Na+ dependence of extracellular Ca2+-sensing mechanisms leading to activation of an outwardly rectifying Cl− channel in murine osteoclasts

Bone Vol. 31, No. 3 September 2002:374 –380

ORIGINAL ARTICLES

Naⴙ Dependence of Extracellular Ca2ⴙ-sensing Mechanisms Leading to Activation of an Outwardly Rectifying Clⴚ Channel in Murine Osteoclasts K. SAKUTA, H. SAKAI, H. MORI, H. MORIHATA, and M. KUNO Department of Physiology, Osaka City University Graduate School of Medicine, Osaka, Japan

Introduction

An elevation in the extracellular Ca2ⴙ concentration ([Ca2ⴙo) is a key signal for bone remodeling by inhibiting the resorbing activity of osteoclasts. The [Ca2ⴙ]o-sensing responses include a variety of morphological and functional changes, but the underlying mechanisms are yet to be defined. This study was aimed at investigating the [Ca2ⴙ]osensing mechanisms leading to the activation of the Clⴚ channel in murine osteoclasts. A rise in either Ca2ⴙ or Gd3ⴙ activated an outwardly rectifying Clⴚ (ORcl) channel reversibly and dose-dependently, which was characterized by rapid activation kinetics, little inactivation, and blockage by DIDS. The concentration required for a half-maximal response was estimated to be >20 –30 mmol/L for Ca2ⴙ. Intracellular dialysis with an ATP-free pipette solution or application of an actin destabilizer, cytochalasin D, decreased the [Ca2ⴙ]oactivated ORcl current. Substitution of extracellular Naⴙ by an impermeable cation, N-methyl-D-glucamineⴙ, inhibited the [Ca2ⴙ]o-activated ORcl channel, suggesting that the activation depended on extracellular Naⴙ. A blocker for the Naⴙ-Ca2ⴙ exchanger, 2ⴕ4ⴕ-dichlorobenzamil hydrochloride (DCB), inhibited the [Ca2ⴙ]o-activated ORcl channel as well. Although 10 mmol/L Ca2ⴙ activated the ORcl current only slightly at a standard intracellular pH (7.3), decreasing pH by dialyzing cells with an acidic pipette solution (pH 6.6) enhanced the [Ca2ⴙ]o-activated ORcl current. This potentiation by cell acidosis was eliminated by amiloride, a blocker for the Naⴙ-Hⴙ exchanger. Zinc ion (0.1 mmol/L) and a polycation, neomycin (0.2 mmol/L), activated the ORcl current at intracellular pH 6.6, whereas the effects of those cations were negligible at intracellular pH 7.3. These results suggest that [Ca2ⴙ]o-sensing mechanisms, leading to activation of the ORcl channel in murine osteoclasts, are regulated by ATP and actin cytoskeletal organization, and are sensitized greatly by cell acidosis. Contributions of Naⴙ-dependent transporters in this activating process are examined in the context of a possible intermediate signal of cell swelling caused by Naⴙ influx. (Bone 31:374 –380; 2002) © 2002 by Elsevier Science Inc. All rights reserved.

Osteoclasts secrete protons and various enzymes to degrade bone matrix and play crucial roles in bodily Ca2⫹ homeostasis. During the bone resorption cycle, osteoclasts face an elevation of extracellular Ca2⫹ concentration ([Ca2⫹]o). Exposure to high [Ca2⫹]o inhibits bone resorption by direct action on osteoclasts, such as retraction, a decrease in podosomes, de-adhesion, inhibition of releasing enzymes,18,34 and apoptosis.15 Although [Ca2⫹]o-sensing is no doubt crucial in osteoclast function and Ca2⫹ metabolism, the underlying mechanisms remain undefined. A [Ca2⫹]osensing receptor coupled to a GTP-binding protein (CaR) was initially cloned in bovine parathyroid cells.4 Its homologs have been found widely in various types of cells including the osteoclast lineage.5 These CaRs are polycation receptors that can be activated by various divalent, trivalent, and organic polyvalent cations. Northern blot analysis has shown that mature rabbit osteoclasts express the CaR,9 but the pharmacological profile of [Ca2⫹]o-sensing differs distinctly from that found in parathyroid cells, such as low-affinity to [Ca2⫹]o6,26 and heterologous effects of polyvalent cations.35 Therefore, osteoclasts appear to possess [Ca2⫹]o-sensing mechanisms distinct from or in addition to the CaR. Transmembrane Cl⫺ flux is closely related to functional states of osteoclasts24 and loss of Cl⫺ channels leads to osteopetrosis.10 We have previously found that either a rise in [Ca2⫹]o or hypotonic stress activates an outwardly rectifying Cl⫺ (ORcl) channel in in vitro-generated murine osteoclasts,22,27 which have the characteristics of mature osteoclasts.1,29 Although a strong elevation of [Ca2⫹]o (ⱖ20 mmol/L) is needed to activate the ORcl current,27 a minor change in osmolarity has been shown to increase the sensitivity to [Ca2⫹]o.22 The present study was undertaken to clarify the [Ca2⫹]o-sensing mechanisms leading to activation of the ORcl channel in murine osteoclasts. The ORcl channel was activated by di-, tri- and polycations (Ca2⫹, Zn2⫹, Gd3⫹, and neomycin). This activation depended on extracellular Na⫹ and was inhibited by a blocker for the Na⫹/Ca2⫹ exchanger. As activation of the ORcl channel by hypotonic stress was potentiated by cell acidosis and inhibited by cytochalasin D and removal of ATP,22 the effects of these factors on activation by [Ca2⫹]o were also examined. Cell acidosis enhanced the sensitivity to cations probably via an amiloride-sensitive Na⫹/H⫹ exchanger. Intracellular ATP level and actin cytoskeletal organization also modified the [Ca2⫹]o-activated ORcl channel. Consequently, the [Ca2⫹]o-sensing responses of osteoclasts change at different cellular conditions or functional states and Na⫹-depen-

Key Words: Osteoclast; Chloride channel; Calcium-sensing; Na⫹ dependency; Cell acidosis; Cell swelling.

Address for correspondence and reprints: Dr. Miyuki Kuno, Department of Physiology, Osaka City University Graduate School of Medicine, Abenoku, Osaka 545-8585, Japan. E-mail: [email protected] © 2002 by Elsevier Science Inc. All rights reserved.

374

8756-3282/02/$22.00 PII S8756-3282(02)00838-4

Bone Vol. 31, No. 3 September 2002:374 –380

dent antiporters may contribute to this activating process in either physiological or in certain pathological situations. A preliminary report has been published.23 Materials and Methods Cell Culture Osteoclasts were generated from a coculture of bone marrow cells of 5– 8 week-old male mice (C3H/HeN) with a marrow-derived stromal cell line (ST2; Riken Cell Bank, Tsukuba, Japan).26,27,29 ST2 cells were maintained in ␣-minimum essential medium (␣-MEM) with 5% fetal calf serum (FCS). The mice were killed by cervical dislocation. Bone marrow cells obtained from the femurs and tibias were centrifuged at 300g for 7 min at 4°C, and incubated in ␣-MEM supplemented with 10% FCS, streptomycin (0.1 mg/mL), and penicillin (100 U/mL) at 37°C in a 95% air/5% CO2 atmosphere overnight. Nonadherent cells were collected by centrifugation at 300g at 4°C for 7 min and incubated in a phosphate-buffered saline solution containing 0.02% pronase and 1.5 mmol/L ethylene-diamine tetraacetic acid (EDTA) for 15 min at 37°C. The pronase reaction was stopped by heat-inactivated horse serum (0.2 mL/10 mL pronase solution), and the cell suspension was layered on ice-cold horse serum. After 15 min of sedimentation at unit gravity on the ice-cold horse serum, the uppermost part of the layers was collected, transferred on cold horse serum, and then centrifuged at 800g at 4°C for 10 min. The bone marrow cell pellet was suspended in fresh ␣-MEM supplemented with 10% FCS at 1 ⫻ 106 cells/mL and cocultured with ST2 cells (1 ⫻ 105 cells/mL) in the presence of 1␣,25(OH)2D3 (10⫺8 mol/L) and dexamethasone (10⫺7 mol/L) at 37°C in a 95% air/5% CO2 atmosphere. The total medium was changed twice per week. ST2 cells were removed by incubation with 0.1% bacterial collagenase (collagenase S-1)/0.1% bovine serum albumin (BSA) in ␣-MEM for 10 –30 min at 37°C before recording. Identification of Cells After culturing for 5–7 days, osteoclasts were identified with phase-contrast microscopy as multinucleated cells with a unique morphology (a thick cell body surrounded with developed lamellipodia and retraction fibers) and tartrate-resistant acid phosphatase (TRAP) activity.27,29 The planar area of the cell body was estimated from the longest and shortest diameters using cell images displayed on a video monitor (PVM-1454Q, Sony, Japan) via a CCD camera (KY-F55MD, Olympus, Japan). This method was effective in monitoring swelling induced by hypotonic stress,19 although the two-dimensional parameter was only a semiquantitative estimate for changes in cell volume. As the complex shape of osteoclasts impeded the quantification, swelling was judged by cells with round or ovoid cell bodies, which swelled symmetrically in the focal plane. Solutions The standard external solution contained (in millimoles per liter): 150 NaCl, 1 CaCl2, 1 MgCl2, 10 glucose, 0.1% BSA, and 10 HEPES-NaOH (pH 7.3) (290 –310 mOsm/L). To eliminate the IRK current, the K⫹ was omitted from the extracellular media.27 The osmolality of solutions was measured using a freezing-point depression OS osmometer (Fiske, Inc.). The high Ca2⫹ solution (10 –100 mmol/L) was made by reducing the concentration of NaCl to adjust the osmolarity. The Na⫹-free solution was prepared by substituting NaCl by N-methyl-D-glucamine (NMDG) chloride. The standard pipette solution contained (in millimoles per liter): 150 K-glutamate, 3 MgCl2, 1 BAPTA, 1 Na2ATP, and

K. Sakuta et al. Calcium-sensing mechanisms of osteoclasts

375

10 HEPES-KOH (pH 7.3). Acidic pipette solutions (pH 5.5– 6.6) were buffered with 120 mmol/L Mes. In the solutions containing high Mes, K-glutamate was reduced to compensate for the osmolarity. The osmolality of these pipette solutions was maintained to be between 280 –290 mOsm/L, slightly lower than that of the external medium to reduce spontaneous swelling during whole-cell clamp recordings. Electrophysiological Recordings Recordings were made from osteoclasts cultured for 5–14 days.22,27 The borosilicate glass pipettes had a resistance of 5– 8 M⍀. The series resistance compensation (60%– 80%) was conducted to reduce the voltage error. The reference electrode was a Ag-AgCl wire connected to the bath solution through a Ringeragar bridge. The zero current potential before formation of the gigaseal was taken to be 0 mV. The recording chamber was perfused with the medium containing di-, tri-, and polycations at about 0.1 mL/sec (volume of the recording chamber; 2 mL). Whole-cell currents were recorded by an amplifier (Axopatch 200A, Axon Instruments, Inc.) at room temperature (20 –24°C). Currents were filtered at 2 kHz for cell capacitance cancellation of an analog signal displayed on an oscilloscope. As the ORcl currents did not contain frequency components of ⬎0.5 kHz, the signals were digitized at 2 kHz with an analog-to-digital converter and analyzed by a personal computer (Maclab/4S, AD Instruments, Australia). Voltage ramps (0.24 mV/msec from ⫺120 –120 mV) were applied at the holding potential of ⫺60 mV every 10 sec. Leak current was determined from the linear portion of the current-voltage (I-V) relation when either inward or outward current was absent or when the currents were eliminated by the blockers. The outward conductance was obtained from the I-V relation between ⫹80 and ⫹100 mV, after subtraction of the leak current. Within this voltage range, the conductance estimated from the voltage-ramp method was almost identical to that obtained by measuring the peak current amplitude evoked by voltage steps. The activated conductance was obtained by subtracting the data before the stimulation from the maximum evoked conductance. Data were expressed as means ⫾ SEM, and tested using Student’s unpaired t-test. Chemicals Mes and BAPTA were purchased from Dojindo Laboratories (Kumamoto, Japan), collagenase S-1 was from Nitta Gelatin Co. (Osaka, Japan), and 1␣,25(OH)2D3 was from Duphar (Weesp, The Netherlands). All other chemicals were obtained from Sigma Chemical Co. (St. Louis, MO). A condensed stock solution of Na2ATP (500 mmol/L) was prepared in 1 mol/L Tris-Cl, stored in a freezer, and added to the internal medium before use. Cytochalasin D, 2⬘4⬘-dichlorobenzamil hydrochloride (DCB), amiloride, and 4,4⬘-diisothiocyanato-2,2⬘-stilbenesulfonate (DIDS) were dissolved in dimethylsulfoxide (DMSO). The final DMSO concentration was ⬍0.1%, which itself did not affect the membrane currents. Results Divalent and Trivalent Cations Activate an Outwardly Rectifying Cl⫺ (ORcl) Channel A rise in [Ca2⫹]o reversibly activated an outwardly rectifying Cl⫺ (ORcl) current in murine osteoclasts.22,26,27 Figure 1A shows a representative time course of the activation and the current-voltage (I-V) relations before (a), during (b), and after (c) exposure to 40 mmol/L [Ca2⫹]o. A trivalent ion, Gd3⫹, also activated a similar ORcl current reversibly and dose-dependently

376

K. Sakuta et al. Calcium-sensing mechanisms of osteoclasts

Bone Vol. 31, No. 3 September 2002:374 –380

Figure 2. ORcl currents activated by extracellular Gd3⫹. (A) Left: Time course of ORcl conductance activated by 10 mmol/L Gd3⫹ (left). Right: I-V relations for the currents in the presence of 1, 3, and 10 mmol/L Gd3⫹. The left and right are obtained from two different cells. (B) ORcl currents in control (left), during exposure to 10 mmol/L Gd3⫹ (middle), and after addition of 100 ␮mol/L DIDS (right).

Figure 1. Outwardly rectifying Cl⫺ (ORcl) currents activated by extracellular Ca2⫹. (A) Time course of the ORcl conductance in a cell exposed to 40 mmol/L Ca2⫹ (left). Conductance was calculated from I-V relations obtained by voltage ramps applied at ⫺60 mV. The I-V relations at times indicated by a, b, and c are superimposed (right). (B) ORcl currents evoked by a series of 500 msec voltage steps in control (left), during exposure to 40 mmol/L Ca2⫹ solution (middle), and after addition of 100 ␮mol/L DIDS (right). The pipette solution contained K⫹-glutamate. (C) I-V relations obtained by voltage ramp (left) and the current records evoked by voltage steps (right) of the ORcl currents activated by 40 mmol/L Ca2⫹ under a symmetrical Cl⫺ gradient. The voltage steps were same as those in (B). Both inward and outward currents were blocked by 100 ␮mol/L DIDS. The pipette solution contained K-glutamate in (A) and (B) and CsCl in (C).

(Figure 2A). The ORcl currents activated by these cations (Figures 1B and 2B) shared the same features, such as rapid activation, little inactivation, strong outward rectification, and blockage by a Cl⫺ channel blocker, DIDS (100 ␮mol/L), indicating that the same class of ORcl channels underlay both currents. This Cl- conductance exhibited outward rectification even under a symmetrical Cl⫺ gradient, although there were small inward currents at potentials negative to the reversal potential (Figure 1C). Figure 3 shows dose-response relationships for the cationactivated ORcl conductance normalized by the cell capacitance. The concentration required for half-maximal response was estimated to be ⬎20 –30 mmol/L for Ca2⫹. Half-activation concentration was not determined for Gd3⫹ because high concentrations of Gd3⫹ tended to form precipitates and might be only nominal. However, 10 mmol/L Gd3⫹ almost activated the half-maximal conductance evoked by Ca2⫹. Therefore, activation of the ORcl current seemed to be mediated via low-affinity [Ca2⫹]o-sensing mechanisms. Effects of Intracellular ATP and Cytochalasin on [Ca2⫹]o-activated ORcl Current Because the omission of intracellular ATP and the addition of an actin destabilizer inhibited the hypotonically activated ORcl cur-

rent,22 the effects of these treatments on the activation by [Ca2⫹]o were examined. Figure 4 summarizes the ORcl conductance recorded immediately before (filled bar) and at 5 min after (open bar) exposure to 40 mmol/L Ca2⫹. The stimulus was applied at 10 min after the whole-cell configuration was made. With an ATP-omitting pipette solution, the [Ca2⫹]o-activated conductance (42 ⫾ 20 pS/pF, n ⫽ 11) was significantly smaller than that recorded in the presence of ATP (116 ⫾ 25 pS/pF, n ⫽ 21) (p ⬍ 0.1). Intracellular dialysis with the pipette solution containing 0.1 ␮mol/L cytochalasin D, an actin destabilizer, also decreased the high [Ca2⫹]o-activated ORcl current (45 ⫾ 17 pS/pF, n ⫽ 8) (p ⫽ 0.1) (Figure 4). Both removal of ATP and addition of cytochalasin D did not affect the conductance before stimulation (filled bars). Thus, the two stimuli, high [Ca2⫹]o and hypotonic stress, seemed to share common regulatory mechanisms for the ORcl channel at least partially.

Figure 3. Dose-response relationships for the ORcl currents activated by Ca2⫹ and Gd3⫹. ORcl conductance normalized by the cell capacitance was plotted against extracellular concentrations of Ca2⫹ (A) and Gd3⫹ (B) on a semilogarithmic scale. In (B), the external solution contained 1 mmol/L Ca2⫹. The conductance was obtained at 5 min following the onset of the perfusion of the cation-containing medium. Data expressed as mean and SEM.

Bone Vol. 31, No. 3 September 2002:374 –380

K. Sakuta et al. Calcium-sensing mechanisms of osteoclasts

377

Figure 4. Effects of ATP removal, cytochalasin D, and Na⫹ removal on the [Ca2⫹]o-activated ORcl current. The ORcl conductance immediately before (filled bar) and after (open bar) activation by 40 mmol/L Ca2⫹. Data were obtained with ATP-omitting or cytochalasin D (0.1 ␮mol/L)containing pipette solutions in standard extracellular medium or with the standard pipette solution in Na⫹-omitting or DCB (10 ␮mol/L)-containing external media. In the Na⫹-free solution, Na⫹ was entirely replaced by NMDG⫹. Cells were stimulated at 10 min after the whole-cell configuration was made. The [Ca2⫹]o-activated ORcl conductance was measured at 5 min after the stimulation. Data expressed as mean and SEM. *p ⬍ 0.1 and †p ⫽ 0.1 compared with the [Ca2⫹]o-activated ORcl conductance in the control.

Na⫹ Dependency of [Ca2⫹]o-activated ORcl Current The [Ca2⫹]o-activated ORcl current was inhibited when the extracellular Na⫹ was replaced by an impermeable cation, NMDG⫹ (Figure 4) (p ⬍ 0.1), suggesting that Na⫹-dependent mechanisms were involved in the activating process. This inhibition was mimicked by the treatment of cells with a blocker for the Na⫹/Ca2⫹ exchanger, DCB (10 ␮mol/L) (p ⬍ 0.1) (Figure 4). As the sensitivity of the ORcl channel to [Ca2⫹]o is increased by a minor hypotonic stress,22 there is a possibility that osmotic swelling caused by Na⫹ influx via the antiporter might contribute to activation of the ORcl channel. An attempt was made to judge cell swelling from the changes in the planar area of the cell body (see Materials and Methods). The planar area was increased to 142 ⫾ 11% (n ⫽ 6) of control in all cells exhibiting large ORcl currents evoked by Ca2⫹ or Gd3⫹ (82 ⫾ 34 pS/pF, n ⫽ 6), but only to 110 ⫾ 5% (n ⫽ 11) when Na⫹ was omitted or DCB was added, and to 105 ⫾ 1% (n ⫽ 35) in nonstimulated cells. Thus, cells stimulated by cations tended to swell in the presence of Na⫹. The ORcl conductance before the stimulation was altered neither by the removal of Na⫹ nor by the addition of DCB. In intact cells, 40 mmol/L Ca2⫹ increased the planar area to 117 ⫾ 2% of control (n ⫽ 11) in the Na⫹-containing solution, to 101 ⫾ 2% (n ⫽ 10) in the Na⫹-free solution, and to 102 ⫾ 1% (n ⫽ 10) in the presence of DCB. The increment was smaller than that in whole-cell recordings, but the [Ca2⫹]o-induced swelling in intact cells also depended on Na⫹ significantly (p ⬍ 0.001). Cell Acidosis Potentiates [Ca2⫹]o-activated ORcl Current ORcl conductance increased when cells were dialyzed with acidic pipette solutions. The conductances measured at 10 min dialysis at different pHs were 14 ⫾ 3 pS/pF at pH 7.8 (n ⫽ 13), 16 ⫾ 2 pS/pF at pH 7.3 (n ⫽ 41), 30 ⫾ 7 pS/pF at pH 6.6 (n ⫽ 18), and 35 ⫾ 9 pS/pF at pH 5.5 (n ⫽ 11). The value at pH 5.5– 6.6 was significantly larger than that at pH 7.3 (p ⬍ 0.05). The effects of decreasing pH on the channel activation by [Ca2⫹]o were examined in cells in which changes the conductance within the 10 min dialysis were small or negligible. At pH 7.3, stimulation with 10

Figure 5. Cell acidosis enhances the cation-activated ORcl current. Traces (A), (C), and (D) show averaged time courses of the ORcl conductance activated by 10 mmol/L Ca2⫹ (A), 100 ␮mol/L ZnCl2 (C), and 200 ␮mol/L neomycin (D). Data were obtained from cells intracellularly perfused with the pipette solutions of pHs 7.3 (filled circles) and 6.6 (open circles). The abscissa indicates time after the whole-cell configuration was made. (B) ORcl conductance activated by exposure to those cations for 5 min. Data represent mean and SEM. Number of cells examined is given in parentheses. **p ⬍ 0.05; ***p ⬍ 0.001; †p ⫽ 0.12. Traces (E) and (F) show family of currents evoked by voltage steps at around 5 min after addition of either neomycin or ZnCl2 at pH 6.6.

mmol/L Ca2⫹ hardly increased ORcl conductance (Figure 5A, filled circles), but intracellular acidification (pH 6.6) greatly potentiated the response to this low Ca2⫹ (open circles). This remarkable enhancement in responsiveness by lowering pH was observed for other cations, Zn2⫹ and neomycin, both of which hardly activated the current at pH 7.3 (Figure 5C,D). Figure 5B summarizes the ORcl conductance activated by a 5 min application of 10 mmol/L Ca2⫹, 100 ␮mol/L Zn2⫹, and 200 ␮mol/L neomycin at pH 7.3 and 6.6. The currents evoked by Zn2⫹ and neomycin at pH 6.6 shared the same kinetics (Figure 5E,F) and sensitivity to DIDS (data not shown) with the [Ca2⫹]o-activated ORcl current. Amiloride, a blocker for the Na⫹-H⫹ exchanger, decreased the ORcl current activated by 10 mmol/L Ca2⫹ at pH 6.6 (Figure 6A). The current decreased by 31 ⫾ 10% (n ⫽ 4) at 10 min after the addition of 200 ␮mol/L amiloride. Direct action of amiloride on the channel was unlikely because of the slow time course of inhibition. Pretreatment of cells with amiloride for 15– 60 min eliminated the sensitization of the [Ca2⫹]o-activated ORcl current by cell acidosis (Figure 6B). The ORcl conductance activated by 10 mmol/L Ca2⫹ at pH 6.6 (268 ⫾ 30 pS/pF, n ⫽ 5) decreased to 14 ⫾ 5 pS/pF (n ⫽ 7) (p ⬍ 0.001) in the presence of 200 ␮mol/L amiloride (Figure 6C). These observations, together with previous findings that intracellular acidification potentiated the

378

K. Sakuta et al. Calcium-sensing mechanisms of osteoclasts

Figure 6. Inhibitory effects of amiloride on acidosis-induced potentiation of [Ca2⫹]o-activated ORcl current. Traces (A) and (B) show time courses of the ORcl current activated by 10 mmol/L Ca2⫹ at pH 6.6. Addition of amiloride (200 ␮mol/L) decreased the current once activated (A). Ten millimoles per liter (10 mmol/L) Ca2⫹ did not evoke ORcl conductance in a cell pretreated with 200 ␮mol/L amiloride (B). The abscissa indicates the time after the whole-cell configuration was made. (C) ORcl conductance activated by perfusion of 10 mmol/L Ca2⫹ for 5 min with the acidic pipette solution (pH 6.6) in the cells pretreated with or without 200 ␮mol/L amiloride. Data represent mean and SEM. ***p ⬍ 0.001.

hypotonically activated ORcl channel,22 further suggest that Na⫹ influx via the Na⫹-H⫹ exchanger would contribute to activating the channel by [Ca2⫹]o. At pH 6.6, stimulation with 10 mmol/L Ca2⫹ increased the planar area to 138 ⫾ 7% of control (n ⫽ 4), but only to 110 ⫾ 3% (n ⫽ 7) in the presence of amiloride. Thus, amiloride significantly inhibited the potentiation induced by cell acidosis (p ⬍ 0.01). Discussion The present study has shown that the ORcl current of murine osteoclasts is activated by di-, tri-, and polycations (Ca2⫹, Zn2⫹, Gd3⫹, and neomycin). The [Ca2⫹]o-sensing response differs from that induced via a G-protein-coupled CaR in parathyroid cells as follows. First, a strong elevation of Ca2⫹ and Gd3⫹ was needed to activate the ORcl current. Second, neomycin, an agonist for the CaR, activated the ORcl current only at cell acidosis. In addition, although either pertussis toxin (PTX)sensitive or -insensitive G-proteins can be directly activated by the CaR,5 involvement of G-proteins in the [Ca2⫹]o-activated ORcl currents in osteoclasts remains unclear; that is, activation is suppressed by PTX partially, but is not potentiated by a nonhydrolyzable GTP analog, GTP␥S.27 Osteoclasts seem to possess a distinct [Ca2⫹]o-sensing cell machinery or a different isoform of the CaR, which leads to activation of the ORcl channel. Na⫹ Dependence of [Ca2⫹]o-activated ORcl Channel It has been shown that removal of extracellular Na⫹ suppressed the [Ca2⫹]o-activated ORcl channel. Moreover, blockers for the Na⫹-Ca2⫹ exchanger and the Na⫹-H⫹ exchanger inhibited the channel activation, suggesting that these Na⫹-dependent anti-

Bone Vol. 31, No. 3 September 2002:374 –380

porters may contribute to the activating process. Removal of Na⫹ might cause an increase in the intracellular Ca2⫹ level ([Ca2⫹]i) and cause a decrease in the intracellular pH (pHi) by inhibiting the Na⫹-Ca2⫹ exchanger and the Na⫹-H⫹ exchanger in intact cells. However, both [Ca2⫹]i and pHi were unlikely to modulate the channel activity directly. [Ca2⫹]i in the range of ⬍10 nmol/L to 10 ␮mol/L did not alter activities of the ORcl channel of murine osteoclasts,27 although mobilization of [Ca2⫹]i activated a Cl- channel in rabbit osteoclasts6 and a nonselective cation channel in rat osteoclasts.33 Amiloride inhibited the potentiation of the [Ca2⫹]o-activated ORcl channel, whereas pHi was maintained at a low level with an acidic pipette solution. Thus, activation of the Na⫹-dependent antiporters, rather than a rise in [Ca2⫹]i or a decrease in pHi, might contribute to the channel activation. Exposure to high [Ca2⫹]o increases the [Ca2⫹]i to ⱖ200 –300 nmol/L by releasing Ca2⫹ from the internal stores and/or Ca2⫹ influx through the plasma membrane in osteoclasts.3,7,16,18,34 High [Ca2⫹]o also decreases pHi to approximately 6.5, which in turn activates an amiloride-sensitive Na⫹-H⫹ exchanger in human osteoclasts.7 DCB and amiloride are not only relatively specific inhibitors for the Na⫹-Ca2⫹ exchanger and the Na⫹-H⫹ exchanger, but also have nonselective inhibitory effects on ion transporters.20 Therefore, involvement of unidentified Na⫹-dependent mechanisms cannot be excluded. The Na⫹-dependent antiporters induce Na⫹ influx to compensate for either elevated [Ca2⫹]i or decreased pHi. As Na⫹ is an osmolyte, Na⫹ influx would result in cell swellings. Cell acidosis is known to accumulate Na⫹ via actions of the Na⫹-H⫹ exchangers.8,19 In the present study, the planar area of the cell body tended to increase in association with development of the cation-activated ORcl current, whereas the change was small or negligible when Na⫹ was omitted or blockers for the Na⫹dependent antiporters were added. Moreover, there is evidence suggesting a possible role of swelling in controlling the [Ca2⫹]oactivated ORcl channel. First, the [Ca2⫹]o-activated ORcl channel and the hypotonically activated ORcl channel are likely to be the same class of channels. They share common features, including electrophysiological properties, the requirement of intracellular ATP and actin cytoskeletal networks, and potentiation by cell acidosis.22 Second, activation by the two stimuli, an elevated [Ca2⫹]o, and hypotonic stress, is synergetic.22 A small increase in [Ca2⫹]o (ⱕ10 mmol/L) and a minor hypotonicity, which can not activate the ORcl channel alone, produce large currents if they coexist. Similar upregulation of the volume-sensitive Cl⫺ channel by [Ca2⫹]o is confirmed in epithelial cells that possess a low-affinity CaR.28 Third, a rise in [Ca2⫹]i to pCa 6.5 increases sensitivity of the ORcl channel to hypotonic stress.22 In rat osteoclasts, cell swelling induced an increase in [Ca2⫹]i through a nonselective cation channel and inhibition of bone resorption.32 Cell swellings share a part of the cellular conditions accompanying the [Ca2⫹]o-sensing responses, such as cell acidosis and a rise in [Ca2⫹]i,13 which supports the idea that reorganization of the cytoskeleton with swelling could be an intermediate signal transmitted from the [Ca2⫹]o-sensing molecule to the ORcl channel. Our current hypothesis of the mechanisms involved in activation of the ORcl channel by extracellular Ca2⫹ is given in Figure 7. Cell Acidosis Potentiates [Ca2⫹]o-sensing Responses The pHi of osteoclasts is influenced by a culture substrate,14 pHo,21,31 and diverse pHi-regulatory mechanisms, including the HCO3⫺/Cl⫺ exchanger, the voltage-gated H⫹ channel, and the

Bone Vol. 31, No. 3 September 2002:374 –380

Figure 7. Hypothetical mechanisms leading to activation of the ORcl channel in response to a rise in extracellular Ca2⫹. An elevation in [Ca2⫹]o increases intracellular Ca2⫹ ([Ca2⫹]i) and, perhaps, decreases intracellular pH (pHi), which may stimulate Na⫹-dependent transporters, such as the Na⫹-Ca2⫹ and Na⫹-H⫹ exchangers. The resultant Na⫹ influx induces cell swelling, and, in turn, activates the volume-sensitive ORcl channel. Swelling caused by any mechanisms (e.g., hypotonic stress) could activate the ORcl channel synergistically with increases in [Ca2⫹]o. There is a possibility that another unidentified second message might modulate this process.

Na⫹-H⫹ exchanger.21,25 During bone resorption, osteoclasts actively secrete massive amounts of H⫹ into the resorbing pit and are thus exposed to acidic environment. Extracellular acidosis decreases pHi, enhances formation of podosomes,31 potentiates bone resorption,2,11,17 and downregulates the transient rise in [Ca2⫹]i induced by high [Ca2⫹]o. The present study has proposed that cell acidosis sensitizes the [Ca2⫹]o-sensing mechanisms of osteoclasts and that the pH dependence of the [Ca2⫹]oactivated ORcl channel may modify resorbing actions. 2⫹

[Ca

]o-sensing Responses of Osteoclasts

Low sensitivity to [Ca2⫹]o is probably a crucial feature of the [Ca2⫹]o-sensing in osteoclasts, because [Ca2⫹]o in the resorptive pit is estimated to increase to 40 mmol/L.30 Sensitivity to cations is, however, variable among species or the evaluation method, because use of 100 ␮mol/L neomycin, 0.1– 0.2 mmol/L Gd3⫹, and 10 mmol/L Ca2⫹ has been shown to inhibit pit formation in mature rabbit osteoclasts9 and activate a nonselective cation channel in freshly isolated rat osteoclasts.33 These concentrations are lower than those required to activate the ORcl channel in murine osteoclasts. The heterogeneity might be ascribed to multiple [Ca2⫹]o receptors/sensing mechanisms, such as CaRs, coupled to GTP-binding proteins,9 a ryanodine receptor-like molecule,35 and other undetermined molecules. On the other hand, sensitivity to [Ca2⫹]o depends on activation phases of osteoclasts.12 The present study provides evidence that the sensitivity to cations can be altered by various factors modulating cell volume, such as pH environment, actin cytoskeletal organization, actions of Na⫹-dependent transporters, and the level of ATP. Hypotonic stress, which increases the sensitivity of the ORcl channel to [Ca2⫹]o, may not be a physiological stimulus during bone resorption. However, bone resorption cycles seem to be maintained by a series of transmembrane fluxes of osmolytes. For example, secretion of H⫹ via H⫹-ATPase results in accumulation of HCO3⫺, which would lead to Cl⫺ influx via the

K. Sakuta et al. Calcium-sensing mechanisms of osteoclasts

379

HCO3⫺-Cl⫺ antiporter. Na⫹ influx may be activated by a rise in [Ca2⫹]i and cell acidosis during bone resorption, as described earlier. Otherwise, liberation of Ca2⫹ from bone might put osteoclasts in a hypertonic environment, in which Na⫹ influx may be stimulated for regulatory volume increase. Generation of metabolic osmolytes inside cells during vigorous activity may also induce swelling. Therefore, although high concentrations of [Ca2⫹]o were required to activate osteoclasts cultivated in the culture dishes, sensitivity to [Ca2⫹]o might be enhanced during bone resorption. Chloride channels in osteoclasts may have several roles, such as maintenance of electroneutral H⫹ secretion via vacuolar-type H⫹-ATPase; and regulation of membrane potential, pH, and cell volume.24,25 The ORcl channel is postulated to be a volume regulator that could either enhance or reduce swelling depending on the direction of the Cl⫺ flux. Extrusion of Cl⫺ leads to regulatory volume decreases, which might protect cells from cytotoxic swelling. Influx of Cl⫺, on the other hand, may advance swelling. It is presently unclear whether activation of the ORcl channel induces or decreases the inhibitory actions of [Ca2⫹]o on osteoclast-resorbing functions. The cell volume is, however, a crucial second message in cellular events,13 thus suggesting that the ORcl channel may be an important controller in bone remodeling.

Acknowledgments: The authors thank Dr. S. Matsuura, Dr. Y. Watanabe, and Dr. F. Nakamura for encouragement throughout this study. We also thank Dr. H. Amano for helpful advice on culture of osteoclasts, J. Kawawaki for technical assistance, C. H. Kim for preparation of the manuscript, and Y. Hozumi for secretarial assistance. This work was supported by grants from the Assistant Program of Graduate Student Fellowships of Osaka City University, the Hoansha Foundation, and a Grant-in Aid for Scientific Research from The Ministry of Education, Science and Culture, Japan.

References 1. Akatsu, T., Tamura, T., Takahashi, N., Udagawa, N., Tanaka, S., Sasaki, T., Yamaguchi, A., Nagata, N., and Suda, T. Preparation and characterization of a mouse osteoclast-like multinucleated cell population. J Bone Miner Res 7:1297–1306; 1992. 2. Arnett, T. R. and Dempster, D. W. Effect of pH on bone resorption by rat osteoclasts in vitro. Endocrinology 119:119 –124; 1986. 3. Bennett, B. D., Alvarez, U., and Hruska, K. A. Receptor-operated osteoclast calcium sensing. Endocrinology 142:1968 –1974; 2001. 4. Brown, E. M., Gamba, G., Piccardi, D., Lombardi, M., Butters, R., Kifor, O., Sun, A., Hediger, M. A., Lytton, J., and Herbert, S. C. Cloning and characterization of an extracellular Ca2⫹-sensing receptor from bovine parathyroid. Nature Lond 366:575–580; 1993. 5. Brown, E. M. and MacLeod, R. J. Extracellular calcium sensing and extracellular calcium signaling. Physiol Rev 81:239 –297; 2001. 6. Fujita, H., Matsumoto, T., Kawashima, H., Ogata, E., Fujita, T., and Yamashita, N. Activation of Cl⫺ channels by extracellular Ca2⫹ in freshly isolated rabbit osteoclasts. J Cell Physiol 169:217–225; 1996. 7. Grano, M., Faccio, R., Colucci, S., Paniccia, R., Baldini, N., Zallone, A. Z., and Teti, A. Extracellular Ca2⫹ sensing is modulated by pH in human osteoclastlike cells in vitro. Am J Physiol 267:961–968; 1994. 8. Grinstein, S., Goetz, J. D., Furuya, W., Rothstein, A., and Gelfand, E. W. Amiloride-sensitive Na⫹-H⫹ exchange in platelets and leukocytes: Detection by electronic cell sizing. Am J Physiol 247:293–298; 1984. 9. Kameda, T., Mano, H., Yamada, Y., Takai, H., Amizuka, N., Kobori, M., Izumi, N., Kawashima, H., Ozawa, H., Ikeda, K., Kameda, A., Hakeda, Y., and Kumegawa, M. Calcium-sensing receptor in mature osteoclasts, which are bone resorbing cells. Biochem Biophys Res Commun 245:419 –422; 1998. 10. Kormak, U., Kasper, D., Bo¨ sl, M. R., Kaiser, E., Schweizer, M., Schulz, A., Friedrich, W., Delling, G., and Jentsch, T. J. Loss of the ClC-7 chloride channel leads to osteopetrosis in mice and man. Cell 104:205–215; 2001.

380

K. Sakuta et al. Calcium-sensing mechanisms of osteoclasts

11. Kraut, J. A., Mishler, D. R., and Kurokawa, K. Effect of colchicine and calcitonin on calcemic response to metabolic acidosis. Kidney Int 25:608 –612; 1984. 12. Lakkakorpi, P. T., Lehenkari, P. P., Rautiala, T. J., and Va¨ a¨ na¨ nen, H. K. Different calcium sensitivity in osteoclasts on glass and bone and maintenance of cytoskeletal structures on bone in the presence of high extracellular calcium. J Cell Physiol 168:668 –677; 1996. 13. Lang, F., Busch, G. L., Ritter, M., Vo¨ lkl, H., Waldegger, S., Gulbins, E., and Ha¨ ussinger, D. Functional significance of cell volume regulatory mechanisms. Physiol Rev 78:247–306; 1998. 14. Lehenkari, P. P., Laitala-Leinonen, T., Linna, T.-J., and Va¨ a¨ na¨ nen, H. K. The regulation of pHi in osteoclasts is dependent on the culture substrate and on the stage of the resorption cycle. Biochem Biophys Res Commun 235:838 –844; 1997. 15. Lorget, F., Kamel, S., Mentaverri, R., Wattel, A., Naassila, M., Maamer, M., and Brazier, M. High extracellular calcium concentrations directly stimulate osteoclast apoptosis. Biochem Biophys Res Commun 268:899 –903; 2000. 16. Malgaroli, A., Meldolesi, J., Zallone, A. Z., and Teti, A. Control of cytosolic free calcium in rat and chicken osteoclasts: The role of extracellular calcium and calcitonin. J Biol Chem 264:14342–14347; 1989. 17. Meghji, S., Morrison, M. S., Henderson, B., and Arnett, T. R. pH dependence of bone resorption: Mouse calvarial osteoclasts are activated by acidosis. Am J Physiol 280:E112–E119; 2001. 18. Miyauchi, A., Hruska, K. A., Greenfield, E. M., Duncan, R., Alvarez, J., Barattolo, R., Colucci, S., Zambonin-Zallone, A., and Teitelbaum, S. L. Osteoclast cytosolic calcium, regulated by voltage-gated calcium channels and extracellular calcium, controls podosome assembly and bone resorption. J Cell Biol 111:2543–2552; 1990. 19. Morihata, H., Nakamura, F., Tsutada, T., and Kuno, M. Potentiation of a voltage-gated proton current in acidosis-induced swelling of rat microglia. J Neurosci 20:7220 –7227; 2000. 20. Murata, Y., Harada, K., Nakajima, F., Maruo, J., and Morita, T. Non-selective effects of amiloride and its analogues on ion transport systems and their cytotoxicities in cardiac myocytes. Jpn J Physiol 68:279 –285; 1995. 21. Nordstro¨ m, T., Shrode, L. D., Rotstein, O. D., Romanek, R., Goto, T., Heersche, J. N. M., Manolson, M. F., Brisseau, G. F., and Grinstein, S. Chronic extracellular acidosis induces plasmalemmal vacuolar type H⫹ ATPase activity in osteoclasts. J Biol Chem 272:6354 –6360; 1997. 22. Sakai, H., Nakamura, F., and Kuno, M. Synergetic activation of outwardly rectifying Cl⫺ currents by hypotonic stress and external Ca2⫹ in murine osteoclasts. J Physiol Lond 515:157–168; 1999. 23. Sakuta, K., Sakai, H., Mori, H., Nakamura, F., and Kuno, M. Activation mechanisms of outwardly-rectifying Cl⫺ currents by external calcium in murine osteoclasts. Jpn J Physiol 50:S108; 2000 Suppl.

Bone Vol. 31, No. 3 September 2002:374 –380 24. Schlesinger, P. H., Blair, H. C., Teitelbaum, S. L., and Edwards, J. C. Characterization of the osteoclast ruffled border chloride channel and its role in bone resorption. J Biol Chem 272:18636 –18643; 1997. 25. Schlesinger, P. H., Mattsson, J. P., and Blair, H. C. Osteoclastic acid transport: Mechanism and implications for physiological and pharmacological regulation. Miner Electrol Metab 20:31–39; 1994. 26. Shibata, T., Sakai, H., and Nakamura, F. Membrane currents of murine osteoclasts generated from bone marrow/stromal cell co-culture. Osaka City Med J 42:93–107; 1996. 27. Shibata, T., Sakai, H., Nakamura, F, Shioi, A., and Kuno, M. Differential effect of high extracellular Ca2⫹ on K⫹ and Cl⫺ conductances in murine osteoclasts. J Membr Biol 158:59 –67; 1997. 28. Shimizu, T., Morishima, S., and Okada, Y. Ca2⫹-sensing receptor-mediated regulation of volume-sensitive Cl⫺ channels in human epithelial cells. J Physiol Lond 528:457–472; 2000. 29. Shioi, A., Ross, F. P., and Teitelbaum, S. L. Enrichment of generated murine osteoclasts. Calcif Tissue Int 55:387–394; 1994. 30. Silver, I. A., Murrills, R. J., and Etherington, D. J. Microelectrode studies on the acid microenvironment beneath adherent macrophages and osteoclasts. Exp Cell Res 175:266 –276; 1988. 31. Teti, A., Blair, H. C., Schlesinger, P., Zambonin-Zallone, A., Kahn, A. J., Teitelbaum, S. L., and Hruska, K. A. Extracellular protons acidify osteoclasts, reduce cytosolic calcium, and promote expression of cell-matrix attachment structures. J Clin Invest 84:773–780; 1989. 32. Tsuzuki, T., Okabe, K., Kajiya, H., and Habu, T. Osmotic membrane stretch increases cytosolic Ca2⫹ and inhibits bone resorption activity in rat osteoclasts. Jpn J Physiol 50:60 –76; 2000. 33. Yoshida, N., Sato, T., Kobayashi, K., and Okada, Y. High extracellular Ca2⫹ and Ca2⫹-sensing receptor agonists activate nonselective cation conductance in freshly isolated rat osteoclasts. Bone 22:495–501; 1998. 34. Zaidi, M., Alam, A. S. M. T., Huang, C. L.-H., Pazianas, M., Bax, C. M. R., Bax, B. E., Moonga, B. S., Bevis, P. J. R., and Shankar, V. S. Extracellular Ca2⫹ sensing by the osteoclast. Cell Calcium 14:271–277; 1993. 35. Zaidi, M., Shankar, V. S., Tunwell, R., Adebanjo, O. A., Mackrill, J., Pazianas, M., O’Connell, D., Simon, B. J., Rifkin, B. R., Venkitaraman, A. R., Huang, C. L.-H., and Lai, F. A. A ryanodine receptor-like molecule expressed in the osteoclast plasma membrane functions in extracellular Ca2⫹ sensing. J Clin Invest 96:1582–1590; 1995.

Date Received: July 12, 2001 Date Revised: April 22, 2002 Date Accepted: April 30, 2002