help button home button Endocrine Society Endocrinology
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
Endocrinology, doi:10.1210/en.2006-0547
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow All Versions of this Article:
148/2/831    most recent
Author Manuscript (PDF)
Right arrow Purchase Article
Right arrow View Shopping Cart
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow Request Copyright Permission
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Bouyer, P.
Right arrow Articles by Insogna, K. L.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Bouyer, P.
Right arrow Articles by Insogna, K. L.
Endocrinology Vol. 148, No. 2 831-840
Copyright © 2007 by The Endocrine Society

Colony-Stimulating Factor-1 Increases Osteoclast Intracellular pH and Promotes Survival via the Electroneutral Na/HCO3 Cotransporter NBCn1

Patrice Bouyer1, Hiroaki Sakai1, Takashi Itokawa, Tsutomu Kawano, Christiaan M. Fulton, Walter F. Boron and Karl L. Insogna

Departments of Cellular and Molecular Physiology (P.B., C.M.F., W.F.B.) and Internal Medicine (H.S., T.I., T.K., K.L.I.), Section of Endocrinology and Metabolism, Yale University School of Medicine, New Haven, Connecticut 06520

Address all correspondence and requests for reprints to: Patrice Bouyer, Department of Cellular and Molecular Physiology, Yale University School of Medicine, 333 Cedar Street, POB 208026, New Haven, Connecticut 06520-8026. E-mail: Patrice.Bouyer{at}Yale.edu.


   Abstract
Top
 Abstract
Introduction
 Materials and Methods
 Results
 Discussion
 References
 
Colony-stimulating factor-1 (CSF-1) promotes the survival of osteoclasts, short-lived cells that resorb bone. Although a rise in intracellular pH (pHi) has been linked to inhibition of apoptosis, the effect of CSF-1 on pHi in osteoclasts has not been reported. The present study shows that, in the absence of CO2/HCO3–, CSF-1 causes little change in osteoclast pHi. In contrast, exposing these cells to CSF-1 in the presence of CO2/HCO3– causes a rapid and sustained cellular alkalinization. The CSF-1-induced rise in pHi is not blocked by 4,4'-diisothiocyanatostilbene-2,2'-disulfonic acid, an inhibitor of HCO3– transporters but is abolished by removing extracellular sodium. This inhibition profile is similar to that of the electroneutral Na/HCO3 cotransporter NBCn1. By RT-PCR, NBCn1 transcripts are present in both osteoclasts and osteoclast-like cells (OCLs), and by immunoblotting, the protein is present in OCLs. Moreover, CSF-1 promotes osteoclast survival in the presence of CO2/HCO3– buffer but not in its absence. Preventing the activation of NBCn1 markedly attenuates the ability of CSF-1 to 1) block activation of caspase-8 and 2) prolong osteoclast survival. Inhibiting caspase-3 or caspase-8 in OCLs prolongs osteoclast survival to the same extent as does CSF-1. This study provides the first evidence that osteoclasts express a CSF-1-regulated Na/HCO3 cotransporter, which may play a role in cell survival.


   Introduction
Top
 Abstract
 Introduction
Materials and Methods
 Results
 Discussion
 References
 
OSTEOCLASTS ARE TERMINALLY differentiated, multinucleated cells that resorb mineralized skeletal tissue. Resorption is a multistep process in which the osteoclast migrates to a site of future resorption, develops a specialized attachment zone against the bone surface and secretes degradative enzymes and protons into the lacuna between the basolateral surface of the cell and the bone. Acid secretion is critical to the ability of osteoclasts to resorb bone (1, 2, 3). A vacuolar-type H+ pump transports H+ from the cytosol into the lacuna (1, 4). Carbonic anhydrase II, which is highly expressed in osteoclasts, plays an important role in acid secretion by replenishing cytosolic H+ (5, 6). This cytosolic enzyme catalyzes the hydration of CO2, yielding H+ and HCO3–. A Cl/HCO3– exchanger on the nonlacunar membrane disposes of this byproduct by exchanging newly formed intracellular HCO3– for extracellular Cl– (7).

After acute intracellular acid loads, the H+ pump, the Na-H exchanger NHE1, and a proton conductance participate in the extrusion of acid and the return of intracellular pH (pHi) toward normal (2, 3, 8, 9, 10).

Colony-stimulating factor-1 (CSF-1) is one of two cytokines, the other being receptor activator of nuclear factor-{kappa}B ligand, that is absolutely required for osteoclast formation. Op/op mice, which lack a functional CSF-1 gene, develop osteopetrosis because of a failure of osteoclastogenesis (11). CSF-1 also appears to play a role in regulating the function of mature osteoclasts by acting as a chemoattractant for osteoclasts and prolonging the survival both of osteoclast precursors and mature osteoclasts (12, 13, 14, 15, 16). For example, Fuller et al. (12) showed that CSF-1 prolonged by 2-fold the survival of freshly isolated rat osteoclasts cultured under serum-free conditions. Indeed, regulating osteoclast apoptosis appears to be an important mechanism for controlling rates of bone resorption. For example, estrogen is a potent stimulus for preosteoclast apoptosis (17). In estrogen-deficient states, prolonged osteoclast survival likely contributes to the observed increase in rates of bone loss.

One mechanism by which CSF-1 prolongs osteoclast survival is by increasing the ubiquitination of the proapoptotic factor Bim (15). Another pathway by which CSF-1 promotes survival is by inhibiting the caspase cascade. In osteoclast precursors, CSF-1 prevents apoptosis by inhibiting the activity of caspase-3 and -9 and by up-regulating Bcl-XL, an inhibitor of procaspase-9 cleavage (16). Consistent with this finding, Okahashi et al. (18) reported that CSF-1 and IL-1{alpha} suppress the activation of caspases in mouse osteoclast-like cells (OCLs). Interestingly, in a cell-free system, an alkaline pH also inhibits the cleavage of procaspase-3 and -9 (19), but whether this inhibition is relevant to osteoclast survival is not known.

In the present study, we investigated the effect of CSF-1 on pHi in freshly isolated mature osteoclasts and in OCLs cultured in vitro. We found that CSF-1 induces a sustained alkalinization in osteoclasts and OCLs that depends on the presence of CO2/HCO3– in the media. Functional and molecular evidence indicate that the alkalinization reflects activation of the electroneutral Na/HCO3 cotransporter NBCn1. Finally, we report that the prosurvival effect of CSF-1 on OCLs requires CO2/HCO3– and appears to be associated with inhibition of caspase-8 and possibly caspase-3.


   Materials and Methods
Top
 Abstract
 Introduction
 Materials and Methods
Results
 Discussion
 References
 
Preparation of freshly isolated osteoclasts and OCLs
Mature osteoclasts were isolated from neonatal rat long bones by mechanical disaggregation, as reported (20). Cells were plated on 12-mm2 coverslips in {alpha}-MEM, 10% fetal bovine serum (FBS), 1% penicillin/streptomycin, 1% L-glutamine, and 20 mM HEPES (pH 7.36). The coverslips were placed in a 5% CO2 incubator at 37 C for 2 h, after which the FBS concentration was reduced to 2% for 2 h before pHi studies.

OCLs were used to prepare total RNA and whole-cell lysates for Western blotting. OCLs were also used for pHi measurements and in cell survival assays. For total RNA preparation, OCLs were generated in vitro by coculturing murine osteoblasts and bone marrow cells on 10-cm tissue culture plates as previously reported (20). Briefly, primary murine osteoblasts were obtained by serial collagenase/dispase digestion of neonatal CD1 mouse calvariae. Osteoblasts were plated at an initial density of 2.5 x 104 cells/cm2 for coculture. In this coculture system, osteoblasts provide key cytokines to promote the differentiation of osteoclast precursors (supplied from bone marrow) into mature osteoclasts. Bone marrow cells were prepared by flushing the marrow from the tibiae and femurs of 7-wk-old CD1 mice. Marrow was spun at 200 x g for 5 min, washed with {alpha}-MEM containing 10% FBS, and cocultured with osteoblasts at an initial density of 1.5 x 105 nucleated cells/cm2. Cocultures were grown in {alpha}-MEM with 10% FBS, 1% penicillin/streptomycin, 1% L-glutamine, and 20 mM HEPES (pH 7.36) containing 10–8 M 1,25-dihydroxyvitamin D3 and 10–6 M prostaglandin E2, with a medium change every other day for 6 d. Contaminating mononuclear cells were removed by treating with 5 mM EDTA for 10 min. Approximately 90% of the cellular material derived from these purified cultures is from OCLs. Purified OCLs were directly lysed with TRIzol (Invitrogen, Carlsbad, CA) for total RNA extraction.

For pHi measurements and cell-survival assays, OCLs were prepared by coculturing murine osteoblasts and bone marrow as described above except that the cells were plated on collagen-coated 10-cm dishes (21). We used OCLs grown on collagen for the pHi and cell survival studies because they exhibited the same response to CSF-1, in particular the same biochemical characteristics of cellular alkalinization, as authentic osteoclasts, whereas OCLs grown on plastic did not. Six days after coculturing the cells, the collagen was digested with collagenase and OCLs replated either on 12-mm2 coverslips for pHi measurements or in 48-well tissue-culture plates for cell survival experiments (TUNEL assay). Replated OCLs were allowed to adhere for 3 h before being studied. Tartrate-resistant acid phosphatase (TRAP) staining was performed as reported (22). The use of animals was approved by the Yale Animal Care and Use Committee.

Solutions
The composition of each of the solutions used in this study is given in Table 1Go. The osmolalities, measured with a vapor pressure osmometer (model 5520C; Wescor, Inc., Logan, UT), were adjusted to 300 ± 3 Osm by adding NaCl (except for solution 3, where N-methyl D-glucamine/Cl was used) or H2O. Solutions were delivered at 4 ml/min using 140-ml gas-tight syringes mounted on pumps (model 55–2222; Harvard Apparatus, South Holliston, MA). The syringes were connected to a valve array via Tygon tubing. The output from each valve was water-jacketed to ensure that the temperature of the delivered solution was 37 C (23, 24).


View this table:
[in this window]
[in a new window]

  		TABLE 1. Composition of experimental solutions

 
The pH-calibration solutions, which contained nigericin, were introduced into the perfusion chamber via an auxiliary gravity-fed inlet to avoid nigericin contamination of the plumbing and fixtures (25).

Fluorescence measurements
A coverslip with cells attached, either neonatal rat osteoclasts or murine OCLs, was fixed to the bottom of a perfusion chamber by sealing it with vacuum grease. The chamber was then secured to the stage of an inverted microscope (Olympus IX 70) equipped for epifluorescence. A differential interference contrast filter system was used to identify osteoclasts. Osteoclasts were then loaded with 10 µM 2',7'-bis-2-carboxyethyl)-5-(and-6)-carboxyfluorescein acetoxy methyl (Molecular Probes Inc., Eugene, OR) in solution 1 (Table 1Go) at room temperature for 10–15 min.

The details of the fluorescence measurement have been reported (23, 24). Briefly, light was generated at two excitation wavelengths, 440 and 490 nm, by alternately placing two excitation filters (440 ± 5 and 490 ± 5 nm; Omega Optical, Inc., Brattleboro, VT) mounted on a filter wheel (Ludl Electronic Products Ltd., Hawthorne, NY) in the light path of a 75-W xenon arc lamp. The excitation light intensity was equalized using neutral density filters mounted on a second filter wheel. The excitation light was directed to the cells via a long-pass dichroic mirror (through a x40 oil-immersion objective, NA 1.35). The emitted light at 440 nm (I440) and 490 nm (I490) was collected via a band-pass filter (530 ± 35 nm; Omega Optical) and then directed to an intensified CCD camera (model 350F; Video Scope International Ltd., Dulles, VA). The pixel intensity of the area of interest (background subtracted) at I490 was divided by the pixel intensity at I440 (background subtracted). The excitation fluorescence ratios I490/I440 were converted into pHi values by using the high-K+ nigericin technique (26) and modified to a one-point calibration (27). Cell viability was assessed during the experiment by monitoring the rate of dye loss (–k440) as described by Bevensee et al. (25).

Protein sample preparation and Western blot
Mouse kidney tissues were homogenized in 85 mM NaCl, 5 mM MgCl2, 10 mM HEPES (pH 7.6), supplemented with a protease inhibitor cocktail set I (Calbiochem EMD Biosciences Inc., La Jolla, CA). Homogenates were centrifuged at 4000 x g for 10 min at 4 C to remove cellular debris. The supernatant was respun at 100,000 x g for 1 h at 4 C. The protein concentration was measured in the supernatant and used for Western blotting. Purified OCLs were lysed in the following buffer: 150 mM NaCl, 50 mM HEPES (pH 7.5), 1% Triton X-100. 10% glycerol, containing 1 µM phenylmethylsulfonyl fluoride, 10 mg/ml aprotinin, 10 µg/ml leupeptin, 1 mM and sodium vanadate. Lysates were clarified by spinning at 16,000 x g for 10 min at 4 C. The protein concentration was measured in the supernatant and used for Western blotting. Western blotting was performed by resolving protein samples (30 µg per sample) on a 7.5% polyacrylamide gel, with subsequent transfer to a nitrocellulose membrane. The membrane was blocked in 5% dry milk dissolved in 145.4 mM NaCl, 84.7 mM Tris-HCl, 14.0 mM Trizma base (pH 7.5), and 0.1% Tween 20 [Tris-buffered saline with Tween 20 (TBS-T)] for 1 h at room temperature. The membrane was then incubated overnight at 4 C with a rabbit polyclonal antibody (1/100 dilution) against the NH2 teminus of NBCn1 (28), which was generously provided to us by Dr. Jeppe Praetorius (Water and Salt Research Center, Institute of Anatomy, University of Aarhus, Aarhus, Denmark). The membrane was then washed with TBS-T several times and incubated for 1 h at room temperature with antirabbit horseradish peroxidase conjugate (Promega, Madison, WI) at a final dilution of 1/10,000 blocking solution. Excess of antibody was removed by washing with TBS-T, and ECL chemiluminescence (Amersham, Piscataway, NJ) was used to detect bound antibody.

RT-PCR
For single-cell RT-PCR, one osteoclast was retrieved at a time from a coverslip by suction into a patch pipette backfilled with RNAlater (Ambion, Austin, TX). The tip of the pipette was broken into a 0.5-ml thin-walled Eppendorf tube containing the PCR mixture described below and then analyzed by RT-PCR. Single-cell RT-PCR was performed using the following primer pair designed on the basis of the rat cDNA sequence for NBCn1 (NM_003615). The forward (sense) primer was 5'-AAGTTCCTCGGAATTCGTGAACA-3', and the reverse (antisense) primer was 5'-TGGTTCATCTTCGAAATTTATTGTCAC-3'. The expected PCR fragment size is 490 bp. PCR was performed with the Expand Long Template kit and polymerase from Roche (Nutley, NJ) on a Genius thermocycler (Techne, Burlington, NJ). Amplification was carried out for 32 cycles of 94 C denaturation for 35 sec, 57–55 C annealing for 1 min, and 68 C extension for 1 min (68 C extension for 10 min during the last cycle). Reactions performed in the absence of reverse transcriptase served as negative controls. PCR products were analyzed on a 1.5% agarose gel. PCR products cut from the gel were sequenced by the W. M. Keck Biotechnology Resource Laboratory at Yale.

RT-PCR was also performed on RNA isolated from OCLs, based on the mouse cDNA sequence (NM_001033270). The forward (sense) primer was 5'-GTTGAATCTGAATGTTCTGCTCCAGGG-3', and the reverse (antisense) primer was 5'-TGGACCTTCCAGAGCGGCACATAAC-3'. The expected PCR fragment size is 284 bp. SuperScript One-Step RT-PCR with Platinum Taq (Invitrogen) was used for these reactions. cDNA synthesis and pre-denaturation were carried out for one cycle at 50 C, cDNA synthesis for 30 min, and denaturation at 94 C for 2 min. Amplification was carried out for 37 cycles of denaturation at 94 C for 15 sec, annealing at 55 C for 30 sec, and extension at 72 C for 30 sec. Glyceraldehyde-3-phosphate dehydrogenase was used as a control.

Cell survival
To determine the effect of CSF-1 on OCL survival in the presence (solution 2) and absence (solution 1) of CO2/HCO3–, OCLs were prepared on collagen as described above and then replated to 48-well plates either in HEPES buffer in a 37 C air incubator or in CO2/HCO3–, buffer in a 5% CO2 incubator. OCLs cultured in these two media were then either treated with vehicle or 2.5 nM CSF-1. Thus, there were four experimental conditions: 1) cells in HEPES buffer (solution 1) with CSF-1, 2) cells in HEPES buffer without CSF-1, 3) cells in CO2/HCO3– buffer (solution 2) with CSF-1, and 4) cells in CO2/HCO3– buffer without CSF-1. TRAP-positive cells containing three or more nuclei were counted as osteoclasts. Apoptotic osteoclasts were identified as TRAP-positive cells that also stained positively for DNA fragments as determined using a commercially available kit and following the manufacturer’s recommended protocol (Calbiochem, EMD Biosciences). In an initial time-course experiment, the number of apoptotic cells was determined after 3, 4, 5, and 6 h of culture in the two buffers. At 4 h, there was 50% survival among cells cultured with HEPES in the absence of CSF-1. Thus, in subsequent experiments, cells were cultured for 4 h under the four conditions noted above and stained for TRAP and DNA fragments. All cells in five fields (at x2 magnification) in each well of every plate were counted. The total number of TRAP-positive cells as well as the number of cells that also stained for DNA fragmentation was recorded. The results from five fields were averaged to yield one set of numbers for that well. Ten wells were analyzed for each treatment condition, and the entire experiment was performed twice.

The effect of inhibiting caspase-3 or -8 was compared with the effect of CSF-1 on OCL survival. Cells were cultured in CO2/HCO3– in the presence of Z-DEVD-FMK, a caspase-3 inhibitor (16, 29) or Z-IETD-FMK, a caspase-8 inhibitor (30) (both from BD Biosciences, San Jose, CA). Thus, cells were cultured under four experimental conditions: 1) in the presence of CO2/HCO3– (solution 2) alone, 2) solution 2 with 2.5 nM CSF-1, 3) solution 2 with 10 nM caspase 3 inhibitor, and 4) solution 2 with caspase-8 inhibitor. The number of apoptotic cells was determined as described above. For experiments examining the effect of etoposide on cell survival, osteoclasts were cultured in the presence of 100 µM etoposide (Sigma Chemical Co., St. Louis, MO) for 6 h.

Assessing caspase activation in OCLs
Activation of caspase-2, -3, -8, and -9 was assayed using BD APoAlert caspase assay plates (BD Biosciences) according to the manufacturer’s protocol. Briefly, OCLs prepared on collagen were incubated either in HEPES buffer (solution 1) or in CO2/HCO3–-buffer (solution 2) with or without CSF-1. OCLs were harvested by collagenase digestion and gently spun at 400 x g for 5 min and the supernatant removed. The pellet was resuspended in ice-cold cell lysis buffer and incubated on ice for 10 min. The cell lysate was then centrifuged at maximum speed at 4 C for 5 min in an Eppendorf microfuge and the supernatant transferred to a new tube and kept on ice. Reactions using these lysates were carried out according to the manufacturer’s protocol and enzyme activity quantified by fluorescence intensity using a 96-well plate reader (Fluostar/Polarstar Galaxy; BMG Technologies, Durham, NC), a 360 ± 5-nm excitation filter and 460 ± 5-nm emission filter. All measurements were done at room temperature.

Data analysis
Data are reported as mean ± SEM (computed on the basis of n – 1). Means were compared using either paired or unpaired Student’s t test (two tailed) as indicated in the figure legends and text. One-way ANOVA with Dunnet’s post hoc testing was used to evaluate the results using the caspase inhibitors. P < 0.05 was considered significant.


   Results
Top
 Abstract
 Introduction
 Materials and Methods
 Results
Discussion
 References
 
Effect of CSF-1 on pHi in freshly isolated osteoclasts: dependency on CO2/HCO3–
The effect of 2.5 nM CSF-1 on the pHi of freshly isolated rat osteoclasts was examined using the protocol illustrated in Fig. 1AGo. Cells were initially perfused with a solution equilibrated with 5% CO2/22 mM HCO3– (solution 2, Table 1Go). After pHi had stabilized, the bath solution was changed to the same solution containing 2.5 nM CSF-1. The addition of CSF-1 caused a substantial increase in pHi over a period of approximately 2 min. In a total of seven similar experiments, mean pHi increased from 7.14 ± 0.03 to 7.34 ± 0.03 (P < 0.001 compared with baseline), as summarized in Fig. 1BGo. In experiments in which the exposure to CSF-1 was prolonged for approximately 5 min, the alkalinization persisted as long as CSF-1 was present in the solution. Removing CSF-1 from the solution caused pHi to return toward its initial value, although the final pHi remained slightly higher (7.21 ± 0.04; n = 7 cells) than the initial value. Subsequently exposing the cells to a HEPES-buffered solution (solution 1, Table 1Go) caused an initial rise in pHi (presumably due to the efflux of CO2), which then fell to a mean value of 7.13 ± 0.04. With the osteoclast in a HEPES buffer, a second exposure to CSF-1 elicited a markedly attenuated response, with pHi increasing only from 7.13 ± 0.04 to 7.16 ± 0.04 (P < 0.05). In 5% CO2/22 mM HCO3–, CSF-1 caused a mean pHi increase ({Delta}pHi) of 0.21 ± 0.03 (n = 7 cells), whereas in HEPES buffer, CSF-1 induced a pHi increase of only 0.03 ± 0.01 pH units (P < 0.001 for comparing the pHi increment in the two buffers). Reversing the sequence in which cells were exposed to the two buffers did not alter the results. Thus, the rise in pHi induced by CSF-1 depends critically on the presence of CO2/HCO3–.


Figure 1
View larger version (14K):
[in this window]
[in a new window]

  		FIG. 1. Effect of CSF-1 on pHi in freshly isolated rat osteoclasts in the presence and absence of CO2/HCO3–. A, Representative pHi recording in a single osteoclast. An osteoclast loaded with 2',7'-bis-2-carboxyethyl)-5-(and-6)-carboxyfluorescein, was initially superfused with a solution buffered with 5% CO2/22 mM HCO3– (solution 2). During the indicated period, 2.5 nM CSF-1 was added to the solution, inducing a significant increase in pHi. In the same osteoclast, the solution was then switched to one containing no CO2/HCO3– (solution 1), and during the indicated period, the osteoclast was again exposed to 2.5 nM CSF-1. Under these conditions no change in pHi was observed. B, Mean CSF-1-induced change in pHi ({Delta}pHi) in the presence (black bar) or absence (white bar) of CO2/HCO3– in seven osteoclasts. {Delta}pHi was calculated by subtracting the pHi just before the application of CSF-1 from the peak or plateau pHi during CSF-1 administration. ***, P < 0.001, paired t test for the difference in mean {Delta}pHi. The CSF-1-induced change in pHi from baseline (7.14 ± 0.03 to 7.34 ± 0.03) was significant in CO2/HCO3– solution (P < 0.001, paired t test). In HEPES, CSF-1 induced a trivial rise in pHi (7.13 ± 0.04 to 7.16 ± 0.04; P < 0.05, paired t test).

 
Effect of external Na+ removal on CSF-1-induced alkalinization in freshly isolated osteoclasts
The data in Fig. 1Go are consistent with the idea that CSF-1 activates a CO2/HCO3–-dependent transporter. To examine the possibility that the CSF-1-induced osteoclast alkalinization requires a Na+-coupled HCO3– transporter, we examined the effect of removing external Na+ on the subsequent response to CSF-1. A representative pHi tracing is shown in Fig. 2AGo. Perfusing osteoclasts with Na+-free CO2/HCO3– (solution 3, Table 1Go) caused a downward drift in pHi, which may reflect the reversal of a Na-H exchanger or a Na+-coupled HCO3– transporter. The subsequent addition of CSF-1 did not affect the downward trajectory of pHi, and in the absence of Na+, CSF-1 did not increase pHi (mean {Delta}pHi, –0.01 ± 0.02; n = 9 cells; P = 0.5). Reintroducing extracellular Na+ led to a pHi recovery and restored the ability of CSF-1 to alkalinize the cell ({Delta}pHi = 0.15 ± 0.03; n = 9 cells; P < 0.001). As summarized in Fig. 2BGo, CSF-1 induced a significantly greater increase pHi in the presence of extracellular Na+ than in its absence (n = 9 cells; P < 0.001 for comparison of the {Delta}pHi under the two culture conditions). Taken together, these data show that the alkalinization induced by CSF-1 requires external Na+ and is consistent with the activation of a Na+-coupled HCO3– transporter.


Figure 2
View larger version (16K):
[in this window]
[in a new window]

  		FIG. 2. Effect of removing Na+ from the perfusate on the CSF-1-induced rise in pHi in freshly isolated rat osteoclasts. A, Representative pHi recording in a single osteoclast. The osteoclast was superfused with solution 2. During the indicated periods, the osteoclast was exposed to CSF-1, first in the absence of extracellular Na+ and a second time in the presence of extracellular Na+. B, Mean CSF-1-induced change in pHi ({Delta}pHi) in the presence (black bar) or absence (checkered bar) of extracellular Na+ in nine osteoclasts. **, P < 0.01, paired t test. The CSF-1-induced change in pHi from baseline (7.13 ± 0.03 to 7.28 ± 0.05) was significant in the presence of sodium (P < 0.001, paired t test) but not in the absence of sodium (7.07 ± 0.07 to 7.06 ± 0.08; P = 05, paired t test).

 
Effect of 4,4'-diisothiocyanatostilbene-2,2'-disulfonic acid (DIDS) on CSF-1-induced alkalinization in freshly isolated osteoclasts
We next determined whether DIDS, an inhibitor of many but not all HCO3– transporters, would inhibit the ability of CSF-1 to alkalinize osteoclasts. Figure 3AGo shows the representative response of an individual osteoclast perfused with a 5% CO2/22 mM HCO3– buffer. Exposing the cell to CSF-1 caused pHi to rise by somewhat more than 0.1. In 13 cells with a mean initial pHi of 7.12 ± 0.03, the mean {Delta}pHi after exposure to CSF-1 was 0.13 ± 0.02, resulting in a mean final pHi of 7.25 ± 0.03 (P < 0.001 compared with baseline). After removing CSF-1 in the experiment shown in Fig. 3AGo, 200 µM DIDS was added to the solution. Two minutes later, 2.5 nM CSF-1 was reapplied in the continued presence of DIDS. Despite the presence of DIDS, CSF-1 was still able to induce significant alkalinization with mean pHi rising from 7.14 ± 0.04 to 7.22 ± 0.03 (n = 13 cells; P < 0.001). As summarized in Fig. 3BGo, the mean {Delta}pHi induced by CSF-1 in the presence of DIDS ({Delta}pHi = 0.08 ± 0.02) was not significantly different from the value obtained in the absence of DIDS ({Delta}pHi = 0.13 ± 0.02; P = 0.13 for the difference; n = 13 cells). These data demonstrate that the alkalinization induced by CSF-1 is not substantially DIDS sensitive.


Figure 3
View larger version (24K):
[in this window]
[in a new window]

  		FIG. 3. Effect of DIDS on the CSF-1-induced rise in pHi in freshly isolated rat osteoclasts. A, Representative pHi recording in a single osteoclast. The osteoclast was superfused with solution 2. During the indicated periods, the osteoclast was exposed to CSF-1, once in the absence and once in the presence of 200 µM DIDS. B, Mean CSF-1-induced change in pHi ({Delta}pHi) in the absence (black bar) or presence (striped bar) of CO2/HCO3– in 13 osteoclasts (P = 0.13 for the difference between the two mean {Delta}pHi values, paired t test). The CSF-1-induced change in pHi from baseline was significant both in the absence of DIDS (7.12 ± 0.03 to 7.25 ± 0.03; P < 0.001, paired t test) and in its presence (7.14 ± 0.04 to 7.22 ± 0.03; P < 0.001, paired t test).

 
Effect of 5-(N-ethyl-N-isopropyl) amiloride (EIPA) and concanamycin A on CSF-1-induced alkalinization in freshly isolated osteoclasts
Among the Na+-coupled HCO3– transporter family, only the electroneutral Na/HCO3 cotransporter NBCn1 is relatively DIDS insensitive (31, 32). On the other hand, work with out-of-equilibrium CO2/HCO3– solutions (33) in the renal proximal tubule indicates that CO2, even at fixed extracellular [HCO3–] and pH, can stimulate acid-base transporters, presumably including the Na-H exchangers and H+ pumps responsible for transepithelial HCO3– transport (34). Thus, we entertained the hypothesis that sodium and bicarbonate, together, are required to stimulate a Na-H exchanger and/or H+ pump.

To explore the possibility that CSF-1 acts through a Na-H exchanger or H+ pump, we examined the effect of EIPA, an inhibitor of the Na-H exchanger (35), and concanamycin A, an inhibitor of the vacuolar-type H+ pump (36) on CSF-1-dependent osteoclast alkalinization. The experimental protocol was identical to that used in examining the effects of DIDS (Fig. 3Go) except that either 50 µM EIPA or 1 nM concanamycin A replaced DIDS. As summarized in Fig. 4Go, neither EIPA nor concanamycin A significantly attenuated the rise in pHi induced by CSF-1.


Figure 4
View larger version (36K):
[in this window]
[in a new window]

  		FIG. 4. Effect of EIPA and concanamycin A on the CSF-1-induced rise in pHi in freshly isolated rat osteoclasts. The experimental protocol was the same as described in Fig. 3AGo except that the osteoclasts were exposed either to EIPA or concanamycin A. A, Mean CSF-1-induced change in pHi ({Delta}pHi) in the absence (black bar) or presence (striped bar) of 50 µM EIPA in eight osteoclasts (P = 0.4 for the difference in the mean {Delta}pHi values, paired t test). The CSF-1-induced change in pHi from baseline was significant both in the absence of EIPA (7.21 ± 0.04 to 7.29 ± 0.04; P < 0.001, paired t test) and in its presence (7.19 ± 0.04 to 7.26 ± 0.04; P < 0.01, paired t test). B, Mean CSF-1-induced change in pHi ({Delta}pHi) in the absence (black bar) or presence (checkered bar) of 1 nM concanamycin A in 15 osteoclasts (P = 0.06 for the difference in the mean {Delta}pHi values, paired t test). The CSF-1-induced change in pHi from baseline was significant both in the absence of concanamycin A (7.12 ± 0.04 to 7.19 ± 0.04; P < 0.001, paired t test) and in its presence (7.13 ± 0.03 to 7.21 ± 0.03; P < 0.001, paired t test).

 
Effect of CSF-1 on pHi in OCLs
Because the numbers of isolated cells are so few, performing biochemical experiments with freshly isolated osteoclasts is not feasible. Therefore, OCLs are required for biochemical analyses. To confirm that OCLs respond to CSF-1 in a manner similar to that observed in freshly isolated authentic osteoclasts, we performed three series of experiments.

First, we exposed OCLs to 2.5 nM CSF-1 in the absence or presence of CO2/HCO3– following the protocol shown in Fig. 1AGo. As summarized in Fig. 5AGo, and as was true for freshly isolated osteoclasts, exposing OCLs to CSF-1 in the presence of CO2/HCO3– caused a significant rise in pHi, ({Delta}pHi = 0.14 ± 0.02; n = 6, P < 0.01 compared with baseline), whereas the exposing OCLs to CSF-1 in the absence of CO2/HCO3– (i.e. HEPES buffer) did not cause a significant pHi change ({Delta}pHi = –0.01 ± 0.01; n = 6; P = 0.6 compared with baseline).


Figure 5
View larger version (18K):
[in this window]
[in a new window]

  		FIG. 5. Effect of CSF-1 on pHi in mouse OCLs. A, Mean CSF-1–induced change in pHi ({Delta}pHi) in the presence (filled bar) or absence (open bar) of CO2/HCO3– in 6 OCLs (** P < 0.01, paired t test). The protocol was identical to that used in Fig. 1AGo. B, Mean CSF-1–induced change in pHi ({Delta}pHi) in the presence (filled bar) or absence (sphered bar) of extracellular Na+ in 4 OCLs (* P < 0.05, paired t test). The protocol was identical to that used in Fig. 2AGo. C, Mean CSF-1-induced change in pHi ({Delta}pHi) in the absence (filled bar) or presence (hatched bar) of 200 µM DIDS in 3 OCLs (*, P < 0.05, paired t test). The protocol was identical to that used in Fig. 3AGo.

 
Second, we verified that removing external Na+ abolishes the rise in pHi. Figure 5BGo summarizes the results obtained using the same protocol outlined in Fig. 2AGo. The mean pHi increment in the presence of Na+ was 0.39 ± 0.10 (n = 4; P < 0.05), whereas in the absence of Na+, it was 0.003 ± 0.01 (n = 4; P = 0.8 compared with baseline).

Finally, using the same protocol shown in Fig. 3AGo, we verified that DIDS does not have a major impact on the CSF-1-induced rise in pHi. As summarized in Fig. 5CGo, the rise in pHi induced by CSF-1 in the presence of DIDS ({Delta}pHi = 0.12 ± 0.03; P < 0.01 compared with baseline) was only slightly less than that caused by CSF-1 in the absence of DIDS (0.14 ± 0.01; n = 3; P < 0.01 compared with baseline), although the difference in mean {Delta}pHi was statistically significant (P < 0.05).

Collectively, these data demonstrate that the pHi response of OCLs to CSF-1 is analogous to that of freshly isolated osteoclasts.

Freshly isolated osteoclasts and OCLs express NBCn1
Our cellular-pH studies are consistent with the hypothesis that NBCn1 is a key participant in CSF-1-induced cellular alkalinization in osteoclasts. To determine whether rat osteoclasts express transcripts for NBCn1, we undertook single-cell PCR using RNA isolated from individual, freshly prepared rat osteoclasts as well as from cultured OCLs. As shown in Fig. 6AGo, we observed a PCR product of the expected size (490 bp) in two separate single-cell PCR experiments (lanes 1 and 2). The results of direct DNA sequencing confirmed that the products from these two separate single-cell PCR matched the published sequence for rat NBCn1 (GenBank accession no. NM058211). As shown in Fig. 6BGo, we also observed a PCR product of the expected size (284 bp), using RNA prepared from mouse OCLs and a primer pair based on the sequence of murine NBCn1.


Figure 6
View larger version (39K):
[in this window]
[in a new window]

  		FIG. 6. Expression of NBCn1 mRNA and protein in freshly isolated rat osteoclasts and OCLs. A, NBCn1 mRNA expression in a single freshly isolated neonatal rat osteoclast. Single-cell RT-PCR was performed as described in the Materials and Methods. A single amplicon of the expected size, 490 bp, was observed in two separate reactions carried out on two individually isolated cells (lanes 1 and 2). When reverse transcriptase was eliminated from the reaction, no amplicon was observed (lane 3). B, NBCn1 mRNA expression in OCLs. RT-PCR was performed on total RNA isolated from OCLs (lane 1, upper panel) and from murine kidney (lane 2, upper panel; as a positive control). A single amplicon of the expected size, 284 bp, was observed in both reactions. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as an internal control for the PCR (lower panels), and an amplicon of the expected size (260 bp) and of equal intensity was observed in both lanes. C, NBCn1 protein expression in osteoclasts. Immunoblotting was performed using a polyclonal antibody directed against the NH2 terminus of NBCn1. In both mouse OCLs (lane 1) and in kidney lysates (lane 2, positive control), bands at 150 and 135–140 kDa were detected, the latter probably representing the unglycosylated form of NBCn1.

 
To confirm expression of NBCn1 protein in OCLs, we performed immunoblotting on OCL lysates, using mouse kidney as a positive control. As shown in Fig. 6CGo, using an anti-NH2 NBCn1 antibody (28), we detected an approximately 150-kDa protein in both OCL and kidney lysates. A second 130-kDa protein was detected in OCL lysates (lane 1) and a second approximately 135-kDa protein in kidney lysates (lane 2). The lower-molecular-weight proteins likely represent unglycosylated protein because the predicted molecular mass of unglycosylated NBCn1 splice variants ranges from 135–140 kDa. In rat kidney and other tissues, this antibody detects bands at approximately 180 and 140 kDa (28). The difference in apparent mass of the proteins in rat and mouse preparations may be due to species differences.

CSF-1 increases the survival of OCLs
To determine whether the NBCn1-dependent rise in pHi induced by CSF-1 plays a role in the prosurvival effects of CSF-1 in these cells, the following studies were undertaken. OCLs prepared on collagen-coated dishes were replated to 48-well plates. The OCLs were then exposed for 4 h to solution 1 (i.e. no CO2/HCO3–) in the absence or presence of 2.5 nM CSF-1 or to solution 2 (i.e. with CO2/HCO3–) in the absence or presence of 2.5 nM CSF-1. After 4 h, DNA fragmentation was quantified by TUNEL, as detailed in the Materials and Methods. As summarized in Fig. 7AGo, the percentage of apoptotic cells was significantly greater when OCLs were cultured in HEPES-buffered solution than when they were cultured in a solution containing CO2/HCO3– (21.4 ± 1.4 vs. 6.1 ± 0.7% respectively; n = 10 wells; P < 0.001, unpaired t test). When cells were cultured in HEPES buffer, CSF-1 had no impact on the apoptotic rate (21.0 ± 1.3% with CSF-1 vs. 21.4 ± 1.4% without CSF-1; n = 10 wells; P = 0.8, unpaired t test). In contrast, when cells were cultured in the presence of CO2/HCO3–, 2.5 nM CSF-1 induced a significant reduction in the number of apoptotic cells (3.2 ± 0.4 vs. 6.1 ± 0.7%; P < 0.01; n = 10 wells, unpaired t test).


Figure 7
View larger version (27K):
[in this window]
[in a new window]

  		FIG. 7. Effect of caspase-8 and -3 inhibitors and of CSF-1 on OCL survival in the presence and absence of CO2/HCO3–. A, OCLs were cultured as described in the Materials and Methods and replated into media containing either CO2/HCO3– (solution 2) or HEPES (solution 1). The effect of CSF-1 to reduce apoptosis at 4 h under each of these experimental conditions was quantified by counting the number of apoptotic osteoclasts. Plotted is the percentage of apoptotic cells, normalized to the total number of OCLs in each well. In the HEPES-buffered solutions, adding CSF-1 did not significantly affect the number of apoptotic OCLs (P = 0.8, unpaired t test). Even without CSF-1 treatment, there were significantly fewer numbers of apoptotic OCLs at the end of 4 h when cells were incubated in CO2/HCO3– compared with HEPES (P < 0.001, unpaired t test). Adding CSF-1 to the CO2/HCO3– buffer significantly reduced the number of apoptotic OCLs (P < 0.001). NS, Not significantly different from OCLs cultured in HEPES without CSF-1; ***, P < 0.001 compared with OCLs cultured in HEPES alone; ###, P < 0.001 vs. cells cultured in HEPES plus CSF-1 and vs. cells cultured in CO2/HCO3– alone (n = 10 fields for all four conditions). B, Effects of a CO2/HCO3–-buffered solution 1, 2.5 nM CSF-1 in presence of CO2/HCO3–, 10 nM Z-IETD caspase-8 inhibitor, and 10 nM Z-DEVD caspase-3 inhibitor on the apoptotic rate of mature OCLs incubated in CO2/HCO3– buffer for 4 h. The histogram summarizes the results from three separate experiments (n = 42, 46, 46, and 44 fields, respectively). The percentage of apoptotic cells was determined by the following formula: number of apoptotic OCLs per field/total number of OCLs per field x 100. To sum the three experiments, the individual data in each experiment were divided by the mean control value for that experiment (CO2/HCO3–) resulting in a mean control value of 1.0 in all three experiments. There was a significant overall treatment effect by one-way ANOVA. *, Significantly different from the control value by Dunnet’s post hoc testing (P < 0.05). There were no significant differences among the three treatment groups (CSF-1, caspase-8 inhibitor, and caspase-3 inhibitor). Data are mean ± SEM. All fields in both experiments were counted at x20 magnification.

 
It has been reported that an alkaline pHi may protect against apoptosis (37, 38, 39). Because, as just noted (Fig. 7AGo), OCLs cultured with CO2/HCO3– have a lower basal rate of apoptosis, we wondered whether osteoclasts cultured under these conditions might be better able to maintain a higher steady-state pHi, which presumably would protect them from proapoptotic stresses. To test this idea, we cultured osteoclasts in the absence or presence of CO2/HCO3– and exposed them to 100 µM etoposide, a well known apoptotic agent (39, 40, 41, 42), for 6 h. We measured pHi after 6 h of exposure to etoposide because it has been previously reported that etoposide-induced caspase activation peaks at 6 h (39, 40, 41).

The mean steady-state pHi, measured over a period of 10–15 min in 11 osteoclasts treated with etoposide in CO2/HCO3–-buffered solution (solution 2), was 7.23 ± 0.04. In eight osteoclasts incubated with etoposide in the absence of CO2/HCO3– (solution 1), the mean steady-state pHi measured over a period of 10–15 min was 7.09 ± 0.03, which was significantly lower than the mean pHi in CO2/HCO3– buffer (P < 0.05, unpaired t test).

Apoptosis is often associated with the activation of the caspases cascade. The activity of caspase-2, -3, -8, and -9 was therefore measured in OCL lysates after 30 min of incubation in CO2/HCO3–-buffered solution containing either 2.5 nM CSF-1 or vehicle. Background was high in these experiments, reducing the sensitivity of the assay. Although CSF-1 did not have a statistically significant effect on the activity of caspase-2, -3, and -9, the cytokine significantly reduced caspase-8 activity by 14.9 ± 7.3% (P < 0.05, unpaired t test; n = 15).

There was no significant effect of CSF-1 on caspase-3 activity, which lies downstream from caspase-8. However, in seven of 10 experiments, CSF-1 did reduce the activity of caspase-3 (P < 0.02 for those seven experiments, unpaired t test). To explore further the involvement of caspase-3 and -8 in OCL apoptosis, we incubated cells for 4 h in CO2/HCO3–-buffered solution containing either 10 nM Z-DEVD, a caspase-3 inhibitor (29), or 10 nM Z-IETD, a caspase-8 inhibitor (30). We compared the apoptotic rate in these cells with the rate of apoptosis in OCLs incubated in CO2/HCO3– alone or CO2/HCO3– with 2.5 nM CSF-1. As summarized in Fig. 7BGo, treatment with 2.5 nM CSF-1, the caspase-8 inhibitor, or the caspase-3 inhibitor reduced the percentage of apoptotic cells by a comparable 40%. There was an overall treatment effect by one-way ANOVA (P < 0.05). Post hoc testing confirmed that there was no statistically significant difference in the response to any of the three treatments (e.g. CSF-1 vs. caspase-8, P = 0.9; CSF-1 vs. caspase-3 inhibitor, P = 0.5; caspase-3 inhibitor vs. caspase-8 inhibitor, P = 0.2, Dunnet’s).


   Discussion
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
References
 
Considerable interest has focused on the prosurvival effects of CSF-1 in osteoclasts. Previous work has established that cellular acidification plays a key role in the apoptotic pathway (37, 43, 44, 45, 46). Because growth factors modulate both HCO3–-dependent and HCO3–-independent acid-base transporters (47, 48, 49) responsible for regulating pHi, the present study was designed to determine the effect of CSF-1 on osteoclast pHi and whether any observed pHi changes were relevant to the known ability of this growth factor to prolong osteoclast survival.

CSF-1 activates the electroneutral Na/HCO3 cotransporter NBCn1
CSF-1 was found to significantly increase pHi in both osteoclasts and OCLs (Figs. 1Go and 5Go). Moreover, the rise in pHi induced by CSF-1 requires CO2/HCO3– (Figs. 1Go and 5Go) as well as extracellular Na+ (Figs. 2Go and 5Go). These data support the involvement of a Na/HCO3 cotransporter in CSF-1-induced alkalinization. The failure of DIDS to block the effect of CSF-1 (Figs. 3Go and 5Go) excludes most HCO3– transporters as candidates for mediating CSF-1’s actions. In fact, NBCn1 is the one known Na+-coupled HCO3– transporter that is largely DIDS resistant (31, 32). This finding suggests NBCn1 as a candidate effector for CSF-1-dependent alkalinization in osteoclasts and OCLs. Consistent with this notion, transcripts for NBCn1 are present in freshly isolated mature osteoclasts (Fig. 6AGo) and OCLs (Fig. 6BGo), and immunoblotting experiments confirmed NBCn1 protein expression (Fig. 6CGo).

When the removal of ammonium chloride (50) is used to impose intracellular acidification in osteoclasts, the subsequent pHi recovery in the absence of CO2/HCO3– is mediated almost exclusively by the Na-H exchanger, the H+ pump, and by an H+ conductance (8, 9, 10). However, the Na-H exchanger and the H+ pump are unlikely to participate in the alkalinization induced by CSF-1 because neither EIPA (an inhibitor of the Na-H exchanger) nor concanamycin A (an inhibitor of the H+ pump) significantly reduced the effect of CSF-1 on pHi (Fig. 4Go).

We are unaware of previous reports of a Na/HCO3 cotransporter in osteoclasts. In avian osteoclasts, adding CO2/HCO3– causes a sustained increase in pHi and removing CO2/HCO3– has the opposite effect (7). These data are consistent with the presence of a HCO3– -uptake mechanism.

A role for NBCn1 in osteoclast survival
Because CSF-1 promotes osteoclast survival, we entertained the hypothesis that the alkalinization induced by NBCn1 played a role in mediating this effect. Culturing cells in a CO2/HCO3– solution alone provided a substantial survival benefit compared with cells cultured in HEPES buffer. The reason for this survival advantage is unclear, although Tsao and Lei (51) reported the same finding in thymocytes. These investigators found that approximately 27% of the thymocytes were apoptotic when cultured in a HEPES-buffered solution, whereas only about 16% were apoptotic in CO2/HCO3–, a reduction of 40%. In our study, we observed 21% of osteoclasts to be apoptotic when cultured in HEPES compared with 6% for cells grown in CO2/HCO3–, a 70% difference. The greater prosurvival effect in our study may reflect our use of 22 mM HCO3– instead of the 10 mM HCO3– used by Tsao and Lei. One explanation for the marked reduction in apoptosis observed in CO2/HCO3– is that cells are better able to regulate pHi in the presence of a CO2/HCO3– (52), thereby protecting them from large swings in pHi, which are associated with apoptosis (38). Consistent with this latter notion, we found that osteoclasts exposed to etoposide are able to maintain a higher pHi in the presence of CO2/HCO3– than in HEPES buffer.

In addition to the prosurvival effect of CO2/HCO3– per se, CSF-1 further reduced osteoclast apoptosis by 50% in the presence of CO2/HCO3– but not when cells were cultured in HEPES. Given our data demonstrating that the pHi increase elicited by CSF-1 requires NBCn1, the most straightforward explanation is that the CSF-1 activates NBCn1, which in turn causes an alkalinization that promotes survival.

Although some reports indicate that maintaining an alkaline pHi does not inhibit programmed cell death (38, 53, 54), several others demonstrate that preventing a fall in pHi does inhibit apoptosis. In HL-60 cells, lovastatin-induced apoptosis was accompanied by a fall in pHi. Activation of the Na-H exchanger by phorbol ester prevented both the lovastatin-induced fall in pHi and apoptosis (37, 38, 44). Similarly, Barrière et al. (45) found that lovastatin-induced apoptosis in CFTR-transfected hamster fibroblasts was also associated with a fall in pHi. In their model, overexpressing the Na-H exchanger prevented the fall in pHi and reduced apoptotic rates. The Na-H exchanger is not the sole pHi regulator that has been reported to have an effect on apoptotic rate. In neutrophils, Gottlieb et al. (43) showed that the prosurvival effect of granulocyte CSF was due to stimulation of the H+ pump. Blockade of the pump by bafilomycin abrogated this effect of granulocyte CSF. Finally, Okahashi et al. (44) reported that inhibiting the vacuolar H+ pump in OCLs with concanamycin accelerated OCL apoptosis. Although these investigators did not directly measure pHi, one would expect blockade of the H+ pump to result in a fall in pHi. The beneficial effect of an alkaline pHi seems to occur over a relatively narrow pH range. Tsao and Lei (51) have reported that an increase in pHi above 7.6 in thymocytes was associated with increased apoptotic rates. Similar results were reported in megakaryocytic cells (55).

Molecular mechanism of the prosurvival effect of CSF-1
Having shown that CSF-1’s prosurvival effect occurred only in CO2/HCO3– buffer, we sought to determine what cellular factors contributed to this effect. It has been previously reported that serum starvation increases OCL apoptosis but that caspase inhibitors extend OCL survival under these conditions (18). We used a fluorometric approach to measure the activity of several key caspases. Treatment with CSF-1 in CO2/HCO3– buffer did not reduce the activity of caspase-2 and -9 in OCLs below that observed in CO2/HCO3– buffer alone. In contrast, CSF-1 significantly inhibited activation of caspase-8, suggesting that this may be one mechanism for the prosurvival effect of CSF-1. Although caspase-3 lies downstream from caspase-8 and is considered one of the final execution caspases, we were unable to demonstrate a significant effect of CSF-1 on caspase-3 activity. Variability in the assay may have contributed to the lack of statistical significance. Consistent with a role of caspase-8 in osteoclast apoptosis, we found that an inhibitor of this caspase significantly prolongs the survival of OCLs. The magnitude of the prosurvival effect of caspase-8 and -3 inhibitors was comparable to the effect of CSF-1, which also inhibited apoptosis by approximately 40% (Fig. 7BGo). In the aggregate, these data support the conclusion that the caspase cascade, including caspase-8 and possibly caspase-3 are targets for CSF-1’s prosurvival effect.

Does an increase in pHi or [HCO3–] directly suppress caspase activity? In a cell-free system, the ability of cytochrome c to activate caspase-9 is pH dependent (56). Cytochrome c binds the protein Apaf-1, inducing its oligomerization, which results in the recruitment and activation of procaspase-9. The optimal pH for this reaction is approximately 6.4. Increasing pH to 7.4 inhibits caspase activation by 75% (56). A similar observation has been reported for caspase-3 activation by cytochrome c, where increasing pH above 7.6 completely inhibits caspase-3 activation (19). Finally, the direct effects of HCO3– on caspase-3 and -9 activity have been analyzed in vitro. Thus, when the [HCO3–] in the reaction mixture was increased from 10 mM (corresponding to a pHi of ~7.0 at 5% CO2) to 25 mM (corresponding to a pHi of ~7.4), the ability of cytochrome c to activate caspase-3 and -9 was reduced by approximately 30% (57). The bicarbonate ion itself appears to have direct effects on the catalytic activity of caspases, albeit at relatively high concentrations. Thus, Dong et al. (57) showed that 50 mM HCO3– inhibited the activity of recombinant caspase-3 and -9 by approximately 20 and 30%, respectively. In our experiments, CSF-1 in the presence of CO2/HCO3– raised pHi in freshly isolated osteoclasts from 7.14 to 7.36, corresponding to a rise in [HCO3–]i from approximately 12 to 20 mM. These data are consistent with the hypothesis that a CSF-1-dependent, NBCn1-mediated increase in pHi and cellular [HCO3–] could inhibit caspase activity in osteoclasts.

In summary, our results show that CSF-1 causes a substantial increase in pHi via the activation of the Na/HCO3 cotransporter NBCn1. The prosurvival effect of CSF-1 is enhanced when OCLs are bathed in a CO2/HCO3– medium compared with a HEPES-buffered solution. The molecular mechanism by which CSF-1-dependent alkalinization increases OCL survival may be associated with inhibition of caspase-8 and -3. NBCn1 may represent a novel therapeutic target for preventing pathological bone loss.


   Acknowledgments
 
We thank Drs. J. Prætorius and C. Aalkjær for providing the NBCn1 antibody and Dr. S. Krueger for help with isolating osteoclasts for the single-cell PCR studies.


   Footnotes
 
Present address for C.M.F.: Kilguss Research Center, 200 Chestnut Street, Providence, Rhode Island 02905.

This work was supported by National Institutes of Health Grants DK45228, DE12459, and P30 AR46032 (to K.L.I.) and by NS18400 (to W.F.B.). P.B. was supported by the National Kidney Foundation.

Disclosure Statement: All of the authors have nothing to declare.

First Published Online October 26, 2006

1 P.B. and H.S. contributed equally to the work. Back

Abbreviations: CSF-1, Colony-stimulating factor-1; DIDS, 4,4'-diisothiocyanatostilbene-2,2'-disulfonic acid; EIPA, 5-(N-ethyl-N-isopropyl) amiloride; FBS, fetal bovine serum; OCL, osteoclast-like cell; pHi, intracellular pH; TBS-T, Tris-buffer saline with Tween 20; TRAP, tartrate-resistant acid phosphatase.

Received April 25, 2006.

Accepted for publication October 18, 2006.


   References
Top
 Abstract
 Introduction
 Materials and Methods
 Results
 Discussion
 References
 

  1. Baron R, Neff L, Louvard D, Courtoy PJ 1985 Cell-mediated extracellular acidification and bone resorption: evidence for a low pH in resorbing lacunae and localization of a 100-kD lysosomal membrane protein at the osteoclast ruffled border. J Cell Biol 101:2210–2222[Abstract/Free Full Text]
  2. Zimolo Z, Wesolowski G, Rodan GA 1995 Acid extrusion is induced by osteoclast attachment to bone. Inhibition by alendronate and calcitonin. J Clin Invest 96:2277–2283[Medline]
  3. Rousselle AV, Heymann D 2002 Osteoclastic acidification pathways during bone resorption. Bone 30:533–540[Medline]
  4. Bastani B, Ross FP, Kopito RR, Gluck SL 1996 Immunocytochemical localization of vacuolar H+-ATPase and Cl–-HCO–3 anion exchanger (erythrocyte band-3 protein) in avian osteoclasts: effect of calcium-deficient diet on polar expression of the H+-ATPase pump. Calcif Tissue Int 58:332–336[Medline]
  5. Gay CV, Ito MB, Schraer H 1983 Carbonic anhydrase activity in isolated osteoclasts. Met Bone Dis Relat Res 5:33–39
  6. Hunter SJ, Rosen CJ, Gay CV 1991 In vitro resorptive activity of isolated chick osteoclasts: effects of carbonic anhydrase inhibition. J Bone Miner Res 6:61–66[Medline]
  7. Teti A, Blair HC, Teitelbaum SL, Kahn AJ, Koziol C, Konsek J, Zambonin-Zallone A, Schlesinger PH 1989 Cytoplasmic pH regulation and chloride bicarbonate exchange in avian osteoclasts. J Clin Invest 83:227–233[Medline]
  8. Nordström T, Rotstein OD, Romanek R, Asotra S, Heersche JNM, Manolson MF, Brisseau GF, Grinstein S 1995 Regulation of cytosolic pH in osteoclasts. Contribution of proton pumps and a proton-selective conductance. J Biol Chem 270:2203–2212[Abstract/Free Full Text]
  9. Ravesloot JH, Eisen T, Baron R, Boron WF 1995 Role of Na-H exchangers and vacuolar H+ pumps in intracellular pH regulation in neonatal rat osteoclasts. J Gen Physiol 105:177–208[Abstract/Free Full Text]
  10. Lees RL, Heersche JN 2000 Differences in regulation of pHi in large (>10 nuclei) and small (<5 nuclei) osteoclasts. Am J Physiol Cell Physiol 279:C751–C761
  11. Wiktor-Jedrzejczak W, Bartocci A, Ferrante Jr AW, Ahmed-Ansari A, Sell KW, Pollard JW, Stanley ER 1990 Total absence of colony-stimulating factor 1 in the macrophage-deficient osteopetrotic (op/op) mouse. Proc Natl Acad Sci USA 87:4828–4832[Abstract/Free Full Text]
  12. Fuller K, Owens JM, Jagger CJ, Wilson A, Moss R, Chambers TJ 1993 Macrophage colony-stimulating factor stimulates survival and chemotactic behavior in isolated osteoclasts. J Exp Med 178:1733–1744[Abstract/Free Full Text]
  13. Tanaka S, Takahashi N, Udagawa N, Tamura T, Akatsu T, Stanley ER, Kurokawa T, Suda T 1993 Macrophage colony-stimulating factor is indispensable for both proliferation and differentiation of osteoclast progenitors. J Clin Invest 91:257–263[Medline]
  14. Suda T, Udagawa N, Nakamura I, Miyaura C, Takahashi N 1995 Modulation of osteoclast differentiation by local factors. Bone 17:87S–91S
  15. Akiyama T, Bouillet P, Miyazaki T, Kadono Y, Chikuda H, Chung UI, Fukuda A, Hikita A, Seto H, Okada T, Inaba T, Sanjay A, Baron R, Kawaguchi H, Oda H, Nakamura K, Strasser A, Tanaka S 2003 Regulation of osteoclast apoptosis by ubiquitylation of proapoptotic BH3-only Bcl-2 family member Bim. EMBO J 22:6653–6664[CrossRef][Medline]
  16. Woo KM, Kim HM, Ko JS 2002 Macrophage colony-stimulating factor promotes the survival of osteoclast precursors by up-regulating Bcl-XL. Exp Mol Med 34:340–346[Medline]
  17. Zecchi-Orlandini S, Formigli L, Tani A, Benvenuti S, Fiorelli G, Papucci L, Capaccioli S, Orlandini GE, Brandi ML 1999 17ß-Estradiol induces apoptosis in the preosteoclastic FLG 29.1 cell line. Biochem Biophys Res Commun 255:680–685[CrossRef][Medline]
  18. Okahashi N, Koide M, Jimi E, Suda T, Nishihara T 1998 Caspases (interleukin-1ß-converting enzyme family proteases) are involved in the regulation of the survival of osteoclasts. Bone 23:33–41[Medline]
  19. Segal MS, Beem E 2001 Effect of pH, ionic charge, and osmolality on cytochrome c-mediated caspase-3 activity. Am J Physiol Cell Physiol 281:C1196–C1204
  20. Insogna KL, Sahni M, Grey AB, Tanaka S, Horne WC, Neff L, Mitnick M, Levy JB, Baron R 1997 Colony-stimulating factor-1 induces cytoskeletal reorganization and c-src-dependent tyrosine phosphorylation of selected cellular proteins in rodent osteoclasts. J Clin Invest 100:2476–2485[Medline]
  21. Suda T, Jimi E, Nakamura I, Takahashi N 1997 Role of 1 {alpha},25-dihydroxyvitamin D3 in osteoclast differentiation and function. Methods Enzymol 282:223–235[Medline]
  22. Yao GQ, Sun B, Hammond EE, Spencer EN, Horowitz MC, Insogna KL, Weir EC 1998 The cell-surface form of colony-stimulating factor-1 is regulated by osteotropic agents and supports formation of multinucleated osteoclast-like cells. J Biol Chem 273:4119–4128[Abstract/Free Full Text]
  23. Bouyer P, Zhou Y, Boron WF 2003 An increase in intracellular calcium concentration that is induced by basolateral CO2 in rabbit renal proximal tubule. Am J Physiol Renal Physiol 285:F674–F687
  24. Zhao J, Zhou Y, Boron WF 2003 Effect of isolated removal of either basolateral HCO3– or basolateral CO2 on HCO3– reabsorption by rabbit S2 proximal tubule. Am J Physiol Renal Physiol 285:F359–F369
  25. Bevensee MO, Bashi E, Boron WF 1999 Effect of trace levels of nigericin on intracellular pH and acid-base transport in rat renal mesangial cells. J Membr Biol 169:131–139[CrossRef][Medline]
  26. Thomas JA, Buchsbaum RN, Zimniak A, Racker E 1979 Intracellular pH measurements in Ehrlich ascites tumor cells utilizing spectroscopic probes generated in situ. Biochemistry 18:2210–2218[CrossRef][Medline]
  27. Boyarsky G, Ganz MB, Sterzel B, Boron WF 1988 pH regulation in single glomerular mesangial cells. I. Acid extrusion in absence and presence of HCO–3. Am J Physiol 255:C844–C856
  28. Damkier HH, Nielsen S, Praetorius J 2006 An anti-NH2-terminal antibody localizes NBCn1 to heart endothelia and skeletal and vascular smooth muscle cells. Am J Physiol Heart Circ Physiol 290:H172–H180
  29. Rickers A, Brockstedt E, Mapara MY, Otto A, Dorken B, Bommert K 1998 Inhibition of CPP32 blocks surface IgM-mediated apoptosis and D4-GDI cleavage in human BL60 Burkitt lymphoma cells. Eur J Immunol 28:296–304[CrossRef][Medline]
  30. Gregoli PA, Bondurant MC 1999 Function of caspases in regulating apoptosis caused by erythropoietin deprivation in erythroid progenitors. J Cell Physiol 178:133–143[CrossRef][Medline]
  31. Choi I, Aalkjær C, Boulpaep EL, Boron WF 2000 An electroneutral sodium/bicarbonate cotransporter NBCn1 and associated sodium channel. Nature 405:571–575[CrossRef][Medline]
  32. Pushkin A, Abuladze N, Lee I, Newman D, Hwang J, Kurtz I 1999 Cloning, tissue distribution, genomic organization, and functional characterization of NBC3, a new member of the sodium bicarbonate cotransporter family. J Biol Chem 274:16569–16575[Abstract/Free Full Text]
  33. Zhao J, Hogan EM, Bevensee MO, Boron WF 1995 Out-of-equilibrium CO2/HCO–3 solutions and their use in characterizing a new K/HCO3 cotransporter. Nature 374:636–639[CrossRef][Medline]
  34. Zhou Y, Zhao J, Bouyer P, Boron WF 2005 Evidence from renal proximal tubules that HCO–3 and solute reabsorption are acutely regulated not by pH but by basolateral HCO–3 and CO2. Proc Natl Acad Sci USA 102:3875–3880[Abstract/Free Full Text]
  35. Frelin C, Vigne P, Barbry P, Lazdunski M 1987 Molecular properties of amiloride action and of its Na+ transporting targets. Kidney Int 32:785–793[Medline]
  36. Dröse S, Bindseil KU, Bowman EJ, Siebers A, Zeeck A, Altendorf K 1993 Inhibitory effect of modified bafilomycins and concanamycins on P- and V-type adenosinetriphosphatases. Biochemistry 32:3902–3906[CrossRef][Medline]
  37. Pérez-Sala D, Collado-Escobar D, Mollinedo F 1995 Intracellular alkalinization suppresses lovastatin-induced apoptosis in HL-60 cells through the inactivation of a pH-dependent endonuclease. J Biol Chem 270:6235–6242[Abstract/Free Full Text]
  38. Matsuyama S, Reed JC 2000 Mitochondria-dependent apoptosis and cellular pH regulation. Cell Death Differ 7:1155–1165[CrossRef][Medline]
  39. Furlong IJ, Ascaso R, Lopez Rivas A, Collins MK 1997 Intracellular acidification induces apoptosis by stimulating ICE-like protease activity. J Cell Sci 110:653–661[Abstract]
  40. Filomenko R, Prévotat L, Rébé C, Cortier M, Jeannin JF, Solary E, Bettaieb A 2006 Caspase-10 involvement in cytotoxic drug-induced apoptosis of tumor cells. Oncogene 25:7635–7645[CrossRef][Medline]
  41. Plotkin LI, Manolagas SC, Bellido T 2006 Dissociation of the pro-apoptotic effects of bisphosphonates on osteoclasts from their anti-apoptotic effects on osteoblasts/osteocytes with novel analogs. Bone 39:443–452[Medline]
  42. Szymczyk KH, Shapiro IM, Adams CS 2004 Ionizing radiation sensitizes bone cells to apoptosis. Bone 34:148–156[Medline]
  43. Gottlieb RA, Giesing HA, Zhu JY, Engler RL, Babior BM 1995 Cell acidification in apoptosis: granulocyte colony-stimulating factor delays programmed cell death in neutrophils by up-regulating the vacuolar H+-ATPase. Proc Natl Acad Sci USA 92:5965–5968[Abstract/Free Full Text]
  44. Okahashi N, Nakamura I, Jimi E, Koide M, Suda T, Nishihara T 1997 Specific inhibitors of vacuolar H+-ATPase trigger apoptotic cell death of osteoclasts. J Bone Miner Res 12:1116–1123[CrossRef][Medline]
  45. Barrière H, Poujeol C, Tauc M, Blasi JM, Counillon L, Poujeol P 2001 CFTR modulates programmed cell death by decreasing intracellular pH in Chinese hamster lung fibroblasts. Am J Physiol Cell Physiol 281:C810–C824
  46. Schneider D, Gerhardt E, Bock J, Müller MM, Wolburg H, Lang F, Schulz JB 2004 Intracellular acidification by inhibition of the Na+/H+-exchanger leads to caspase-independent death of cerebellar granule neurons resembling paraptosis. Cell Death Differ 11:760–770