ClC-7 Cl− channels expressed in osteoclasts are important for bone resorption since it has been shown that disruption of the ClCN7 gene in mice leads to severe osteopetrosis. We have previously reported that Cl− currents recorded from mouse osteoclasts resemble those of ClC-3 Cl− channels. The aim of the present study was to determine the expression of ClC-3 channels in mouse osteoclasts and their functional role during bone resorption. We detected transcripts for both ClC-7 and ClC-3 channels in mouse osteoclasts by RT-PCR. The expression of ClC-3 was confirmed by immunocytochemical staining. Mouse osteoclasts lacking ClC-3 Cl− channels (ClC-3−/− osteoclasts) derived from ClCN3 gene-deficient mice (ClC-3−/−) showed lower bone resorption activity compared with ClC-3+/+ osteoclasts derived from wild-type mice (ClC-3+/+). Treatment of ClC-3+/+ osteoclasts with small interfering RNA (siRNA) against ClC-3 also significantly reduced bone resorption activity. Electrophysiological properties of basal and hypotonicity-induced Cl− currents in ClC-3−/− osteoclasts did not differ significantly from those in ClC-3+/+ osteoclasts. Using immunocytochemistry, ClC-3 was colocalized with lysosome-associated membrane protein 2. Using pH-sensitive dyes, organelle acidification activity in ClC-3−/− osteoclasts was weaker than in ClC-3+/+ osteoclasts. Treatment of ClC-3+/+ osteoclasts with siRNA against ClC-3 also reduced the organelle acidification activity. In conclusion, ClC-3 Cl− channels are expressed in intracellular organelles of mouse osteoclasts and contribute to osteoclastic bone resorption in vitro through organelle acidification.
- knockout mice
- Cl− current
- lysosome-associated membrane protein
- pH-sensitive dye staining
osteoclasts are multinucleated cells that play a critical role in bone resorption. Following adhesion to the bone surface, osteoclasts form ruffled borders and clear zones with a highly polarized cytoplasmic organization that is a prerequisite for organizing an enclosed microenvironment. For the digestion of inorganic and organic components of the bone matrix, osteoclasts secrete H+ and lysosomal enzymes such as acid proteases into this enclosed microenvironment through the ruffled border membrane (2). Cl− channels act as a transport pathway for electric neutralization of the charge accumulation associated with H+ secretion. Disruption of ClC-7 channels in mice leads to severe osteopetrosis due to failure of H+ secretion from osteoclasts (15). Recently, ClC-7 Cl− channels expressed in Xenopus laevis oocytes were shown to be activated by extracellular acidity, and these currents exhibited strong outward rectification with slow activation kinetics (3).
Electrophysiological studies using the whole cell patch-clamp technique revealed that mammalian osteoclasts possess Cl− channels on their plasma membranes (11, 14, 17, 18). These Cl− currents are characterized by outward rectification, rapid activation kinetics, blocked by DIDS, stimulated by extracellular hypotonicity, and have an anion permeability of I− > Cl−. These electrophysiological properties resemble those of Cl− currents mediated by the ClC-3 channel, which is thought to underlie swelling-activated Cl− currents (4), rather than those of ClC-7 channels. Additionally, it has been suggested that the ClC-3 channel might be an intracellular Cl− channel, present on intracellular organelles such as endosomes, lysosomes, and synaptic vesicles, that contributes to intracellular acidification of the organelles' interior by providing an electrical shunt path for more efficient H+ pumping by vacuolar-type H+-ATPase (V-ATPase) (16, 21). However, the presence and location of ClC-3 channels in mouse osteoclasts and whether ClC-3 is involved in Cl− transports though plasma membrane and/or intracellular organelle membrane are not known.
Accordingly, we focused on ClC-3 as a possible candidate for Cl− transport during bone resorption. The purpose of this study was to determine the expression and functional role of ClC-3 in mouse osteoclasts. We examined the expression of ClC-3 using RT-PCR and immunocytochemistry and compared Cl− channel activity, acidity of intracellular organelles, and bone resorption activity in ClC-3-deficient (ClC-3−/−) osteoclasts with those of normal wild-type (ClC-3+/+) osteoclasts.
MATERIALS AND METHODS
Heterozygous (ClC-3+/−) mice of the ClCN3 gene were generously provided by Dr. Uchida (Homeostasis Medicine and Nephrology, Graduate School, Tokyo Medical and Dental University) and bred under conditions approved by the Council on Animal Care of Fukuoka Dental College (approval number 04011). We used homozygous ClC-3−/− and wild-type ClC-3+/+ mice for osteoclast cell culture.
Bone marrow cells were obtained from the tibia and femora of 4- to 5-wk-old ClC-3+/+ (Crlj:ICR) and ClC-3−/− (Crlj:ICR;Clcn3tm1) male mice. Primary osteoblastic cells were prepared from the calvaria of newborn wild-type ddy mice. Bone marrow cells (5 × 105 cells/ml) were cocultured with primary osteoblastic cells (1 × 105 cells/ml) in α-MEM (Invitrogen, Grand Island, NY) containing 10% FBS (Invitrogen), 1α,25-dihydroxyvitamin D3 [1α,25(OH)2D3; 10−8 M], PGE2 (10−6 M), and antibiotics (100 U/ml penicillin G and 0.15 mg/ml streptomycin sulfate) in plastic culture dishes (35 mm diameter) coated with collagen gel (Nitta Gelatin, Osaka, Japan). About 5–6 days after the coculture, osteoclast formation had reached a maximum, and all adherent cells including osteoclasts were removed from the culture plates by treatment with 0.2% collagenase (Wako Pure Chemical, Osaka, Japan). The removed cells were recultured onto glass coverslips (5 × 5 mm), plastic dishes (35 mm diameter), or dentine slices with culture medium (α-MEM containing 10% FBS and antibiotics) for several hours. To examine the number of osteoclasts generated from the coculture, we counted the number of tartrate-resistant acid phosphatase (TRAP)-positive multinucleated cells (MNCs; >3 nuclei) after cell fixation. The number of TRAP-positive MNCs derived from ClC-3−/− mice bone marrow cells (12,984 ± 2,104 cells/35-mm-diameter plastic dish; n = 5) was not significantly different from that derived from ClC-3+/+ mice bone marrow cells (14,302 ± 2,547 cells/35-mm-diamter plastic dish, n = 5). In addition, no significant differences could be seen in the percentages of TRAP-positive MNCs with F-actin distinctly organized in so-called “actin-rings” (87 ± 6%, n = 4, and 85 ± 8%, n = 4, in osteoclasts from ClC-3+/+ and ClC-3−/− mice, respectively). To acquire a pure osteoclast preparation, adherent cells on glass coverslips or plastic dishes were incubated with PBS containing 0.02% EDTA for 10 min at 37°C, and osteoblasts and other bone marrow cells were then removed by pipette. More than 95% of the adherent cells were TRAP-positive mononuclear cells and MNCs. In some experiments, osteoclasts were generated from osteoclast precursor cells (12). Briefly, bone marrow cells (4 × 106 cells/ml) derived from ClC-3+/+ mice were cultured with α-MEM containing 10% FBS and macrophage colony-stimulating factor (M-CSF; 50 ng/ml). After 3 days in culture, all nonadherent cells were removed by pipette. Adherent cells were used as osteoclast precursor cells and then further cultured for 1–3 days with α-MEM containing 10% FBS and receptor activator of NF-κB ligand (RANKL; 100 ng/ml) or IL-1β (10 ng/ml).
TRAP and actin microfilament staining.
To identify osteoclasts, cells on glass coverslips, plastic dishes, or dentine slices were fixed with 3.7% paraformaldehyde and then stained for TRAP using the Leukocyte Acid Phosphatase Kit (Sigma Chemical, St. Louis, MO). After being stained for TRAP, cells were incubated with rhodamine-conjugated phalloidin (5 U/ml) to stain for F-actin and inspected under a fluorescence microscope (TMD 300, Nikon, Tokyo, Japan) with the appropriate filter set (excitation: 510- to 540-nm bandpath filter, dichroic mirror: 550 nm, and emission: 590-nm long-pass barrier filter).
Bone resorption assay.
Cell suspensions containing osteoclasts generated from the cocultures were plated onto dentine slices (4 mm diameter) or glass coverslips and cultured for 15 h with culture medium. At the end of this period, cells were fixed with 3.7% paraformaldehyde and stained for TRAP. The number of TRAP-positive MNCs on glass coverslips was counted to give an estimate of the osteoclast number on dentine slices. An average of 430–530 osteoclasts were plated onto each dentine slice (4 mm diameter). Osteoclasts were then removed from the dentine slices by ultrasonication in 0.25 M ammonium hydroxide, and resorption pits were stained with Mayer's hematoxylin solution. Images of the resorption pits were acquired using a digital camera (Coolpix 990, Nikon), and the total resorption pit area was measured by image-analysis software [National Institutes of Health (NIH) Image, version 1.62, Bethesda, MD]. The ratio of resorption pit area per total number of osteoclasts was calculated to assess bone resorption activity.
Glass coverslips with adherent cells were placed in a recording chamber (volume: 1 ml) attached to an inverted microscope (TMD 300, Nikon). Only MNCs having more than three nuclei were used for these experiments. Membrane capacitances of these cells ranged from 45 to 95 pF. At end of the electrophysiological recordings, cells were stained for TRAP to conclusively identify them as osteoclasts. Cl− current was recorded using the whole cell configuration patch-clamp technique as previously described (17). To isolate Cl− current from other whole cell currents, we used the following K+-free extracellular and Cs+-rich patch pipette solutions: extracellular solution (in mM) contained 140 NaCl, 10 glucose, 0.5 MgCl2, 1.25 CaCl2, and 10 HEPES, adjusted to pH 7.3 with Tris; and patch pipette solution contained (in mM) 100 Cs-gluconate, 45 CsCl, 3 MgCl2, 2 ATP-2Na, 0.3 EGTA, and 10 HEPES, adjusted to pH 7.3 with Tris. The osmolarity of all solutions was measured with a freezing-point depression osmometer (Osmometer Automatic, Knauer, Berlin, Germany) and adjusted to 290–310 mosM with mannitol. In some experiments, a hypotonic solution (210 mosM) was made by reducing the NaCl concentration in the K+-free extracellular solution to 100 mM. Low-NaCl isotonic solution (290 mosM) was prepared by the addition of 80 mM mannitol to the hypotonic solution as a control. All electrophysiological experiments were performed at 26–27°C.
Small interfering RNA experiments.
The 21-nucleotide small interfering RNA (siRNA) duplexes with two overhang dT nucleotides at the 3′-end targeted to the mouse ClC-3 channel (GenBank Accession No. AF029347) were designed and synthesized by B-Bridge (Sunnyvale, CA). The siRNA sequence for ClC-3 was as follows: sense siRNA 5′-CGAGAGAAGUGUAAGGACATT-3′ and antisense siRNA, 5′-UGUCCUUACACUUCUCUCGTT-3′. The nonsense siRNA sequence was as follows: sense siRNA 5′-AUCCGCGCGAUAGUACGU ATT-3′ and antisense siRNA: 5′-UACGUACUAUCGCGCGGAUTT-3′. siRNAs were transfected with Lipofectamine 2000 (Invitrogen) according to the manufacturer's protocol. Prior to the bone resorption assay and staining of acidic organelles, siRNAs were added to the coculture with primary osteoblasts and bone marrow cells for at least 2 days during the total incubation period of 6–7 days. We confirmed that incubation with Lipofectamine 2000, with or without siRNA, did not affect osteoclast formation (data not shown).
For semiquantitative RT-PCR, total RNA from osteoclast precursor cells or osteoclasts prepared with TRIzol (Invitrogen) was amplified by Superscript II and Taq polymerase (Invitrogen). To examine the mRNA expression of mouse ClC-2, ClC-3, ClC-4, ClC-5, ClC-6, and ClC-7 channels, RT-PCR was performed using the following gene-specific PCR primers. For ClC-2 (GenBank Accession No. AF097415), the 5′-primer was 5′-GGAAGATTGTCCAGGTGAT-3′ and the 3′-primer was 5′-GCAGACATCCAGAACTTC-3′, with an expected product size of 313 bp. For ClC-3 (GenBank Accession No. AF029347), the 5′-primer was 5′-TGTGTCTCTGGTGGTTATTG-3′ and the 3′-primer was 5′-GGAAGAGATGGAGTATGCTG-3′, with an expected product size of 456 bp. For ClC-4 (GenBank Accession No. Z49916), the 5′-primer was 5′-GGACTTCCACACCATAGACT-3′ and the 3′-primer was 5′-CAGAAGAAGCTCTGACCACT-3′, with an expected product size of 337 bp. For ClC-5 (GenBank Accession No. AF134117), the 5′-primer was 5′-GCACCGAGAGATTACCAATA-3′ and the 3′-primer was 5′-CCTTGACAAGAGATACAGCA-3′, with an expected product size of 383 bp. For ClC-6 (GenBank Accession No. NM_011929), the 5′-primer was 5′-TCCAGGTCACATCAGAAGAT-3′ and the 3′-primer was 5′-GACTCGATCAGGATGACTGT-3′, with an expected product size of 400 bp. For ClC-7 (GenBank Accession No. NM_011930), the 5′-primer was 5′-GTCTCATTCTGCACTGTTCC-3′ and the 3′-primer was 5′-GAGGAAGCACTTGATCTGAG-3′, with an expected product size of 496 bp. For controls, identical amplification procedures were used with GAPDH [GenBank Accession No. NM_008084; 5′-primer: 5′-AACCTGCCAAGTATGATGAC-3′ and 3′-primer: 5′-TACCAGGAAATGAGCTTGAC-3′, with an expected product size of 190 bp] or β-actin [GenBank Accession No. NM_007393; 5′-primer: 5′-TGAGAGGGAAATCGTGCGT-3′ and 3′-primer: 5′-GCTGGAAGGTGGACAGTGAG-3′, with an expected product size of 449 bp]. cDNA was amplified under the following conditions: 1 min of denaturation at 95°C, 1 min of annealing at 53°C, and 1 min of extension at 72°C using 36 cycles. Fluorescence of each PCR product was detected by the use of an image analyzer (Fluoro Image Analyzer FLA-2000F, Fuji Film, Tokyo, Japan). The mRNA signal for each ClC Cl− channel was normalized to the respective GAPDH or β-actin mRNA expression levels (i.e., calculation of relative intensity) using NIH Image software (version 1.67).
Osteoclasts on glass coverslips were fixed with 3.7% formaldehyde for 10 min at 4°C and permeabilized with 0.05% Triton X-100 in PBS for 10 min. Cells were incubated with goat polyclonal anti-mouse ClC-3 antibody (1:150 dilution, Santa Cruz Biotechnology, Santa Cruz, CA) overnight at 4°C after blockade of nonspecific binding with 5% rabbit serum for 30 min at room temperature. Cells treated with the primary antibody were washed with PBS and then incubated with biotinylated rabbit anti-goat IgG secondary antibody (5 μg/ml, Vector Laboratories, Burlingame, CA) for 40 min at room temperature. After being rinsed with PBS, cells were incubated with Alexa fluor 488-conjugated streptavidin (1 μg/ml, Molecular Probes, Eugene, OR) for 40 min at room temperature. Fluorescence was observed using fluorescence microscopy (TMD 300, Nikon).
For double immunostaining of osteoclasts for ClC-3 and lysosome-associated membrane protein (LAMP)-2, cells were incubated with primary antibody mixture containing goat polyclonal anti-mouse ClC-3 antibody (1:150 dilution) and rat monoclonal anti-mouse LAMP-2 antibody (1:200 dilution, Santa Cruz Biotechnology) overnight at 4°C. After cells had been rinsed with PBS, ClC-3 was detected with Alexa fluor 488-conjugated streptavidin (1 μg/ml, Molecular Probes) after incubation of the cells with biotinylated rabbit anti-goat IgG secondary antibody (5 μg/ml, Vector Laboratories) for 40 min at room temperature. LAMP-2 was visualized by incubating the cells with Alexa fluor 568-conjugated goat anti-rat IgG secondary antibody (2 μg/ml, Molecular Probes). Nuclear staining was performed with 4,6-diamidino-2-phenylindole dye (1:2,000 dilution, Dojindo, Kumamoto, Japan). Cells were imaged using a Zeiss LSM 510 laser scanning confocal microscope (Carl Zeiss, Jena, Germany). The excitation beam was produced by 380/488/596-nm lasers and delivered to the specimen via a Zeiss Apochromat objective. Emitted fluorescence was captured using LSM 510 software (Carl Zeiss). Two-dimensional images of the cells, cut horizontally through approximately the middle of the cell, were captured.
Measurements of pH in intracellular organelles with pH-sensitive dyes.
Osteoclasts on coverslips were incubated for 15 min at 37°C with culture medium containing acridine orange [2 μg/ml 3,6-bis(dimethylamine)acridine] in 5% CO2-95% air. In some experiments, osteoclasts were preincubated in a culture medium containing bafilomycin A1 (30 nM) for 1 h in 5% CO2-95% air. After cells had been washed with PBS to remove excess acridine orange, coverslips were allowed to settle in a temperature-controlled bath (volume: 5 ml, at 33–35°C) filled with physiological saline solution (PSS) containing (in mM) 134 NaCl, 6 KCl, 2.5 CaCl2, 0.5 MgCl2, 10 glucose, and 10 HEPES, adjusted to pH 7.3 with Tris. The accumulation of acridine orange into intracellular organelles was visualized using a fluorescence microscope (TMD 300) with an appropriate filter set (excitation: 450- to 490-nm bandpath filter, dichroic mirror: 510 nm, and emission: 520-nm long-pass filter). Fluorescence images were acquired using a cooled charge-coupled device digital camera (VB-7010, KEYENCE, Osaka, Japan). In parallel experiments, LysoSensor yellow/blue DND-160 (a pH-dependent dual-excitation probe, Molecular Probes) was used to quantify organelle pH using fluorescence ratio imaging. Osteoclasts were incubated for 1 h at 37°C with culture medium containing DND-160 (2 μM) in 5% CO2-95% air and washed with PBS. Fluorescence was measured in single MNCs excited with wavelengths of 330 and 380 nm. Emitted fluorescence was deflected to pass through a dichroic mirror (400 nm), and the transmitted light of 460 ± 20 nm was monitored with a photomultiplier (SPEX Industries, Edison, NJ). Back ground fluorescence was subtracted, and the ratio of fluorescence intensities at 330 and 380 nm was used to give an estimate of intracellular pH. At the end of experiments, cells were stained for TRAP activity.
Goat polyclonal anti-mouse ClC-3 antibody, goat polyclonal anti-mouse V-ATPase antibody, rat monoclonal anti-mouse LAMP-2 antibody, and Alexa fluor-conjugated rabbit anti-goat IgG antibodies were obtained from Santa Cruz Biotechnology. Acridine orange and LysoSensor yellow/blue DND-160 were obtained from Molecular Probes. Bafilomycin A1, M-CSF, and RANKL were obtained from Wako Pure Chemical (Osaka, Japan). IL-1β was purchased from R&D Systems (Minneapolis, MN). 5-Nitro-2-(3-phenylpropylamino)-benzoic acid (NPPB) was obtained from Calbiochem-Novabiochem (San Diego, CA). All other chemicals were purchased from Sigma Chemical. DIDS, NPPB, and bafilomycin A1 were dissolved in DMSO (Sigma) and later diluted to final concentrations in either PSS or normal culture medium (α-MEM containing 10% FBS and antibiotics), which resulted in a final concentration of <0.1% DMSO. This concentration of DMSO had no effect on Cl− current recordings or pH-sensitive dye staining.
Data are expressed as means ± SE of the number of cells (n). Statistical differences were analyzed using a t-test, and P values of <0.05 were considered to be significant.
Expression of ClC-3 channels in mouse osteoclasts.
We examined the expression of ClC Cl− channels, including ClC-3 channels, in purified mouse osteoclasts derived from bone marrow cells from wild-type ClC-3+/+ mice using RT-PCR (Fig. 1A). Osteoclasts expressed not only ClC-7 mRNA but also ClC-3 mRNA. We also confirmed the expression of ClC-4, ClC-5, and ClC-6 channels, but not ClC-2 or Ca2+-activated Cl− channels, in these cells (data not shown). We further examined whether the expression of ClC-3 Cl− channels was affected by osteotropic factors such as RANKL or IL-1β during the differentiation from precursor cells to osteoclasts. Expression levels of ClC-3 mRNA were not altered significantly by treatment with RANKL for 3 days (Fig. 1B). Similar results were obtained in three independent experiments. Treatment with IL-1β also induced no significant changes in the expression levels of ClC-3 mRNA (data not shown). These results indicate that ClC-3 mRNA stably expresses during osteoclast differentiation. Immunocytochemical staining for ClC-3 showed that ClC-3 proteins were present in osteoclasts (ClC-3+/+ osteoclasts, Fig. 1C,a, bottom). In contrast, osteoclasts derived from bone marrow cells from ClC-3−/− mice did not stain for ClC-3 (ClC-3−/− osteoclasts, Fig. 1C,b, bottom). The lack of ClC-3 stain in ClC-3−/− osteoclasts also demonstrates the specificity of our ClC-3 staining method.
Contribution of ClC-3 channels in osteoclastic bone resorption.
Next, we compared bone resorption activities of ClC-3+/+ and ClC-3−/− osteoclasts using a pit formation assay. When ClC-3−/− and ClC-3+/+ osteoclasts were cultured on dentine slices for 15 h, they formed resorption pits, as expected (Fig. 2A). However, the bone resorption activity was significantly weaker in ClC-3−/− osteoclasts (3,059 ± 195 μm2/osteoclast, n = 3) compared with ClC-3 +/+ osteoclasts (4,447 ± 286 μm2/osteoclast, n = 3; Fig. 2B). To confirm the contribution of ClC-3 channels in bone resorption activities, we further examined the effect of siRNA against ClC-3 (ClC-3 siRNA). Using semiquantitative RT-PCR analysis, we confirmed that treatment with ClC-3 siRNA decreased the expression of ClC-3 mRNA in ClC-3+/+ osteoclasts compared with nonsense siRNA treatment (Fig. 2C). There were no significant differences between the bone resorption activities in cells treated with nonsense siRNA (4,367 ± 535 μm2/osteoclast, n = 7) and transfection reagent alone (control: 4,227 ± 609 μm2/osteoclast, n = 7; Fig. 2D). In contrast, treatment with ClC-3 siRNA caused a significant reduction in bone resorption activity (3,175 ± 133 μm2/osteoclast, n = 7, P < 0.05 vs. control).
Comparison of Cl− currents recorded from ClC-3+/+ and ClC-3−/− osteoclasts.
When cell membrane potential in a ClC-3+/+ osteoclast was changed by a voltage ramp from −120 to 90 mV, a membrane current with moderate outward rectification could be elicited (Fig. 3A, a, control). Application of the Cl− channel blocker DIDS (100 μM) inhibited this current with a reversal potential of −21.0 ± 2.1 mV (n = 5), very close to the estimated equilibrium potential for Cl− (ECl = −26 mV). Another Cl− channel blocker, NPPB (50 μM), also inhibited this current (data not shown). These results show that Cl− channels are active in ClC-3+/+ osteoclasts under basal conditions. In ClC-3−/− osteoclasts, a membrane current with moderate outward rectification was also observed and was inhibited by DIDS (100 μM) with a reversal potential of −22.8 ± 2.7 mV (n = 4; Fig. 3A,b). This current was also inhibited by NPPB (50 μM; data not shown). There were no significant differences between the amplitudes of DIDS-sensitive basal Cl− currents in ClC-3+/+ and ClC-3−/− osteoclasts (5.0 ± 0.7 pA/pF, n = 5, in ClC-3+/+ osteoclasts vs. 5.1 ± 0.4 pA/pF, n = 4, in ClC-3−/− osteoclasts, P > 0.05; Fig. 3A,c).
It has been suggested that the ClC-3 channel underlies the swelling-activated Cl− current elicited by extracellular hypotonicity (4). Therefore, we compared hypotonicity-induced currents in ClC-3+/+ and ClC-3−/− osteoclasts. Small outwardly rectifying currents were observed in ClC-3+/+ osteoclasts in low-NaCl isotonic solution (Fig. 3B,a). Current amplitudes increased significantly after a change to the hypotonic solution and were inhibited by the subsequent addition of DIDS (100 μM) with a reversal potential (−11.0 ± 3.2 mV, n = 5) near the estimated ECl of −18 mV, suggesting that the hypotonicity-induced currents in ClC-3+/+ osteoclasts were indeed due to Cl− channels. In ClC-3−/− osteoclasts, hypotonicity-induced currents were also observed (Fig. 3B,b) and were inhibited by DIDS (100 μM) with a reversal potential of −12.8 ± 3.6 mV (n = 4). There were no significant differences between the amplitudes of net currents induced by hypotonic challenge in ClC-3+/+ and ClC-3−/− osteoclasts (40.1 ± 3.5 pA/pF, n = 5, in ClC-3+/+ osteoclasts vs. 38.1 ± 3.8 pA/pF, n = 4, in ClC-3−/− osteoclasts, P > 0.05; Fig. 3B,c). Also, no significant differences were found in the mean times of current activation after hypotonic stimulation (5.8 ± 1.5 min, n = 5, in ClC-3+/+ osteoclasts vs. 6.2 ± 0.7 min, n = 4, in ClC-3−/− osteoclasts, P > 0.05) or mean times for maximal current activation after hypotonic stimulation (13.5 ± 2.8 min, n = 5, in ClC-3+/+ osteoclasts vs. 13.0 ± 3.1 min, n = 4, in ClC-3−/− osteoclasts, P > 0.05). These results suggest that ClC-3 channels contribute neither to basal nor hypotonicity-induced membrane Cl− currents in mouse osteoclasts. Therefore, we next examined the expression and role of ClC-3 channels in intracellular compartments.
Colocalization of ClC-3 channels with LAMP-2 in mouse oseteoclasts.
It has been reported that the ClC-3 channel colocalizes with LAMPs (LAMP-1 or LAMP-2) when heterologously expressed in human hepatoma cells (Huh-7 cells) (16). Therefore, we examined the colocalization of ClC-3 channels with LAMPs in intracellular organelles using a confocal laser scanning microscope. Figure 4 shows the double immunostaining of ClC-3+/+ osteoclasts with antibodies against ClC-3 channels and LAMP-2. As shown in the merged image (Fig. 4D), the ClC-3 signal was colocalized with LAMP-2, a marker for late endosomes and lysosomes (5, 13). This colocalization was most prominently seen in the cell periphery. The same observation was made in four additional cells (data not shown). These results show that ClC-3 channels in mouse osteoclasts are localized on LAMP-2-expressing organelles, probably endosomes and/or lysosomes.
Contribution of ClC-3 channels to organelle acidification in mouse osteoclasts.
It has been proposed that the ClC-3 channel contributes to the acidification of intracellular compartments, including synaptic vesicles (21) and endosomes (6). Therefore, we investigated the possible contribution of ClC-3 channels to organelle acidification in mouse osteoclasts. To measure the acidification of organelles, we used the acridine orange accumulation method (Fig. 5), since acridine orange preferentially accumulates into acidic compartments like intracellular organelles and extracellular resorption lacunae (1). With this method, cell nuclei also stain in green, since acridine orange binds to double-stranded DNA, emitting green fluorescence (23, 26). Images of acidic compartments in ClC-3+/+ and ClC-3−/− osteoclasts are shown in Fig. 5A. Both types of osteoclasts possessed acidic compartments emitting orange fluorescence. These orange-fluorescing compartments are from intracellular acidic organelles and are not extracellular in origin, since we found that these areas could be lifted up together with the cell using mechanical manipulating with a patch pipette. The orange fluorescence was stronger at the cell periphery in both ClC-3+/+ and ClC-3−/− osteoclasts; however, it was weaker in ClC-3−/− osteoclasts compared with ClC-3+/+ osteoclasts, suggesting an attenuation of organelle acidification in ClC-3−/− osteoclasts. We further investigated the acidification of organelles in ClC-3+/+ and ClC-3−/− osteoclasts using LysoSensor yellow/blue DND-160 to better quantify organelle pH. The mean fluorescence ratio in ClC-3−/− osteoclasts was significantly higher (1.75 ± 0.07, n = 9) than that in ClC-3+/+ osteoclasts (1.48 ± 0.14, n = 13), indicating lesser acidification of organelles in ClC-3−/− osteoclasts (Fig. 5B).
To further confirm the contribution of ClC-3 channels to organelle acidification, we examined the effect of ClC-3 siRNA on organelle acidity. Acidic organelles emitting orange fluorescence covered the entire cell area in ClC-3+/+ osteoclasts treated with nonsense siRNA (Fig. 5C, left). Treatment with bafilomycin A1, a V-ATPase inhibitor, totally inhibited the acidification of all MNCs and also of neighboring mononuclear cells (Fig. 5C, right). In contrast, treatment with ClC-3 siRNA attenuated the acidification of organelles in only some of the cells (Fig. 5C, middle, as indicated by arrows), possibly due to the low transfection efficacy of siRNA. Following the fluorescence microscopy, acridine orange-treated cells were stained for TRAP. All MNCs shown in Fig. 5C were TRAP positive (data not shown).
We have previously reported that basal Cl− currents in mouse osteoclasts are characterized by outward rectification, blocked by DIDS and NPPB, and have an anion permeability of I− > Cl− (18). These properties are consistent with currents mediated by ClC-3 channels (4, 25). However, in the present study, we found that the basal Cl− currents recorded from ClC-3+/+ and ClC-3−/− osteoclasts are remarkably similar in appearance, including outward rectification, current amplitude, and DIDS sensitivity (Fig. 3A). Furthermore, the ClC-3 channel has been suggested as a molecular candidate for swelling-activated Cl− current activated by extracellular hypotonicity (4). In the present experiments, we found no significant differences between current density or properties of hypotonicity-induced Cl− currents recorded from ClC-3+/+ and ClC-3−/− osteoclasts (Fig. 3B). These electrophysiological data suggest that ClC-3 channels contribute to neither basal nor hypotonicity-induced Cl− currents in mouse osteoclasts.
The ClC-7 channel has been shown to be highly expressed in the ruffled border membrane of osteoclasts and participates in H+ secretion into resorption lacunae (15). Although the electrophysiological properties of ClC-7 channels are less well understood, one report (3) described that ClC-7 channels expressed in X. laevis oocytes are activated under extracellular acidic conditions with a pH between 4 and 6, but not within neutral pH ranges, and that its current-voltage relationship showed strong outward rectification. In contrast, both basal and hypotonicity-induced Cl− currents in ClC-3+/+ and ClC-3−/− osteoclasts were readily observed at pH 7.3 and exhibited moderate outward rectification (Fig. 3). These properties are clearly distinct from those of ClC-7 channels expressed in oocytes. Since mouse osteoclasts express mRNAs for several Cl− channels, we suggest that channels other than the ClC-3 channel may be involved in basal and hypotonicity-induced Cl− currents recorded from ClC-3+/+ and ClC-3−/− osteoclasts. Further studies are needed to identify the exact nature of these Cl− channels in osteoclasts.
Although the cell membrane ClC-3 channel has been suggested to mediate the swelling-activated Cl− current (4), more recent studies (9, 21) have shown that the ClC-3 channel is present in intracellular organelles. LAMPs have been reported to be present in endosome/lysosome membranes (5, 13). A study (16) of heterologous expression based on colocalization with LAMP-1 and LAMP-2 demonstrated that ClC-3 proteins are present in lysosomal membranes. In our present study, we showed that the ClC-3 channel colocalizes with LAMP-2 in mouse osteoclasts (Fig. 4), suggesting the existence of ClC-3 channels in organelle membranes, including endosomes and lysosomes, rather than in the plasma membrane. Stobrawa et al. (21) showed that the ClC-3 channel is expressed on endosomes and synaptic vesicles in neurons, where it contributes to acidification of their interior by Cl− shunting of the interior-positive membrane potential created by V-ATPase. In fact, it has been demonstrated that endosomal acidification and Cl− accumulation are impaired in hepatocytes lacking ClC-3 channels derived from ClC-3−/− mice (6). Similarly, we found that the acidity of organelles was significantly weaker in ClC-3−/− mouse osteoclasts compared with ClC-3+/+ mouse osteoclasts (Fig. 5, A and B). Toyomura et al. (24) demonstrated that V-ATPase is immunochemically colocalized with LAMP-2 in osteoclasts derived from RAW 264.7 cells and can be detected in acidic organelles. This report supports further the notion that ClC-3 channels expressed in osteoclasts provide an electrical shunt pathway to permit organelle acidification by V-ATPase.
It has been reported that bafilomycin A1 and the specific Cl− channel blocker NS-3736 have strong anti-bone resorption action (19, 22). These drugs also prevent the acidification of lysosomes as well as resorption compartments of osteoclasts (19). Since the ruffled border membrane of osteoclasts is likely created by exocytosis of lysosomes (20), the acidity of lysosomes may directly influence the pH in the resorption lacunae. We (10) recently reported that the Cl− channel blocker NPPB inhibited H+ secretion across the plasma membrane in osteoclasts. Therefore, it appears likely that the absence of ClC-3 channels in osteoclasts would also reduce H+ secretion. In addition, acidification of lysosomes in osteoclasts has been suggested to be crucial for the activation of lysosomal enzyme cathepsin K, which is important for the degradation of the bone matrix (7). In the present study, we found that both organelle acidity and bone resorption activity were indeed significantly lower in ClC-3−/− osteoclasts compared with ClC-3+/+ osteoclasts (Figs. 2B and 5B). Treatment of osteoclasts with ClC-3 siRNA also reduced both organelle acidity and bone resorption activity (Figs. 2D and 5C). This excellent correlation between organelle acidity and bone resorption activity indicates that organelle acidification might be required for osteoclastic bone resorption. Consequently, the reduced bone resorption activity in ClC-3−/− osteoclasts may result from an insufficient acidification of organelles attributable to the lack of ClC-3 channels. However, in our experiments, the absence of ClC-3 channels in osteoclasts did not cause a complete abolition of bone resorption activity (Fig. 2, A and B). It has been suggested that organelle acidification is accomplished by several ClC isoforms, including ClC-3 (8). We have shown that ClC mRNAs other than ClC-3 are also expressed in mouse osteoclasts (10). Therefore, bone resorption activity may remain via ClC-7 channels even in cells that have lost ClC-3 channels.
ClC-3−/− mice have been reported to be smaller than their wild-type ClC-3+/+ littermates and display a pattern of progressive degeneration of the retina, hippocampus, and ileal mucosa (27). However, morphological analysis of the bone structure in ClC-3−/− mice was not performed. Therefore, it is presently unclear whether ClC-3−/− mice develop the osteopetrotic phenotype. It would be very interesting to examine to what extent the reduced bone resorption activity by ClC-3 deficiency seen in the in vitro bone resorption assay has in terms of in vivo consequences.
In conclusion, we found that mouse osteoclasts express not only ClC-7 but also ClC-3 channels and that the latter are present on intracellular organelles. Our work shows that ClC-3 deficiency leads to decreased organelle acidification, resulting in reduced bone resorption activity in vitro.
This work was supported by Ministry of Education, Culture, Sports, Science and Technology of Japan Grant-in-Aids for Scientific Research 15591988, 18592053 (to F. Okamoto), and 18592054 (to K. Okabe) and by a Frontier Research Grant.
The authors thank Dr. Andreas Carl for English editing.
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.
- Copyright © 2008 the American Physiological Society