS, picosiemens; ORCC, outwardly rectifying chloride channel.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				H. Singh and R. H. Ashley Redox Regulation of CLIC1 by Cysteine Residues Associated with the Putative Channel Pore Biophys. J., March 1, 2006; 90(5): 1628 - 1638. [Abstract] [Full Text] [PDF]

				M. Berryman, J. Bruno, J. Price, and J. C. Edwards CLIC-5A Functions as a Chloride Channel in Vitro and Associates with the Cortical Actin Cytoskeleton in Vitro and in Vivo J. Biol. Chem., August 13, 2004; 279(33): 34794 - 34801. [Abstract] [Full Text] [PDF]

				G. Novarino, C. Fabrizi, R. Tonini, M. A. Denti, F. Malchiodi-Albedi, G. M. Lauro, B. Sacchetti, S. Paradisi, A. Ferroni, P. M. Curmi, et al. Involvement of the Intracellular Ion Channel CLIC1 in Microglia-Mediated {beta}-Amyloid-Induced Neurotoxicity J. Neurosci., June 9, 2004; 24(23): 5322 - 5330. [Abstract] [Full Text] [PDF]

				D. R. Littler, S. J. Harrop, W. D. Fairlie, L. J. Brown, G. J. Pankhurst, S. Pankhurst, M. Z. DeMaere, T. J. Campbell, A. R. Bauskin, R. Tonini, et al. The Intracellular Chloride Ion Channel Protein CLIC1 Undergoes a Redox-controlled Structural Transition J. Biol. Chem., March 5, 2004; 279(10): 9298 - 9305. [Abstract] [Full Text] [PDF]

				K. L. Berry, H. E. Bulow, D. H. Hall, and O. Hobert A C. elegans CLIC-like Protein Required for Intracellular Tube Formation and Maintenance Science, December 19, 2003; 302(5653): 2134 - 2137. [Abstract] [Full Text] [PDF]

				X. X. Li and M. Colombini Catalyzed Insertion of Proteins into Phospholipid Membranes: Specificity of the Process Biophys. J., November 1, 2002; 83(5): 2550 - 2559. [Abstract] [Full Text] [PDF]

				F. Thevenod Ion channels in secretory granules of the pancreas and their role in exocytosis and release of secretory proteins Am J Physiol Cell Physiol, September 1, 2002; 283(3): C651 - C672. [Abstract] [Full Text] [PDF]