Osteoclasts are multinucleated cells of hematopoietic origin that are formed by fusion of mononuclear precursors.1 In response to proresorption factors, osteoblasts and stromal cells release receptor activator of nuclear factor κB ligand (RANKL), which binds to its cognate receptor RANK on the surface of mononuclear precursor cells leading to activation of a number of key transcription factors, such as nuclear factor κB (NF-κB), activator protein 1 (AP-1), and nuclear factor of activated T cells c1 (NFATc1). NFATc1 is the transcription factor regulating expression of osteoclast-specific genes, such as tartrate-resistant acid phosphatase (TRAP), cathepsin K, calcitonin receptor, dendritic cell-specific transmembrane protein (DC-STAMP), osteoclast associated receptor (OSCAR), and the integrin αvβ3. NFATc1 activation, in turn, is controlled by Ca2+-activated phosphatase calcineurin, which dephosphorylates NFATc1, allowing it to translocate into the nucleus and regulate gene expression.2
Another molecule highly expressed in osteoclasts is the proton pump vacuolar H+-ATPase (V-ATPase).3, 4 This enzyme is responsible for acidification of the resorption lacunae and demineralization of the bone surface; by creating a low pH environment, it also provides conditions optimal for tissue-degrading enzymes such as cathepsin K.5 V-ATPases are ubiquitously expressed, multisubunit proton pumps involved in membrane trafficking, protein degradation and vesicular transport, and are necessary for acidification of intracellular (e.g., endosomes, lysosomes) and extracellular (e.g., renal acidification, bone resorption) compartments.6 V-ATPases consist of at least 14 subunits organized into V0 (proton translocation) and V1 (ATP hydrolysis) domains. The “a” subunit has four isoforms (a1–a4) that are tissue and cell type specific, with the a3 isoform highly enriched in osteoclasts,3, 7 where it localizes to lysosomes and plasma membranes in nonresorbing cells8 and to the ruffled border in actively resorbing osteoclasts. The importance of a3 in osteoclasts is further confirmed by the finding that human mutations in this subunit lead to infantile malignant osteopetrosis, due to inability of osteoclasts to acidify the bone surface and, therefore, an inability to resorb bone.9
We previously demonstrated that heterozygous mice with a point mutation (R740S) in the a3 subunit of V-ATPase (+/R740S) have mild osteopetrosis and impaired bone resorption,10 even though the number of osteoclasts in +/R740S bones is increased, a finding consistent with that seen in Tcirg1−/− (a3 knockout)7 or oc/oc (a3 truncation) mouse models.11 We also showed that proton translocation was inhibited by almost 90% in membranes from +/R740S osteoclasts. Arginine at the 740 position in mammals (corresponding to R735 position in yeast) is the evolutionarily conserved amino acid responsible for proton translocation.12 Here we present evidence that osteoclasts carrying the R740S mutation in the a3 subunit of V-ATPase have a higher lysosomal pH compared with +/+ cells, resulting in decreased NFATc1 nuclear translocation and impaired osteoclastogenesis in vitro.
Materials and Methods
Heterozygous mice carrying the R740S mutation bred onto the C3H/HeJ and FVB backgrounds were generated as described previously10 and are referred to as +/R740S mice. All experimental procedures received approval from the local Animal Care Committees and were conducted in accordance with the guidelines of the Canadian Council on Animal Care.
Bone marrow (BM)-derived mononuclear cells from femurs of two month old male mice were plated at a cell density of 1 × 105 cells/mL in 10% fetal bovine serum (FBS) in alpha-MEM and antibiotics (10 µg/mL penicillin G, 50 µg/mL gentamycin, 0.03 µg/mL fungizone) supplemented with 50 ng/mL M-CSF (Calbiochem, EMD Biosciences, Burlington, ON, Canada) for 2 days, and then with 50 ng/mL M-CSF and 200 ng/mL RANKL (made in-house) for another 4 days at 37°C in 5% CO2. For tartrate-resistant acid phosphatase (TRAP) staining, cells were washed, briefly fixed in formalin, and stained with a solution consisting of 50 mM acetate buffer, 30 mM sodium tartrate, 0.1 mg/mL naphthol AS-MX phosphate, 0.1% Triton X-100, and 0.3 mg/mL Fast Red Violet LB for 10 to 20 minutes, then washed three times with dH2O.
BM-derived osteoclasts were generated as described above. At day 5 of culture, cells were loaded with lysine-fixable dextran-FITC (10,000 MW; 0.25 mg/mL) overnight and chased for 3 hours. Cells were fixed in 4% paraformaldehyde (PFA) in PBS, permeabilized using 0.1% saponin, blocked with 1% normal goat serum (NGS) and 1% bovine serum albumin (BSA), probed with anti-LAMP2 (ABL-93, Developmental Studies Hybridoma state bank, University of Iowa, Iowa City, IA, USA) or anti-a3 (rabbit polyclonal antibody, custom made by Cedarlane Laboratories (Burlington, ON, Canada) using RNTQRRPAGQQDEDTDKLLASPDASTLEN peptide) primary antibody and AF568-conjugated secondary antibody. The cells were visualized with a 63 × 1.4 NA oil-immersion lens using a spinning-disk confocal microscope (Quorum Technologies Inc., Guelph, ON, Canada) equipped with a back-thinned EM-CCD camera (C9100-13, Hamamatsu, Quorum Technologies Inc.); images were acquired by and processed with Volocity software (Perkin-Elmer, Waltham, MA, USA). Images acquired throughout the entire cell thickness with 0.5-mm intervals were assembled by maximum-intensity projection for illustration purposes.
Determination of lysosomal pH
BM-derived osteoclasts were cultured on glass coverslips as described above. On day 5 of culture, cells were washed with phenol red free α-MEM medium (Invitrogen, Burlington, ON, Canada) and incubated overnight with pH-sensitive dextran-FITC (Invitrogen, 10,000 MW; 0.5 mg/mL) in 10% FBS in phenol red free α-MEM containing 25 ng/mL M-CSF and 50 ng/mL RANKL. Dextran-FITC was chased for 3 hours in 10% FBS in phenol red free α-MEM, and pH was measured by fluorescence ratiometric imaging as previously described.13, 14 Images were acquired at 30-second intervals on a Leica DM IRB microscope equipped with a 40 × 1.25 NA oil immersion objective and an EM-CCD camera (Cascade II, Photometrics), using MetaFluor software (MDS Analytical Technologies, Quorum Technologies, Guelph, ON, Canada). At the end of each experiment, for every cell, in situ calibration was performed using isotonic K+ buffer solutions (145 mM KCl, 10 mM glucose, 1 mM MgCl2, 20 mM Hepes, pH ranging from 4.0 to 7.0) containing 10 µg/mL nigericin. Fluorescence intensity ratio (485 nm/438 nm) corrected for background was used to interpolate pH values from the experimental data. Four independent experiments were performed, two cells/genotype (only cells with three or more nuclei were used) and eight regions of interest (ROI)/cell were measured in each experiment.
BM-derived osteoclasts were plated into 96-well plates and cultured as described above. On day 5 of culture, medium was replaced with one containing 25 ng/mL M-CSF, 50 ng/mL RANKL, and various concentrations of lysosomal inhibitors (ammonium chloride and chloroquine [Sigma Aldrich, St. Louis, MO, USA]) and the V-ATPase inhibitor concanamycin A (MP Biomedicals, Solon, OH, USA). Cells were incubated for 24 hours and stained for TRAP as described above. TRAP-positive cells containing five or more nuclei were counted. The experiments were performed three times, with n = 4 per group.
BM-derived osteoclasts were plated on eight-well chamber slides (BD Falcon, Mississauga, ON, Canada) and cultured as described above. On day 6, cells were fixed in 3.7% PFA, rinsed in PBS, and used for colorimetric detection of apoptotic cells with the fragEL DNA Fragmentation Kit according to the manufacturer's directions (Calbiochem, EMD Biosciences). The total number of osteoclasts and number of osteoclasts with apoptotic nuclei were counted. The experiments were performed three times.
BM-derived osteoclasts were cultured as described above. On day 5 of culture, the cells were scraped, counted, and replated onto dentine slices at 1 × 104 cells/well. After 48 hours, cells were fixed, stained for TRAP, and images were acquired with a Leica DM-IRE2 microscope. Cells were then removed using sonication, and the dentine was stained with toluidine blue to identify resorption pits. Images of the dentine were taken using a Leica DM-IRE2 microscope; the resorption area was quantified with ImageJ software. The experiments were performed three times.
Total RNA was extracted at days 4 and 6 from osteoclast cultures using TRIzol reagent (Invitrogen, Burlington, ON, Canada). For PCR, total RNA (3 µg) was treated with DNase I (Invitrogen) at 1 unit of DNase I/µg RNA and then reverse transcribed using Revert Aid H− First Strand Kit (Fermentas, Burlington, ON, Canada). PCR reactions were performed in 50-µL volumes using HotStarTaq polymerase (Qiagen, Toronto, ON, Canada). PCR products were visualized using GeneSnap software version 4.00.00 (SynGene, Cambridge, UK) and data were normalized to GAPDH. For quantitative PCR (qPCR) gene expression analysis, 1µg of RNA was treated with DNase I (Invitrogen), and cDNA was synthesized using Maxima Reverse Transcriptase (Fermentas) according to the manufacturer's instructions. qPCR was performed using the Quanti-Tect SYBR Green PCR Kit (Qiagen) in iQ5 BioRad detection system (BioRad, Mississauga, ON, Canada). All data were normalized to expression of L32, a ribosomal protein. Specific primer sequences used in the reactions are summarized in Supplemental Table S1. All experiments were repeated at least three times.
BM-derived osteoclasts were cultured in 100-mm tissue culture dishes as described earlier. On days 4 or 6, plates were washed two times with cold PBS and lysed in lysis buffer (0.1% Triton X-100, 50 mM Tris, 300 mM NaCl, 5 mM EDTA), containing protease inhibitors (protease inhibitors cocktail Sigma P8340 and 1 mM PMSF), and phosphatase inhibitors (Sigma P5726). Protein concentration was determined using Pierce 660 nm Protein Assay Reagent (Thermo Scientific, Nepean, ON, Canada). Whole cell lysates (25 µg protein) were separated on 10% SDS-PAGE, transferred to nitrocellulose membrane, and probed for a3 (rabbit polyclonal antibody, custom made by Cedarlane Laboratories using RNTQRRPAGQQDEDTDKLLASPDASTLEN peptide), E (from Dr Beth Lee, Ohio State University, Columbus, OH, USA), d2 (from Dr G Tremblay, Alethia Biotherapeutics, Montreal, Canada) subunits of V-ATPase, NFATc1 (Santa Cruz, Santa Cruz, CA, USA), RCAN1 (Aviva Systems Biology, San Diego, CA, USA), calcineurin A (Abcam, Cambridge, MA, USA), DYRK1A (Santa Cruz), and actin (Cell Signalling, Pickering, ON, Canada). Images were captured using BioRad ChemiDoc Gel Docking system and analyzed by rolling disk method using Quantity One software (BioRad, Mississauga, ON, Canada). All experiments were repeated three times.
NFATc1 nuclear translocation
BM mononuclear cells were plated onto eight-well glass chamber slides (BD Falcon) and cultured as described above. On day 6, the cells were fixed in 4% PFA in PBS, permeabilized using 0.1% Triton X-100 in PBS for 10 minutes, blocked in 1% NGS for 1 hour, stained with anti-NFATc1 antibody (Santa Cruz) overnight, and rabbit anti-mouse-AF568 secondary antibody (Invitrogen) for 1 hour; slides were mounted using ProLong Gold mounting media with DAPI (Invitrogen). Cells were visualized with a 40X objective, and images were acquired on a Leica DM IRE2 microscope. Experiments were repeated three times. To quantify NFATc1 nuclear translocation, 50 images were taken of each group per experiment, and absence or presence of NFATc1 nuclear translocation was determined in 61 ± 10 cells/group.
Calcineurin activity assay
Osteoclasts were cultured in 100-mm dishes as described above, and calcineurin cellular activity was measured using a Calcineurin Cellular Activity Assay Kit (Enzo Life Sciences, Cedarlane, Burlington, ON, Canada) following the manufacturer's instructions. Briefly, on day 6 of culture, cells were washed twice with ice-cold Tris-buffered saline (TBS; 20 mM Tris pH 7.2, 150 mM NaCl), scraped into 300 µl/dish of lysis buffer provided with the kit, briefly vortexed, and centrifuged at 100,000g for 45 minutes at 4°C. Supernatants were collected, desalted, and analyzed for phosphatase activity by measuring free-phosphate release. Okadaic acid (OA; inhibits PP1 and PP2A) and OA + EGTA (inhibits PP1, PP2A, and PP2B) were added to selected wells to distinguish activities of different phosphatases. Calcineurin (PP2B) activity was determined as PP2B = OA − (OA +EGTA), corrected by background. Results were normalized by total protein, measured using Pierce 660 nm Protein Assay (Thermo Scientific, Nepean, ON, Canada) and presented as nmol PO4/µg protein. Three independent experiments were performed.
Osteoclasts were cultured in 12-well plates as described above. On day 6 of culture, medium was replaced with one containing lysosomal inhibitors (5 mM ammonium chloride or 10 µM chloroquine) or proteasomal inhibitor (1 µM MG-132). Cells were incubated for 2 hours at 37°C, 10% CO2 (an extracellular acidosis condition, known to activate NFATc1 signaling).15 Cells were then washed with ice-cold PBS, lysed, and processed as described above. Whole cell lysates were separated on 8% SDS-PAGE, transferred to nitrocellulose membrane, and probed for RCAN1 and the housekeeping control, actin. Images were captured using BioRad ChemiDoc Gel Docking system and analyzed by rolling disk method using Quantity One software (BioRad). Three independent experiments were performed.
Results are plotted as the mean ± SD of independent experiments; sample size is as indicated in the methods and figures. Statistical analysis was performed using a two-tailed Student's t test, and p values were considered significant at p < 0.05.
Lysosomal pH is increased in +/R740S osteoclasts
In nonresorbing osteoclasts, V-ATPases containing the a3 subunit are localized to the lysosomes8 and are thought to be responsible for maintenance of lysosomal pH. To address whether the +/R740S mutation affected lysosomal pH in osteoclasts, we used pH-sensitive dextran-FITC. Immunostaining with antibodies specific to a3 and LAMP2, a lysosomal marker, showed an extensive colocalization between dextran and both a3 and LAMP2 (Pearson's coefficient 0.83 ± 0.09 and 0.87 ± 0.06, respectively), confirming that dextran-FITC localized to lysosomes (Fig. 1A). Fluorescence ratiometric imaging with in situ calibration demonstrated significantly higher lysosomal pH in +/R740S osteoclasts compared with +/+ osteoclasts, with pH measurements of 5.71 ± 0.44 versus 4.85 ± 0.18, respectively (Fig. 1B), indicating that the R740S mutation impairs V-ATPase function not only at the plasma membrane but also in the intracellular lysosomal compartment.
In vitro osteoclastogenesis is impaired in +/R740S cells
+/R740S BM mononuclear cells cultured for 6 days in the presence of M-CSF and RANKL formed smaller and fewer osteoclasts than +/+ cells (Fig. 2A, B). This decrease in osteoclast number was not the result of increased apoptosis of +/R740S compared with +/+ cells (Fig. 2C). Consistent with decreased osteoclastogenesis, bone resorption was also significantly reduced in +/R740S osteoclast cultures (Fig. 2D, E). Semiquantitative PCR analysis showed that, while there was no difference in gene expression at day 4, by day 6, the levels of all key osteoclast markers, such as TRAP, cathepsin K, calcitonin receptor, DC-STAMP, OSCAR, NFATc1, and the d2 subunit of the V-ATPase, were decreased in +/R740S compared with +/+ cells (Fig. 3A, B); a3 gene expression was not affected. Immunoblotting on whole cell lysates confirmed that the levels of d2 and NFATc1 proteins were decreased in +/R740S compared with +/+ cells, whereas a3 and the ubiquitous V-ATPase E subunit protein levels were not affected (Fig. 3C–E). These results indicate that higher lysosomal pH impairs in vitro osteoclastogenesis, specifically at the later stage of osteoclast formation/fusion. To confirm that lysosomal pH is in fact important for the later stages of osteoclast formation, we treated +/+ osteoclasts with two different lysosomal inhibitors, ammonium chloride and chloroquine, as well as with a V-ATPase inhibitor, concanamycin A, during the last 24 hours of culture. All three inhibitors decreased the number of large (5+ nuclei) osteoclasts in +/+ cells (Fig. 3F, G). Interestingly, none of the inhibitors had any significant effect on osteoclast formation in +/R740S cultures at concentrations that also had no effect on cell viability (Fig. 3F, G, Supplemental Fig. S1), suggesting that a certain lysosomal pH threshold is necessary for late stages of osteoclast formation.
NFATc1 nuclear translocation is decreased in +/R740S osteoclasts
Next, we decided to elucidate the molecular mechanism of impaired osteoclastogenesis in +/R740S cells. Gene expression of osteoclast markers, such as TRAP,16 cathepsin K,17 DC-STAMP, OSCAR,18 NFATc1,19 and the d2 subunit of the V-ATPase, is under NFATc1 transcriptional control. Therefore, we looked at NFATc1 activation in +/+ and +/R740S osteoclasts by immunofluorescence. A wide range of NFATc1 nuclear translocation phenotypes was observed in both +/R740S and +/+ cells (Fig. 4A), varying from osteoclasts with no nuclear translocation to those with all nuclei positive for NFATc1. On average, 53% of +/+ osteoclasts had all nuclei positive for NFATc1 (100% of nuclei), 43% were in the intermediate state, and 4% had no nuclear NFATc1 (0% of nuclei). In comparison, the numbers for +/R740S osteoclasts were 35%, 49%, and 16%, respectively. To simplify quantification, we focused only on osteoclasts with 0% of nuclei or 100% of nuclei displaying NFATc1 translocation. To combine results of three independent experiments, the values for cells with all nuclei positive for NFATc1 (100% of nuclei) were expressed as percent of +/+ cells. +/R740S cultures had significantly fewer osteoclasts with 100% NFATc1 nuclear translocation compared with +/+ cells (Fig. 4B). At the same time, the number of +/R740S osteoclasts displaying no nuclear NFATc1 (0% of nuclei) was significantly increased compared with +/+ cells, suggesting that NFATc1 activation is impaired in +/R740S cells.
Calcineurin enzymatic activity is not affected in +/R740S osteoclasts
NFATc1 activation is a tightly regulated process.20 Calcineurin is the phosphatase responsible for dephosphorylation of NFATc1.21 We detected no significant difference in calcineurin phosphatase activity (Fig. 5A) or calcineurin Aα protein levels (Fig. 5B) in +/R740S versus +/+ osteoclasts at day 6 of culture. Protein expression of dual-specificity tyrosine-phosphorylation-regulated kinase 1A (DYRK1A), one of the kinases responsible for phosphorylation of NFATc1,22, 23 was also not affected in +/R740S osteoclasts (Fig. 5B).
RCAN1 is the factor responsible for NFATc1 inhibition
To elucidate the mechanism of NFATc1 inhibition, we investigated regulator of calcineurin 1 (RCAN1), an endogenous NFATc1 inhibitor and one of the genes dramatically upregulated during osteoclastogenesis.24 It is also associated with the craniofacial and skeletal anomalies seen in Down syndrome.22, 25 Both +/R740S and +/+ osteoclasts expressed Rcan1, as determined by qPCR (Fig. 6A). Because Rcan1 is a target gene of NFATc1, we expected Rcan1 mRNA levels to be decreased in +/R740S cells, similarly to other NFATc1 target genes; however, the expression levels were not significantly different between genotypes. Immunoblotting confirmed that RCAN1 protein was present in bone marrow-derived osteoclasts (Fig. 6B) and that there was no significant differences in protein expression levels between +/R740S and +/+ osteoclasts (Fig. 6C). Furthermore, the RCAN1/NFATc1 ratio was significantly higher in +/R740S compared with +/+ cells (Fig. 6D). Therefore, we hypothesized that RCAN1 is the factor responsible for NFATc1 downregulation. It has been demonstrated that in HEK293 cells, during calcineurin activation, RCAN1 is rapidly degraded in the lysosome,26 and an alkalinization of lysosomal pH by chloroquine increases RCAN1 protein levels and decreases NFAT-calcineurin activity.26 To confirm whether a similar mechanism occurs in osteoclasts, we treated +/+ osteoclasts with lysosomal inhibitors (5 mM ammonium chloride or 10 µM chloroquine), or proteasomal inhibitor (1 µM MG-132; used as a positive control to show inhibition of protein degradation), incubated for 2 hours at 37°C 10% CO2 (extracellular acidosis) to activate NFATc1 signaling,15 and assessed RCAN1 degradation by immunoblotting. Results demonstrated that +/+ osteoclasts exposed to chloroquine and MG-132 had significantly higher RCAN1 protein levels compared with control (Fig. 6E, F). Ammonium chloride also increased RCAN1 levels, but this increase was not statistically significant. These findings confirm published literature26 and indicate that in osteoclasts higher lysosomal pH leads to RCAN1 accumulation and, potentially, to NFATc1 inhibition.
Here we demonstrate that the R740S mutation in the a3 subunit of V-ATPase causes an increase in lysosomal pH and a cell autonomous impairment of osteoclastogenesis in vitro because of decreased NFATc1 activation.
Heterozygous mice with a point mutation in the a3 subunit of V-ATPase (+/R740S) display mild osteopetrosis, even though the number of osteoclasts in bones of +/R740S mice is increased.10 This is similar to Tcirg1−/− (a3 knockout)7 and to oc/oc (a3 truncation)11 mouse models, all of which display reduced proton translocation at the plasma membrane into the resorption lacunae. This increase in osteoclast numbers in vivo is most likely a compensatory response to the increased bone density or decreased resorption capacity of the osteoclasts, and because of the higher expression of Rankl and Csf1 (M-CSF) in +/R740S bones,10 consistent with what is seen in oc/oc mice.27 In contrast, we show that in vitro osteoclastogenesis is impaired, resulting in smaller and fewer osteoclasts in +/R740S cells compared with +/+ cells, a phenomenon not observed in the other murine models. In addition, we found that the presence of both a wild-type and an R740S a3 protein, which does not detectably alter expression levels or targeting of the a3 protein to appropriate cellular sites, impairs lysosomal proton pumping and acidification (Fig. 1B). This finding is different from that observed in the Tcirg1−/− model,7 where it was reported that there was no abrogation in osteoclast lysosomal acidification as determined by acridine orange staining of osteoclasts. This discrepancy suggests that the complete absence of a3 elicits a compensatory response in which other “a” subunits are able to maintain lysosomal pH but not plasma membrane proton translocation. A similar situation has been observed in yeast where the two “a” isoforms, Vph1p and Stv1p, can compensate for each other in some but not all cellular compartments.28 In mammalian cells, it has been shown that a1-containing V-ATPases are also present in the phagolysosomes, with the ratio of a3/a1 much higher in osteoclasts compared with THP.1 monocytic cells.29 Therefore, it is possible that a1-containing V-ATPases maintain the wild-type pH in the Tcirg1−/− osteoclast lysosomes. Notably, the fact that in vitro osteoclastogenesis was not affected in either Tcirg1−/− or oc/oc models but is impaired in +/R740 cells indicates that low lysosomal pH is necessary for osteoclast differentiation. To confirm the importance of lysosomal pH in osteoclastogenesis, we treated the cells with various inhibitors, all commonly used to inhibit lysosomal function.30 Ammonium chloride is a weak base known to neutralize the pH of acidic compartments, chloroquine is a lysosomotropic drug that raises lysosomal pH, whereas concanamycin A inhibits V-ATPase function by binding directly to the V0 domain, also resulting in elevation of lysosomal pH. Others31 have shown that treatment with ammonium chloride and concanamycin A raises pH of acidic secretory granules to 7.15 and 7.27, respectively. All three inhibitors decreased the number of large (5+ nuclei) osteoclasts formed; at the same time, none of the inhibitors had any effect on +/R740S osteoclasts, suggesting that a certain lysosomal pH threshold is necessary for the late stages of osteoclast formation (and possibly fusion). Our results also suggest that lysosomes may serve multiple yet unknown functions in osteoclasts, besides the well-studied roles, such as processing and trafficking of acid hydrolases.32
Amongst novel roles, our data indicate that lysosomes play a role in regulation of NFATc1 activation in osteoclasts. The effect of extracellular pH on NFATc1 activation in osteoclasts has been shown previously;15 however, the effect of intracellular, particularly lysosomal, acidification on NFAT signaling has not been reported. We found no evidence of a change in the activity or protein expression level of calcineurin A, the phosphatase responsible for NFATc1 dephosphorylation, or DYRK1A, one of the kinases responsible for phosphorylation of NFATc1, in spite of the significant diminution in NFATc1 nuclear translocation. Amongst several molecules known to inhibit NFATc1,33–35 RCAN1, also known as Down syndrome critical region 1 (DSCR1), has been shown in neurons and in muscle cells to physically interact with calcineurin A subunit and prevent NFATc1 dephosphorylation.22, 34Rcan1 is also an NFATc1 target gene, resulting in an NFATc1–RCAN1 negative feedback loop.35, 36 Moreover, during osteoclastogenesis, Rcan1 expression levels dramatically increase,24 and Rcan1 is believed to be one of the genes associated with the craniofacial and skeletal defects present in Down syndrome.22, 25 Coincidentally, it has been demonstrated that in HEK293 cells, during calcineurin activation, RCAN1 is rapidly degraded in the lysosome,26 and an alkalinization of lysosomal pH by chloroquine increases RCAN1 protein levels and decreases NFAT-calcineurin activity.26 We found that bone marrow-derived osteoclasts express RCAN1, and gene and protein levels were, somewhat unexpectedly given the reduction in NFATc1 expression, similar in +/+ and +/R740S cells; however, relative to NFATc1 protein (Fig. 6D), +/R740S cells do express higher levels of RCAN1. Therefore, we proposed that the increased lysosomal pH in +/R740S cells abrogates RCAN1 degradation, leading to relatively higher levels of RCAN1 and inhibition of NFATc1 nuclear translocation. Consistent with this hypothesis, treatment with chloroquine, one of the lysosomal inhibitors known to elevate lysosomal pH, increased RCAN1 levels in +/+ cells (Fig. 6E, F). Unexpectedly, another lysosomal inhibitor, ammonium chloride, elicited only a nonsignificant increase in RCAN1 protein levels in +/+ cells. This may be because of differences in the mechanisms of action of ammonium chloride versus chloroquine that may necessitate different treatment durations or other changes in culture conditions. Furthermore, even though the RCAN1 levels are relatively higher in +/R740S osteoclasts, we did not observe inhibition of calcineurin A, possibly because of the nature of the assay, i.e., detergent present in the lysis buffer may result in dissociation of RCAN1 and calcineurin, abrogating potential changes in activity.
Our data indicate that osteoclasts from heterozygous mice with an R740S mutation in the a3 subunit of V-ATPase have an elevated lysosomal pH and a cell autonomous impairment of osteoclastogenesis in vitro because of a decrease in NFATc1 nuclear translocation. The precise roles of the lysosomal pH in osteoclastogenesis are still not known. One of the mechanisms, as shown in this article, is inhibition of NFATc1 signaling via RCAN1 accumulation. It is possible that other signaling pathways, e.g., the Wnt signaling pathway,37 are also affected by elevated lysosomal pH, further contributing to impaired osteoclast formation in vitro.
These findings have important clinical implications—medications, such as the antimalarial lysosomotropic drug chloroquine, are used for treatment of rheumatic diseases.38 It has been shown that a small rise in pH of acidic compartments affects receptor recycling, protein processing, production of cytokines, and other immune mediators.38, 39 Here we show that deregulated NFATc1 signaling in osteoclasts is another pathway affected by changes in lysosomal pH.
All authors state that they have no conflicts of interest.
We thank Dr Sergio Grinstein (Hospital for Sick Children, Toronto) for helpful discussions. We also thank Keying Lee (Manolson Lab) for making the recombinant RANKL, Dr Yeqi Yao for technical assistance, and Dr Beth Lee (Ohio State University) for V-ATPase a3 and E antibodies.
This work was supported by grants from the Canadian Institutes of Health Research (FRN 69198 to JEA and FRN 79322 to MFM and JEA), as well as scholarship support from the Canadian Arthritis Network (IV), the CIHR Strategic Training Program Cell Signaling in Mucosal Inflammation and Pain (NO), the Faculty of Dentistry (University of Toronto) Harron Fund (NO), and Foundation Bettencourt Schueller (VJ).
Authors' roles: Study design: IV. Study conduct: IV, NO, VJ, and CO. Data collection: IV, NO, and VJ. Data analysis: IV, NO, and VJ. Data interpretation: IV and NO. Drafting manuscript: IV. Revising manuscript content: IV, MFM, and JEA. Approving final version of manuscript: all authors. IV takes responsibility for the integrity of the data analysis.