Expression of volume-sensitive Cl⁻ channels and ClC-3 in acinar cells isolated from the rat lacrimal gland and submandibular salivary gland

Abstract

1. The expression of ClC-3 was examined in rat lacrimal gland and submandibular salivary gland cells using RT-PCR and Western analysis. Whole-cell patch clamp methods were used to investigate the expression of volume-sensitive anion channels in acinar cells isolated from these tissues.

2. Expression of mRNA encoding ClC-3, and ClC-3 protein, was found in rat submandibular gland by RT-PCR and Western analysis. Rat lacrimal gland cells, however, did not appear to express mRNA encoding for ClC-3, nor the ClC-3 protein.

3. Volume-sensitive anion conductances were observed in both rat lacrimal gland and submandibular salivary gland acinar cells. The conductance was of a similar size in the two cell types, but was much slower to activate in the lacrimal cells.

4. The properties of the conductances in lacrimal and submandibular cells were similar, e.g. halide selectivity sequence (P_I > P_Cl > P_aspartate) and inhibition by 4,4′-diisothiocyanostilbene-2,2′-disulphonic acid, 5-nitro-2-(3-phenylpropylamino)-benzoate and tamoxifen.

5. The data suggest that the expression of ClC-3 is not an absolute requirement for the activity of volume-sensitive anion channels in rat lacrimal gland acinar cells.

Volume-sensitive anion channels are expressed in most mammalian cells (Nilius et al. 1996; Strange et al. 1996; Okada, 1997). These are channels which, when activated by cell swelling, mediate the efflux of Cl⁻, organic anions and amino acids during cell volume regulation. The broad ion and substrate selectivity displayed by these channels has led them to be called volume-sensitive osmolyte and anion channels or VSOAC by some groups (for review see Strange et al. 1996). There is, however, some debate as to whether a single channel mediates the efflux of anions and other osmolytes or whether a group of channel proteins each with similar properties is involved (Mongin et al. 1999; Stutzin et al. 1999; Valverde, 1999).

This confusion over the phenotype of the channel or channels involved is further complicated by a lack of precise information about their molecular biology. Several candidate proteins have been identified including p-glycoprotein, pI_Cln, ClC-2 and ClC-3 (Okada et al. 1998). However, there is little definitive evidence to support the role of these proteins as volume-sensitive anion channels, and at least one of them, p-glycoprotein, is now thought to function as a channel regulator rather than the actual channel (Okada, 1997). The properties of ClC-3 make it one of the most likely candidate proteins, e.g. it has a structure which is very similar to that of known Cl⁻ channels (ClC-0 and ClC-1), and it produces an outward-rectifying Cl⁻ conductance when expressed in Xenopus oocytes (Kawasaki et al. 1994) or mammalian cell lines (Duan et al. 1997; Valverde, 1999; Shimada et al. 2000). However, others have been unable to express ClC-3 in oocytes (for reviews see Jentsch et al. 1999; Valverde, 1999). Furthermore, when transfected into mammalian cells, ClC-3 gives rise to currents in the absence of cell swelling (Duan et al. 1997; Valverde, 1999; Shimada et al. 2000), although cell swelling does appear to further activate the channels (Duan et al. 1997; Valverde, 1999). The role of ClC-3 as a volume-sensitive anion channel is therefore still somewhat controversial.

Rat lacrimal gland acinar cells are unusual because volume regulation is not thought to involve volume-sensitive anion channels (Strange et al. 1996). Instead, cell swelling causes an increase in intracellular Ca²⁺, which is sufficient to activate Ca²⁺-activated Cl⁻ channels allowing Cl⁻ efflux (Kotera & Brown, 1993; Park et al. 1994; Speake et al. 1998). Lacrimal acinar cells may therefore provide a useful natural ‘knock-out’ in which to study the role of ClC-3, i.e. they can be used to test the hypothesis that ClC-3 is not expressed in cells which do not exhibit volume-sensitive anion channels. The present study has examined the expression of ClC-3 and volume-sensitive anion channels in rat lacrimal acinar cells, using molecular biological and electrophysiological methods respectively. Expression was also examined in salivary gland acinar cells, which are similar in many respects to lacrimal cells, but are known to express volume-sensitive anion channels (Arreola et al. 1995, 1996).

METHODS

Animals and cells

Lacrimal glands, submandibular salivary glands and brain (positive control for ClC-3) were isolated from Sprague-Dawley rats (200–250 g), which were killed by an overdose of Fluothane inhalation (Zeneca), in accordance with Schedule One Procedures specified by the UK Home Office. The tissues were washed in phosphate-buffered saline (Gibco). They were then either snap-frozen and stored in liquid N₂ for use in protein and mRNA determination, or the gland tissue was used immediately for the preparation of acinar cells for patch clamp experiments. In PCR experiments, cDNA prepared from rat insulinoma cells (RINm5F cell line; ECACC, Salisbury) was also used as a positive control (Majid et al. 2000).

Isolation of mRNA, reverse-transcription and PCR

Total RNA was prepared from tissue using an acid phenol- guanidium isothiocyanate method based on that of Chomczynski & Sacchi (1987). mRNA was extracted using magnetic beads (Dynal, Norway), following the manufacturer's instructions. One microgram mRNA was DNase (Gibco) treated, and then reverse transcribed using the avian myeloblastosis virus RNA-dependent DNA polymerase (Boehringer Mannheim, Germany) at 42 °C for 60 min. RT-negative controls were performed to exclude the possibility of genomic or other DNA contamination. Initial experiments involved the use of degenerate primers for the ClC family of channels (forward: 5′CCGGATCCGGSTCYGGMMTCCCNGARMTGAAARA-C3′ and reverse: 5′CCGAATTCNACCTCDATGCTGAANAGSACNCC3′), designed using the Primer Select program (DNASTAR, USA) and published sequences (Kawasaki et al. 1994). Primer sequences specific for ClC-3, 5′GATTCATCATCAGAGGT3′ (nucleotides 1891–1911) for the forward primer, and 5′CGCACCAAAAGCCAC3′ (nucleotides 2130-2150) for the reverse primer, were also employed (designed using the published sequences for rat ClC-3; Kawasaki et al. 1994). Reaction mixtures consisted of 25 μl final volume containing: 10 × PCR buffer (500 mm KCl, 1 % Triton X-100, 100 mm Tris; pH 9), 2.5 mm of each dNTP, 10 pm of each primer, 1.5 mm MgCl₂ and 5 U Taq polymerase (Promega). PCR was performed in a Hybaid thermal cycler (Hybaid). Cycle parameters were: initial denaturation at 94 °C for 5 min, followed by 30 cycles of 94 °C for 1 min, 50 °C for 1 min and 72 °C for 1 min and a final extension at 72 °C for 10 min. To verify the integrity of the cDNA used in these reactions, PCR was also performed using β-actin primer pairs (Promega), and primer pairs for carbonic anhydrase II (forward: 5′ATGTCCCACCACTGGGGATAC3′ (nucleotides 9–29) and reverse: 5′ATGGTGGACAACTGGCATCCA3′ (nucleotides 727–747), designed by Dr P. Hurley, University of Manchester, UK). PCR products were electrophoresed on a 1.5 % agarose Tris-EDTA (pH 8.0) gel at 80 V h⁻¹, and visualised using ethidium bromide. Bands were also excised and products characterised by automated sequencing using the dideoxynucleotide termination method.

Southern analysis

PCR products were transferred overnight onto a nylon membrane (Stratagene) in 10 × sodium salt citrate (SSC; Sigma). The ‘blot’ was washed in diethyl pyrocarbonate (Sigma)-treated water, and cross-linked in a UVP1000 oven (UVP, Cambridge). Blots were prehybridised for 30 min in 5 × SSC, 1 % sodium dodecyl sulphate (SDS), 5 % dextran sulphate and a 1 in 20 dilution of blocking buffer (Amersham Pharmacia Biotech) at 65 °C. They were then hybridised overnight with a 1700 bp partial cDNA probe for ClC-3 spanning nucleotide positions 500–2200 (from Dr B. Horowitz, University of Reno, NV, USA), a 1800 bp partial ClC-2 clone spanning nucleotide positions 683–2483 (from Dr T. J. Jentsch, University of Hamburg, Germany) or full-length clones for carbonic anhydrase II and β-actin (from Dr P. Hurley). The probes were labelled using the Gene Image Random Prime labelling kit following the manufacturer's instructions (Amersham Pharmacia Biotech). After hybridisation, blots were washed in 50 ml buffer A (100 mm Tris HCl, 300 mm NaCl; pH 9.5) and a 1 in 10 dilution of blocking buffer for 1 h at room temperature. The blot was then incubated with anti-fluorescein-AP-conjugated antibody (Amersham; 1 in 5000 dilution) containing 0.5 % BSA in buffer A. Excess antibody was removed by washing in 0.3 % Tween 20 and buffer A (three 10 min washes at 20 °C). Blot detection was carried out using an ECL kit (Amersham) followed by exposure to autoradiograph film overnight.

Western analysis

Crude membrane extracts were prepared from frozen tissue by homogenisation in solubilisation buffer containing: 8.3 mm Tris (pH 7.4), 125 mm NaCl, 1.25 μm pepstatin, 4 μm leupeptin, 4.8 μm phenylmethylsulfonyl fluoride and 1 % (v/v) Triton X-100. The tissue suspension was then centrifuged at 2500 g (15 min) to remove nuclei, etc., and the resulting supernatant was centrifuged at 100 000 g (30 min) to yield a crude membrane pellet. These pellets were resuspended in 200–500 μl fresh solubilisation buffer. The protein content of the samples was determined by the method of Bradford (1976).

Twofold-concentrated Laemmli buffer (0.32 m Tris at pH 6.8, 5 % (w/v) SDS, 25 % (v/v) glycerol, 1 % (w/v) bromophenol blue) was added in the ratio 1:1 to the membrane suspension at 65 °C for 3 min. The samples (each of 70 μg total protein) were resolved using 8 % SDS-polyacrylamide gels. Proteins were then transferred electrophoretically to polyvinylidene difluoride membranes in blotting buffer (25 mm Tris, 200 mm glycine and 15 % (v/v) methanol), containing 0.5 % SDS to improve protein transfer. Membranes were incubated for 1 h at 20 °C in 5 % (w/v) powdered milk dissolved in Tris-buffered saline-Tween (TBST; 150 mm NaCl, 0.1 % (v/v) Tween 20, 15 mm Tris), to block non-specific binding sites, followed by overnight incubation at 4 °C with a 1: 300 dilution of anti-ClC-3 rat polyclonal antibody (Alomone, Israel). After washing in TBST to remove all non-specifically bound antibody, membranes were incubated with anti-rabbit IgG secondary antibody (1:500 dilution; Jackson ImmunoResearch Laboratories Ltd, USA) for 1 h. The blots were developed using an ECL kit following the manufacturer's instructions. In control experiments the antibody was pre-incubated for 1 h (room temperature) in the presence of a 1:3 excess of the antigen (provided by Alomone), before addition to the membranes and overnight incubation.

Patch clamp recording

Patch clamp experiments were performed on single acinar cells isolated from the lacrimal and submandibular glands. Cells from both glands were prepared by very similar methods. The isolated gland tissue was first minced with scissors in a Ca²⁺-free solution containing (mm): 130 NaCl, 4.5 KCl, 1 NaH₂PO₄, 1 MgSO₄, 10 d-glucose, 10 NaHepes-NaOH at pH 7.4. The minced tissue was incubated for 20 min at 37 °C in Ca²⁺-free solution containing 0.4 mg ml⁻¹ trypsin (Type II-S, Sigma). This was followed by a second incubation for 60 min in Ca²⁺-free solution containing 2 mg ml⁻¹ trypsin inhibitor (Sigma), 100 U ml⁻¹ collagenase (Worthington, Lakewood, UK) and 1 % BSA. The tissue was mechanically dissociated by repeated pipetting through a 1 ml pipette tip at 20 min intervals during the incubation. A third 30 min incubation was performed only with the submandibular gland tissue, in Ca²⁺-free solution containing 50 U ml⁻¹ collagenase. The isolated cells were harvested at the end of each incubation by filtration though a nylon mesh (200 μm²), followed by centrifugation. The cells were resuspended in a solution containing (mm): 140 NaCl, 5 KCl, 1 MgCl₂, 1 CaCl₂, 2 glutamine, 10 Hepes-NaOH at pH 7.4, and stored on ice until use.

Channel activity was measured using conventional whole-cell recording methods. Patch pipettes were manufactured from haematocrit capillaries (Oxford Labware, USA). They had a tip resistance of 3–5 MΩ. In most experiments the pipette contained a low-Cl⁻ solution (Table 1). However, the anionic selectivity of the conductance was examined in experiments using a high-Cl⁻ pipette solution (Table 1). In addition, a hypertonic pipette solution (330 mosmol (kg H₂O)⁻¹) was used in a few experiments; this consisted of the low-Cl⁻ solution plus 30 mm mannitol.

Table 1.

Pipette and bath solutions for the patch clamp experiments

	Pipette solutions		Bath solutions
	Low Cl−	High Cl−	Control	Hypotonic	Aspartate	Iodide
[NaCl] (mm)	20	135	140	98	7	7
[Sodium aspartate] (mm)	110	—	—	—	91	—
[NaI] (mm)	—	—	—	—	—	91
[CaCl2] (mm)	—	—	1	0.7	0.7	0.7
[MgCl2] (mm)	3	3	1	0.7	0.7	0.7
[Na2ATP] (mm)	5	5	—	—	—	—
[BAPTA] (mm)	5	5	—	—	—	—
[D-Glucose] (mm)	10	10	10	7	7	7
[Hepes] (mm)	10	10	10	7	7	7
pH	7.2	7.2	7.4	7.4	7.4	7.4
Osmolality (mosmol(kgH2O)−1)	300	300	306	213	213	213

The pH of solutions was adjusted using NaOH. Solution osmolalities were measured by the freezing-point depression method using a Roebling osmometer (Camlab) and adjusted to the desired value by the addition of mannitol.

Whole-cell currents were monitored using an Axopatch 200 amplifier (Axon Instruments, CA, USA). Command potentials were generated by pCLAMP 6 software (Axon Instruments), and the resultant whole-cell currents recorded to hard disk using the same package. The cells were maintained at a holding potential (V_m) of either −40 or −60 mV. Current profiles were generated by applying 0.5 or 1 s step potentials from −100 to 100 mV at 20 mV increments. The time course for channel activation was monitored by repeatedly applying 0.4 s step potentials to −60 and 60 mV (at 2 s intervals), and the resultant currents were recorded on DAT tape using a modified DAT recorder (DTR1204, Bio-logic, France). The bath solution was connected to ground using a Ag-AgCl pellet pipette in most experiments; however, an agar-salt bridge was used in the experiments studying anion selectivity.

Cells in the experimental chamber (volume, 400 μl), were superfused at a rate of 2 ml min⁻¹ with an isotonic control bath solution (Table 1). Cell swelling was produced using the hypotonic solution (Table 1). The rate of cell volume changes on exposure to hypotonic solution was measured in some experiments using a video-imaging method (Park et al. 1994). The anionic selectivity of the conductance was assessed using hypotonic bath solutions in which 91 mm NaI or 91 mm sodium aspartate replaced 91 mm NaCl (Table 1). 4,4′-Diisothiocyanostilbene-2,2′-disulphonic acid (DIDS), 5-nitro-2-(3-phenylpropylamino)-benzoate (NPPB) and tamoxifen (Sigma) were used as Cl⁻ channel blockers (prepared as 10 mm stock solutions in dimethyl sulphoxide). All experiments were performed at room temperature (18–22 °C).

Data are presented as means ±s.e.m. (n, number of cells). Differences between mean values were assessed by Student's t test for unpaired data or by ANOVA followed by Bonferroni's post hoc test. P < 0.05 was considered significant.

RESULTS

Expression of ClC-3 in acinar cells

Preliminary experiments using Northern analysis failed to provide clear evidence for the expression of mRNA for ClC-3 in submandibular or lacrimal gland tissue. RT-PCR methods were therefore employed to increase the sensitivity of mRNA detection. Initial experiments were performed using PCR primers degenerate for ClC-1, ClC-2, ClC-K1 and ClC-3 channels. Figure 1A shows that PCR products (∼300 bp) were obtained using these primers and cDNA isolated from brain (P), submandibular gland (S), lacrimal gland (L) and RINm5F cells (P_β). In a Southern analysis the products from brain and submandibular gland, but not lacrimal gland, hybridised to the ClC-3 cDNA probe (Fig. 1B). Sequencing was also used to confirm that these products were derived from ClC-3 (97 % homology; Kawasaki et al. 1994). The products from lacrimal gland (and also brain, submandibular gland and RINm5F cells) did, however, hybridise with a ClC-2 cDNA probe (data not shown). Sequencing confirmed that these products were derived from ClC-2 (99 %; Thiemann et al. 1992). A further round of PCR on the degenerate products from lacrimal gland, using ClC-3-specific primers, failed to yield any products (data not shown).

Figure 1. Expression of mRNA encoding for ClC-3 in rat submandiular gland, but not in rat lacrimal gland.

A, PCR products obtained using degenerate primers for the ClC family from cDNA prepared from brain (P), lacrimal gland (L), submandibular gland (S) and rat insulinoma (RINm5F) cells (P_β). N, negative control. B, Southern analysis of the degenerate primer products using a cDNA probe for ClC-3. C and D, PCR products obtained using ClC-3 primers (predicted product size, 259 bp; C) and the corresponding Southern analysis of ClC-3 (D). E and F, PCR products obtained using primers for carbonic anhydrase II (predicted product size, 738 bp; E) and Southern analysis of these products using the cDNA probe for carbonic anhydrase II (predicted product size, 738 bp; F).

In a second series of experiments, RT-PCR was performed on cDNA from lacrimal gland, submandibular gland, brain and RINm5F cells using primers specific for ClC-3. A single 259 bp PCR product (i.e. the expected size for the primers used) was obtained with the ClC-3 primers and cDNA from the salivary gland, RINm5F cells and brain tissue (Fig. 1C). No product, however, was obtained from the lacrimal gland cDNA (Fig. 1C). The PCR products were characterised by Southern analysis using a cDNA probe for ClC-3. Figure 1D shows that the cDNA probe hybridised to the PCR products from the positive controls and salivary gland. The Southern blot (Fig. 1D) also confirmed that a ClC-3-derived product was not obtained from the lacrimal gland. The products from the submandibular gland and brain were sequenced and showed 98 % homology with the published sequence for ClC-3 (Kawasaki et al. 1994). In control experiments using primers for β-actin (data not shown) or carbonic anhydrase II (Fig. 1E), PCR products were generated from the cDNA for all four tissues. The products hybridised with the cDNA probes to β-actin (not shown) and carbonic anhydrase II (Fig. 1F). These experiments confirmed the integrity of the cDNA prepared from lacrimal gland (and the other tissues).

Expression of ClC-3 was examined at the protein level by Western analysis. Figure 2A shows that brain and salivary gland expressed a single protein which cross-reacted with the ClC-3 antibody. The immunoreactive protein had a molecular mass of approximately 80 kDa. Immunoreactive bands were not observed when the blot was incubated with the antibody in the presence of an excess of the antigen (Fig. 2B), indicating that the antibody is binding specifically to ClC-3 protein in the brain and salivary gland. The antibody did not cross-react with any protein in the lacrimal gland. Thus, no evidence for ClC-3 protein expression in the rat lacrimal gland was obtained by Western blotting, and results from RT-PCR suggested that lacrimal cells express little, if any, ClC-3 mRNA. By contrast, ClC-3 protein and mRNA were both readily detected in the submandibular salivary gland.

Figure 2. ClC-3 protein is expressed in rat submandibular gland, but not in rat lacrimal gland.

A, Western analysis was performed on crude membrane fractions prepared from brain (P), lacrimal gland (L) and submandibular gland (S), using a ClC-3 antibody. B, control experiment in which the primary antibody was used in the presence of an excess of antigen.

Volume-sensitive anion channels in rat lacrimal and submandibular acinar cells

Figure 3 shows the effects of hypotonic solutions on whole-cell Cl⁻ currents in submandibular gland acinar cells (A) and lacrimal gland acinar cells (B). Increases in the outward (V_m= 60 mV) and inward (V_m= −60 mV) currents were observed in the two cell types on exposure to hypotonic solution. The rate of Cl⁻ current activation, however, appeared to be much greater in the submandibular gland cells than in the lacrimal gland cells. There was also a considerable latency before the currents in the lacrimal gland began to increase. A comparison of the changes in outward Cl⁻ currents in seven lacrimal (^) and seven submandibular (•) cells is shown in Fig. 3C. In submandibular cells, the maximum current was attained within 298 ± 31 s of exposure to the hypotonic solution. By contrast, in the lacrimal cells there was no increase in the current for the first 306 ± 24 s in the hypotonic solution, and maximum activation was observed after 1110 ± 53 s (P < 0.05, compared with the submandibular gland by t test). The maximum current densities observed in the two types of cell, however, were not significantly different (lacrimal cell, 34.8 ± 7.9 pA pF⁻¹; submandibular cell, 32.3 ± 4.3 pA pF⁻¹ at a V_m of 60 mV; P > 0.1 by t test).

Figure 3. Cl⁻ current activation by hypotonic extracellular solution in rat submandibular and lacrimal gland cells.

A, a rat submandibular gland acinar cell; B, a rat lacrimal gland acinar cell. Channel activity was evoked by applying 0.4 s voltage steps to −60 and 60 mV from a holding potential of −40 mV. The arrows indicate the start of the superfusion with the hypotonic solution. C, time course of channel activation in submandibular (•, n = 7) and lacrimal (^, n = 7) acinar cells. Channel activity was assessed by measuring the whole-cell current at V_m= 60 mV; these values were then normalised for cell size by dividing by cell capacitance (current density; pA pF⁻¹). All data are means ±s.e.m.

The changes in cell volume caused by exposure to the hypotonic solution were examined in video-imaging experiments. The lacrimal and submandibular gland cells swelled to maximum relative volumes of 1.30 ± 0.03 in 3 min (n = 5) and 1.29 ± 0.05 in 3 min (n = 6), respectively. Thus, it appears unlikely that the differences in the rates of channel activation can be explained by different rates of cell swelling. Volume-sensitive Cl⁻ currents were also induced in lacrimal cells by using a hypertonic pipette solution to cause cell swelling. Activation, however, was observed only after a latency of 438 ± 78 s (n = 3).

Properties of the volume-sensitive currents in submandibular and lacrimal acinar cells

Figure 4 shows current profiles for volume-sensitive currents in lacrimal (A) and submandibular (B) acinar cells. In these experiments the pipette contained the low-Cl⁻ solution, and the I–V relationships exhibited outward rectification in the lacrimal cells (Fig. 4C) and submandibular cells (Fig. 4D). The mean reversal potential (V_rev) for the volume-sensitive currents was −26 ± 4 mV in five lacrimal gland cells and −29 ± 3 mV in five submandibular gland cells. These values were not statistically different (P > 0.1 by t test), and although close to the equilibrium potential for Cl⁻ (E_Cl; -34 mV), they do suggest some permeability to aspartate ions (Table 2). Time-dependent inactivation of the outward currents was observed at more extreme depolarising potentials in both cell types (Fig. 4A and B).

Figure 4. Volume-sensitive Cl⁻ currents in rat exocrine acinar cells.

Profiles for currents activated maximally by cell swelling in lacrimal gland cells (A and C) and submandibular gland cells (B and D). A and B, the currents were evoked by applying 1 s voltage steps from −100 to 100 mV (at 20 mV increments) from a holding potential of −40 mV (A) and −60 mV (B). The pipette was filled with the low-Cl⁻ solution and the dashed lines indicate the zero current level. C and D, I–V relationships for maximum current densities (pA pF⁻¹; means ±s.e.m.) recorded from five lacrimal (C) and five submandibular (D) cells in the presence of the hypotonic bath solution. V_rev was −26 ± 4 and −29 ± 3 mV in the lacrimal and submandibular cells, respectively.

The anionic selectivity of the volume-sensitive conductance in lacrimal cells to Cl⁻, I⁻ and aspartate was examined by changing the anionic composition of the bath solution (Table 1). Figure 5A shows current profiles recorded from a cell in the presence of the different bath solutions, with the corresponding I–V relationships in Fig. 5B. In the control solution (^), only small inward and outward currents could be observed. The volume-sensitive conductance was activated by exposure to the hypotonic solution (•). Outward rectification of the conductances was still observed in these experiments, which used the high-Cl⁻ pipette solution, but it was reduced compared with that observed in Fig. 4. (The ratio of the chord conductances measured at a V_m of 100 and −100 mV (an index of the outward rectification) was 2.1 ± 0.4 in five lacrimal cells. A similar value was also obtained in submandibular cells under these conditions (4.1 ± 1.5; n = 5; P > 0.1 by t test). Current inactivation at depolarising potentials was not observed in either lacrimal or submandibular cells when using the high-Cl⁻ pipette solution (Fig. 5A and Fig. 6A).) Exchanging the bath solution for the iodide bath solution (▴) caused an increase in the size of the outward currents with a leftward shift in V_rev, suggesting that the conductance is more permeable to I⁻ than to Cl⁻. In the presence of the aspartate bath solution (□), the outward currents were reduced with a rightward shift in V_rev, indicating that the conductance is less permeable to aspartate ions. Table 2 gives mean values for V_rev measured in three cells in the presence of the iodide and aspartate bath solutions. The anionic permeability sequence of the conductance was P_I > P_Cl > P_aspartate.

Figure 5. Anionic selectivity of the volume-sensitive Cl⁻ currents in a rat lacrimal gland acinar cell.

A, current profiles for channels in isotonic solution (Control), and when maximally activated by cell swelling (Hypotonic). Swelling-activated currents are also shown in the presence of the iodide and aspartate bath solutions. The pipette contained the high-Cl⁻ solution. Currents were evoked by applying 0.5 s voltage steps from −100 to 100 mV (at 20 mV increments) from a holding potential of −40 mV. The dashed line indicates the zero current level. B, I–V relationships for the current profiles in A. ^, control bath solution; •, hypotonic bath solution; ▴, iodide bath solution; □, aspartate bath solution. Similar data were obtained in three experiments (see Table 2).

Figure 6. Anionic selectivity of the volume-sensitive Cl⁻ currents in a rat submandibular gland acinar cell.

A, current profiles for channels in isotonic solution (Control) and when maximally activated by cell swelling (Hypotonic). Currents are also shown in the presence of the iodide and aspartate bath solutions, and on return to the control bath solution. The pipette contained the high-Cl⁻ solution. Currents were evoked by applying 0.5 s voltage steps from −100 to 100 mV (at 20 mV increments) from a holding potential of −40 mV. The dashed line indicates the zero current level. B, I–V relationships for the current profiles in A. ^, control bath solution; •, hypotonic bath solution; ▴, iodide bath solution; □, aspartate bath solution. Similar data were obtained in five experiments (see Table 2).

Similar experiments were performed using the high-Cl⁻ pipette solution to study the anionic selectivity of the volume-sensitive conductance in the submandibular gland acinar cells (Fig. 6). The size of the outward currents was increased and decreased, respectively, in the presence of the iodide and aspartate bath solutions (Fig. 6A). The I–V relationships (Fig. 6B) showed a slight shift to the left with iodide, and a marked shift to the right with aspartate. Table 2 summarises data from five experiments, and shows that the anionic permeability sequence of the conductance was P_I > P_Cl > P_aspartate. Figure 6A also shows that current activation by the hypotonic solution could be reversed when submandibular cells were returned to the control bath solution. The currents in lacrimal gland cells could also be inactivated by this manoeuvre (data not shown).

Table 2.

Anion selectivity of the volume-sensitive conductances in rat lacrimal gland and submandibular gland acinar cells

Anion	Vrev(mV)	PA:PCl	n
Lacrimal gland
Aspartate−(pipette)	−26±4	0.2±0.1	5
Aspartate− (bath)	37±1	0.2±0.1	3
I− (bath)	−22±2	3.7±0.3	3
Submandibular gland
Aspartate−(pipette)	−29±3	0.2±0.1	5
Aspartate− (bath)	39±6	0.2±0.1	5
I− (bath)	−9±1	2.1±0.1	5

The conductances were activated by exposing cells to the hypotonic bath solution. V_rev was estimated from the I–V relationships recorded in solutions with different anionic compositions (see Table 1): aspartate⁻ (pipette), low-Cl⁻ pipette, hypotonic bath; aspartate⁻ (bath):high-Cl⁻ pipette, aspartate bath; I⁻ (bath):high-Cl⁻ pipette, iodide bath. The permeability of aspartate⁻ or I⁻ relative to Cl⁻ (P_A:P_Cl) was calculatedusing the Goldman-Hodgkin-Katz equation.

The effects of known inhibitors of volume-sensitive channels (NPPB, DIDS and tamoxifen) were examined in the lacrimal and submandibular cells. Figure 7A shows the outward currents measured at a V_m of 100 mV in lacrimal gland cells in the absence and presence of 100 μm NPPB, 100 μm DIDS and 10 μm tamoxifen; the currents were inhibited by each drug. The outward currents in submandibular acinar cells were also inhibited by 100 μm NPPB, 100 μm DIDS and 10 μm tamoxifen (Fig. 7B). Figure 7C summarises the effects of the channel blockers on the outward currents in lacrimal cells (□) and submandibular cells (▪). This shows that there were no significant differences between the inhibition by NPPB, DIDS and tamoxifen in the lacrimal cells and submandibular cells (P > 0.1 by ANOVA).

Figure 7. Cl⁻ channel blockers inhibit the volume-sensitive Cl⁻ channel in rat lacrimal and submandibular acinar cells.

A and B, outward currents were evoked in lacrimal (A) and submandibular (B) acinar cells by voltage steps (1 s) from −40 to 100 mV, and are shown under control conditions and in the presence of 100 μm NPPB, 100 μm DIDS or 10 μm tamoxifen. C, summary of channel inhibition by 100 μm NPPB, 100 μm DIDS and 10 μm tamoxifen in lacrimal (□) and submandibular (▪) cells. The number of cells is given in parentheses.

DISCUSSION

ClC-3 expression in rat lacrimal gland and submandibular salivary gland

ClC-3 was first identified in rat brain by homology cloning methods (Kawaski et al. 1994). Kawaski et al. (1994) also showed that mRNA encoding ClC-3 was widely expressed throughout the mammalian body, e.g. brain, lung, kidney and adrenal gland. This view was echoed by Horowitz and colleagues who proposed that ClC-3 is a volume-sensitive anion channel in cardiac muscle (Duan et al. 1997), and have since shown that ClC-3 mRNA is also expressed in smooth muscle (Dick et al. 1998; Yamazaki et al. 1998). Co-expression of ClC-3 mRNA and volume-sensitive anion channels has also been determined in pancreas from neonatal rats (Schmid et al. 1998), brain endothelial cells (von Wiekersthal et al. 1999) and in non-pigmented ciliary epithelial cells (Coca-Prados et al. 1996; Wang et al. 2000). Wang et al. (2000) have also demonstrated ClC-3 protein expression in ciliary epithelial cells using a ClC-3 antibody.

In the present study, RT-PCR was used to investigate the expression of mRNA for ClC-3 in rat salivary gland and lacrimal gland. Two different protocols were employed: (1) PCR using degenerate primers for ClC channels, followed by a further round of PCR using nested primers specific for ClC-3, and (2) PCR using specific ClC-3 primers. ClC-3-derived products were observed, and characterised, in submandibular gland and the positive control tissues (brain and RINm5F cells) using both protocols. Neither method, however, provided evidence of ClC-3 expression in the lacrimal gland. It is possible that mRNA is expressed at such low levels in lacrimal tissue that it cannot be detected by RT-PCR. This explanation seems unlikely, however, given that ClC-3 products were not observed after two rounds of PCR, and that expression of ClC-3 was observed in submandibular tissue (which has a very similar morphology and function to the lacrimal gland). Furthermore, mRNAs for β-actin, carbonic anhydrase II and ClC-2 were all identified in the same preparations of mRNA from lacrimal gland. Thus, it seems likely that mRNA for ClC-3 is not expressed in rat lacrimal gland.

The expression of ClC-3 protein was examined by Western analysis using a commercially available antibody. A protein with a molecular mass of approximately 80 kDa was observed in the brain and submandibular gland. This is very similar to the size of the protein identified using the same antibody in liver (Shimada et al. 2000), and is close to the predicted size of the ClC-3 protein (84.5 kDa; Kawasaki et al. 1994). The ClC-3 antibody did not interact, however, with a protein in the lacrimal gland. These data support the conclusion that ClC-3 is either not expressed in lacrimal gland, or is expressed at extremely low concentrations. They are therefore consistent with our working hypothesis that, if ClC-3 is a volume-sensitive anion channel, it will not be expressed in lacrimal gland acinar cells (in which volume-sensitive anion channels have not previously been identified; Kotera & Brown, 1993). This hypothesis was further tested in electrophysiological experiments designed specifically to investigate the expression of volume-sensitive anion channels in the lacrimal gland cells.

Volume-sensitive anion channels in acinar cells

Previous studies of the effects of cell swelling on Cl⁻ channel activity in rat lacrimal acinar cells showed that Ca²⁺-activated Cl⁻ channels are activated on exposure to a hypotonic bath solution (Kotera & Brown, 1993). In the present study, whole-cell patch clamp experiments were performed using a pipette solution containing 5 mm BAPTA to prevent the activation of these channels. Under these conditions, a second Cl⁻ conductance was activated by prolonged exposure to the hypotonic bath solution (i.e. in excess of 5 min). The conductance activated in the lacrimal cells was anion selective (P_I > P_Cl > P_aspartate), exhibited an outward-rectifying I–V relationship, and was blocked by DIDS, NPPB and tamoxifen. These properties are similar to those of the conductances activated by hypotonic solutions in rat submandibular gland acinar, and other acinar, cells, e.g. rat parotid acinar (Arreola et al. 1995) and neonatal rat pancreas (Schmid et al. 1998). They are also similar to those of volume-sensitive anion conductances identified in many other cell types (Nilius et al. 1996; Strange et al. 1996; Okada, 1997).

There was, however, one major difference between the currents observed in the salivary and lacrimal acinar cells. This was the difference between the rates at which the channels activated on exposure to the hypotonic solution. In the submandibular gland cells the channel started to activate almost immediately after the addition of the hypotonic solution and the maximum current was observed after 5 min. By contrast, lacrimal gland cells showed a delay of about 5 min before any increase in current, and the maximum current was not observed until 18 min. This difference between the rates of channel activation could not be explained by a difference in the rates of cell swelling, which were similar in both cell types. This suggests that the difference in the rates of channel activation is a function of either the channel proteins involved, or the pathways responsible for their activation.

The delay in activation may help explain why this volume-sensitive anion conductance was not observed in a previous study of lacrimal cells (Kotera & Brown, 1993). In these earlier experiments, changes in the inward Cl⁻ currents (at V_m= −80 mV) were monitored for up to 7 min in the presence of a hypotonic solution. Over such a relatively short period, any change in the inward currents carried by the volume-sensitive channels would be small, compared with the large increase in currents carried by Ca²⁺-activated Cl⁻ channels. The slow activation of the volume-sensitive anion channels in lacrimal cells also implies that these channels probably make only a minor contribution to regulatory volume decrease (RVD) in these cells, since Park et al. (1994) found that RVD commences within 3 or 4 min of exposure to hypotonic solutions, i.e. before the volume-sensitive anion channel is activated.

The role of ClC-3 as a volume-sensitive Cl⁻ channel

The apparent latency and length of time to reach maximum activation of volume-sensitive anion channels in the lacrimal cells are much greater than those in other cell types including those known to express ClC-3, e.g. smooth muscle cells (Dick et al. 1998; Yamazaki et al. 1998) and non-pigmented ciliary epithelial cells (Wang et al. 2000). A similar rate of activation, however, was observed in ciliary epithelial cells in which ClC-3 expression was reduced using antisense oligonucleotides (Wang et al. 2000). The lack of ClC-3 expression in both lacrimal and the antisense-treated ciliary epithelial cells therefore appears to be a significant factor in retarding the activation of the volume-sensitive anion channels.

There are at least three possible explanations for our data: (1) ClC-3 is the volume-sensitive anion channel in lacrimal cells, which is expressed at very low concentrations; (2) ClC-3 is not the volume-sensitive channel, but is a regulator of volume-sensitive channels; and (3) ClC-3 is a fast-activating volume-sensitive anion channel, but other volume-sensitive channels which activate more slowly are expressed in some cells. The remainder of this discussion will briefly review the evidence in support and against these different hypotheses.

ClC-3 is the volume-sensitive Cl⁻ channel in lacrimal cells

This hypothesis is based largely on the fact that it is difficult to state definitively that ClC-3 is not expressed in lacrimal cells. Significant expression of ClC-3 is unlikely, however, since it could not be detected in lacrimal cells by two different methods. Clearly, if ClC-3 is expressed in lacrimal cells the amount of channel protein present must be extremely small. One would therefore predict that the maximum conductance of the volume-sensitive currents (carried by ClC-3) would be small, but that the time course of activation would be similar to that observed in other cells. In the present study, however, the exact opposite was observed, i.e. the volume-sensitive conductance was of a similar magnitude, but the rate of channel activation was greatly reduced, in lacrimal compared with submandibular acinar cells. Consequently, while we cannot absolutely exclude very low levels of ClC-3 expression in lacrimal cells, we feel it is highly unlikely that ClC-3 makes a major contribution to the observed volume-sensitive Cl⁻ conductance.

ClC-3 is not a volume-sensitive channel, but is a regulator of volume-sensitive channels

The possible expression of regulatory proteins for volume-sensitive anion channels has been discussed by others (Valverde, 1999). Indeed, other candidate proteins for volume-sensitive anion channels may in fact be regulatory proteins, e.g. p-glycoprotein and pI_Cln (Okada, 1997). In the present study, the rate of activation of the volume-sensitive conductance was greatly reduced in lacrimal cells as opposed to submandibular cells. Furthermore, the properties of the anion channel activated by prolonged cell swelling in lacrimal cells were generally quite similar (except for the rate of activation) to those of volume-sensitive channels in other cell types. Thus, it could be argued that the same channel is expressed in both lacrimal and submandibular cells, and that the difference in the rate of activation is due to the presence of ClC-3 (the channel regulator) in the submandibular cells. However, this must remain a tentative conclusion until the molecular identity of the volume-sensitive channels expressed in the acinar cells is firmly established. While the biophysical properties of the volume-sensitive conductances are not significantly different in the two cell types, this infers only that the channels are similar rather than identical.

Standing against the idea that ClC-3 is a regulator, there is a considerable weight of evidence that ClC-3 is a channel protein. First, ClC-3 is a member of a family of proteins which are known to be Cl⁻ channels. Second, the expression of ClC-3 in mammalian cells produces an outward-rectifying Cl⁻ conductance with properties which are similar to those of volume-sensitive anion channels (Duan et al. 1997; Valverde, 1999; Shimada et al. 2000). Third, the properties of the expressed conductance are altered by changing the amino acid sequence of the ClC-3 protein (Duan et al. 1997; Duan et al. 1999). Finally, there is evidence that antisense oligonucleotides to ClC-3 reduce protein expression and the magnitude of the volume-sensitive anion conductance in ciliary epithelial cells (Wang et al. 2000). Overall therefore it seems unlikely that ClC-3 acts as a channel regulator, rather than as a channel per se, in the submandibular acinar cells.

ClC-3 is a ‘fast-activating’ volume-sensitive Cl⁻ channel

This hypothesis presupposes that ClC-3 is one of a number of different volume-sensitive anion channels which may contribute to volume-sensitive Cl⁻ conductances. This was the conclusion reached by Wang et al. (2000) as a result of their ClC-3 antisense experiments. Although the antisense treatment reduced the magnitude of the volume-sensitive Cl⁻ currents in these cells, there was a residual, slow-activating, volume-sensitive Cl⁻ conductance. Wang et al. (2000) found that the pharmacological properties of the residual conductance were slightly different to those of the conductance in untreated cells (Wang et al. 2000). In the present study, significant differences were not observed in the halide selectivity sequence and pharmacology of the channels in the lacrimal cells and submandibular cells. Nonetheless, the significant difference in the rate of channel activation could reflect a difference in the properties of the channel involved. In this view, the fast-activating conductance in submandibular cells would represent current carried by ClC-3, while the slow-activating conductance seen in lacrimal cells would be carried by a different, and as yet unidentified, protein. The lack of ClC-3 in lacrimal gland cells may make them a useful system in which to examine the expression of other putative volume-sensitive anion channel proteins. The possibility that proteins other than ClC-3 may contribute to volume-sensitive currents has recently been reinforced by Stobrawa et al. (2001), who identified volume-sensitive Cl⁻ currents in hepatocytes and pancreatic acinar cells isolated from mice in which the ClC-3 gene was disrupted. Stobrawa et al. (2001) did not, however, comment on the rate of channel activation in these cells.

In conclusion, lacrimal acinar cells do not express ClC-3, but possess slow-activating, volume-sensitive anion channels. It is suggested that ClC-3 may be one of a number of volume-sensitive anion channels, but that the expression of ClC-3 is not an absolute requirement for the activity of volume-sensitive anion channels in rat lacrimal gland acinar cells.