NFATc1 in mice represses osteoprotegerin during osteoclastogenesis and dissociates systemic osteopenia from inflammation in cherubism

Osteoporosis results from an imbalance in skeletal remodeling that favors bone resorption over bone formation. Bone matrix is degraded by osteoclasts, which differentiate from myeloid precursors in response to the cytokine RANKL. To gain insight into the transcriptional regulation of bone resorption during growth and disease, we generated a conditional knockout of the transcription factor nuclear factor of activated T cells c1 (Nfatc1). Deletion of Nfatc1 in young mice resulted in osteopetrosis and inhibition of osteoclastogenesis in vivo and in vitro. Transcriptional profiling revealed NFATc1 as a master regulator of the osteoclast transcriptome, promoting the expression of numerous genes needed for bone resorption. In addition, NFATc1 directly repressed osteoclast progenitor expression of osteoprotegerin, a decoy receptor for RANKL previously thought to be an osteoblast-derived inhibitor of bone resorption. “Cherubism mice”, which carry a gain-of-function mutation in SH3-domain binding protein 2 (Sh3bp2), develop osteoporosis and widespread inflammation dependent on the proinflammatory cytokine, TNF-α. Interestingly, deletion of Nfatc1 protected cherubism mice from systemic bone loss but did not inhibit inflammation. Taken together, our study demonstrates that NFATc1 is required for remodeling of the growing and adult skeleton and suggests that NFATc1 may be an effective therapeutic target for osteoporosis associated with inflammatory states.

Osteoporosis, the most common pathologic condition resulting from excessive bone destruction, afflicts 10 million Americans over the age of 50 and results in 1.5 million fractures annually (1, 2). Congenital and acquired inflammatory conditions, such as cherubism and RA, cause regional bone loss leading to skeletal deformities (3, 4). A scarcity of molecular targets in bone hinders the development of therapeutics for these diseases. Moreover, an inducible genetic system to terminate bone resorption in mice would facilitate the analysis of preclinical models of bone loss.

Osteoclasts are multinucleated, bone resorbing cells that differentiate from myeloid precursors in response to M-CSF and RANKL. Mice deficient in Csf1 (M-CSF), Tnfsf11 (RANKL), or either of their receptors (Csf1r [c-FMS] or Tnfrsf11a [RANK]), have increased bone density (osteopetrosis) with an absence of osteoclasts (5, 6). A decoy receptor for RANKL, TNFRSF11B (also known as osteoprotegerin [OPG]), balances osteoclastogenesis in vivo (7, 8). Although multiple tissues express OPG, most research has focused on osteoblasts as the physiologic source of OPG in the skeletal system (9).

Nuclear factor of activated T cell (NFAT) proteins are a family of 5 inducible transcription factors (NFATc1–4 and NFAT5) that control numerous cellular differentiation processes, including immune cell activation, heart valve formation, and the growth of cartilage and stem cells (10–12). NFATc1 is a target of RANK signaling. During osteoclastogenesis, NFATc1 expression is induced and this transcription factor can be identified at the promoters of osteoclast specific genes (13–19). Studies interrogating NFATc1 in the skeletal system have been limited since Nfatc1-deficient mice die at E13.5 of cardiac valve defects (20, 21). The importance of NFATc1 in osteoclastogenesis in vivo was confirmed through fetal liver complementation experiments (22) and by rescuing embryonic lethality by targeting expression of NFATc1 to the developing heart (23). However, a tractable system to explore the role of NFATc1-mediated osteoclastogenesis in pathologic bone destruction was lacking. Moreover, a source of NFATc1-deficient osteoclast progenitors was not available, hindering discovery of the universe of NFATc1-regulated genes.

Cherubism is a rare, autosomal dominant, pediatric disorder characterized by fibroinflammatory infiltrates in the face, leading to disfiguring swelling and cystic changes in the jaw. Large numbers of multinucleated giant cells that express osteoclast markers are found within the lesion (24). Mutations leading to cherubism were mapped to a signaling adapter molecule, SH3-domain binding protein 2 (SH3BP2) (25, 26), which may promote NFAT activation (27). Recently, a mouse model of cherubism was developed (3). These knock-in (KI) mice contain a missense mutation in Sh3bp2, resulting in a proline to arginine substitution (P416R), the most commonly identified mutation in cherubism (P418R in humans). Homozygous KI/KI mice exhibit multiorgan inflammation and severe systemic bone loss that is dependent on TNF-α derived from myeloid cells. In mouse models and humans, TNF-α has been implicated in the systemic bone loss associated with inflammatory conditions like RA and postmenopausal estrogen withdrawal states (28). Mutant SH3BP2 enhances TNF-α production in macrophages and promotes osteoclastogenesis via ERK and spleen tyrosine kinase pathways, respectively (3). Since NFATs can promote both osteoclastogenesis (22) and TNF-α production (29, 30), here we have asked which aspects of the cherubism phenotype depend on NFATc1?

Using a conditional knockout strategy, we show that genetic ablation of Nfatc1 results in osteoclast-poor osteopetrosis. This result complements previous studies (22, 23) and extends the requirement of NFATc1 to bone remodeling in the growing and adult skeleton. Molecular profiling of highly purified osteoclast precursors reveals 2 families of NFATc1-regulated genes in osteoclasts: those that are augmented by NFATc1 and those that absolutely depend on this transcription factor. Remarkably, in the absence of NFATc1, RANKL stimulates the expression of OPG, identifying a previously unrecognized function for NFATc1 in this pathway. Last, we show that NFATc1 functions downstream of mutant SH3BP2 and prevents the enhanced osteoclastogenesis and systemic bone loss characteristic of “cherubism mice”, which carry a gain-of-function mutation in Sh3bp2 (3). Notably, cherubism mice deficient in NFATc1 still display multiorgan inflammation. Thus, NFATc1 uncouples systemic inflammation from osteopenia in this unique model and may represent a therapeutic target for TNF-α–dependent bone loss.

Creation of an Nfatc1 conditional knockout mouse. To circumvent embryonic lethality (20), the exon encoding the regulatory domain of Nfatc1 (exon 3) was flanked with loxP sites (Nfatc1^fl/fl) (Figure 1A) (31). Nfatc1-deficient mice (Nfatc1^Δ/Δ) were generated by crossing Nfatc1^fl/fl mice with a transgenic line expressing the Cre recombinase from a type I interferon inducible promoter (Mx1-Cre) (32). Activation of Mx1-Cre by i.p. injection of polyinosinic-polycytidylic acid (poly I:C) resulted in complete deletion of exon 3 in BM DNA (Figure 1B), a greater than 95% reduction in Nfatc1 mRNA levels in BM (Figure 1D), and markedly diminished NFATc1 protein levels in splenic lysates (Figure 1E).

Figure 1

Generation of an Nfatc1 conditional knockout mouse. (A) Graphic representation of the Nfatc1 conditional targeting strategy. Flanking exon 3 of Nfatc1 with loxP sites generated the floxed allele (fl). Cre-mediated recombination of the loxP sites results in the delta allele (Δ). The approximate locations of the 5′ probe used for Southern blotting and PCR primers (see Methods) employed for genotyping (a, b, and c) are shown. HSV-1 TK, herpes simplex virus–1 thymidine kinase cassette; neoR, neomycin resistance cassette. (B) Southern blot analysis of tail and BM genomic DNA from poly I:C–treated Nfatc1^fl/fl (n = 4) and littermate Nfatc1^fl/fl, Mx1-Cre (n = 4) mice. Genomic tail DNA from Nfatc1^fl/+ and Nfatc1^fl/Δ mice are shown for comparison. (C) Three primer genotyping PCR, using primers a, b, and c shown in A, of tail genomic DNA from mice of the following genotypes: Nfatc1^+/+ (lane 1), Nfatc1^fl/+ (lane 2), Nfatc1^+/Δ (lane 3), Nfatc1^fl/fl (lane 4), Nfatc1^fl/Δ (lane 5). A 100-bp ladder and DNA diluent negative control (lane 6) are shown. (D) qRT-PCR analysis for Nfatc1 (using primer set Nfatc1ex3; Table 1) in mRNA samples from Nfatc1^fl/fl and Nfatc1^Δ/Δ BM. (E) Western blot analysis for NFATc1 and HSP90 (loading control) in protein lysates from Nfatc1^fl/fl (n = 2) and Nfatc1^Δ/Δ (n = 2) splenocytes.

Nfatc1^Δ/Δ mice develop osteoclast-poor osteopetrosis. Nfatc1 was ablated in 10-day-old pups to examine its role in the skeleton of growing and adult mice. Compared with littermate controls, Nfatc1^Δ/Δ mice displayed (a) short, club-shaped long bones, with higher radiodensity at the distal femur and proximal tibia (Figure 2, A–C, and Supplemental Figure 1A; supplemental material available online with this article; doi: 10.1172/JCI35711DS1); (b) a massive increase in Von Kossa and toluidine blue–positive matrix (calcified cartilage) adjacent to the growth plates (Figure 2, D and E); and (c) growth plate dysplasia, characterized by an increase in thickness, acellular areas, and the retention of hypertrophic chondrocytes beyond the chondro-osseous junction (Figure 2, E and F). Other skeletal sites, such as diaphyseal cortical bone and the epiphysis did not display increased bone density (Supplemental Figure 1B and A.O. Aliprantis, data not shown). Some Nfatc1^Δ/Δ mice developed upper incisor overgrowth with short or absent lower incisors (Figure 2G). Lateral radiographs of the skull did not show impaction of the lower incisors in the jaw but did reveal widened, hyperdense mandibular condyles (Figure 2H). Histologically, the mandibular condyles of Nfatc1^fl/fl mice showed a thin cap of articular cartilage overlying a mature neck of compact bone, while those from Nfatc1^Δ/Δ mice displayed marked retention of hypertrophic cartilage (Figure 2I). These data suggest that the tooth phenotype observed in Nfatc1^Δ/Δ mice is a consequence of malocclusion from abnormal biomechanical properties at the mandibular condyle, though a selective defect in lower incisor growth cannot be excluded. The increased bone density in Nfatc1^Δ/Δ mice is not due to augmented osteoblast function because (a) these mice showed a trend toward reduced osteoblast numbers (Supplemental Figure 1, C and D) and (b) Nfatc1^fl/fl mice expressing the Cre recombinase from the osteoblast-specific osterix promoter (33) did not display a similar high bone mass phenotype (Supplemental Figure 1E). Taken together, Nfatc1^Δ/Δ mice display features of severe osteopetrosis, with improperly shaped long bones and a failure to resorb primary spongiosa (calcified cartilage), resulting in aberrant endochondral growth and ossification (34, 35).

Figure 2

Nfatc1^Δ/Δ mice develop osteopetrosis. (A) Lateral digital radiograph of the leg of 5-month-old female Nfatc1^fl/fl (n = 2) and Nfatc1^Δ/Δ (n = 2) mice. The white arrows show increased radiodensity at the distal femurs. (B) Anteroposterior digital radiograph of the femur of 5-month-old female Nfatc1^fl/fl and Nfatc1^Δ/Δ mice. (C) The bone volume to total volume fraction was determined by microquantitative computed tomography at the distal femur of 6 male Nfatc1^fl/fl (black bar, n = 2 at 2.5 months old and n = 4 at 5 months old) and 4 male Nfatc1^Δ/Δ (white bar, n = 2 at 2.5 months old and n = 2 at 5 months old) mice. The data are the mean + SD; P < 1 × 10^–8. (D) Von Kossa and (E) toluidine blue stains of the distal femur of 5-month-old female Nfatc1^fl/fl and Nfatc1^Δ/Δ mice. Pictures are a montage of low-power images. (F) Toluidine blue stain (original magnification, ×400) of the femoral growth plate of 9-week-old male Nfatc1^fl/fl and Nfatc1^Δ/Δ mice. (G) Photograph of the snout of 5.5-month-old male Nfatc1^fl/fl and Nfatc1^Δ/Δ mice. (H) Lateral radiograph of the skull of 2.5-month-old female Nfatc1^fl/fl and Nfatc1^Δ/Δ mice. White arrowhead denotes lack of impaction of the lower incisors. White arrow shows increased radiodensity at the mandibular condyle. (I) Toluidine blue stain (original magnification, ×100) of the mandibular condyle of 3-month-old female Nfatc1^fl/fl and Nfatc1^Δ/Δ mice. Radiographs and histology images are representative of at least 8 (long bones) or 3 (mandibles) mice analyzed per genotype.

Osteopetrosis results from either a decrease in osteoclast numbers or a reduction in osteoclast function (6). Therefore osteoclast differentiation was investigated in Nfatc1^Δ/Δ mice in vivo and in vitro. Compared with Nfatc1^fl/fl littermates, Nfatc1^Δ/Δ mice displayed (a) a deficiency in tartrate-resistant acid phosphatase–positive (TRAP-positive) multinucleated cells at the growth plates of the long bones and mandibles (Figure 3A and A.O. Aliprantis, data not shown); (b) reduced osteoclast numbers by histomorphometric analysis (Figure 3B); and (c) decreased TRAP5b levels, a serologic marker of osteoclast function (Figure 3C). Rare TRAP-positive, morphologically normal osteoclasts could be identified in some Nfatc1^Δ/Δ mice (Supplemental Figure 1, F and G), consistent with our observation that serum TRAP5b levels were reduced, but not absent, in these mice. Consistent with markedly reduced osteoclast numbers in vivo, BM cells from Nfatc1^Δ/Δ mice failed to generate osteoclasts in vitro in response to either osteoblasts (Figure 3, D and E) or M-CSF and RANKL (Figure 3F). To establish that the differentiation defect was due to a failure of osteoclast progenitors to differentiate, and not to a lack of progenitors, osteoclast precursors were purified. Within the CD11b^lo/–CD3ε^–B220^–c-fms⁺ population in BM are c-kit⁺ and c-kit^– cells, with the former being the most highly osteoclastogenic (36). Similar numbers of each population were found in Nfatc1^fl/fl and Nfatc1^Δ/Δ BM (Figure 3, G and H). Fluorescence-activated cell sorting–purified (FACS-purified) BM-derived CD11b^lo/–CD3ε^–B220^–c-kit⁺c-fms⁺osteoclast precursors (BMOcPs) from Nfatc1^Δ/Δ mice failed to generate osteoclasts (Figure 3, I and J). Importantly, M-CSF induced normal macrophage differentiation in Nfatc1^Δ/Δ BM cells, as determined by CD11b and F4/80 staining (A.O. Aliprantis, data not shown) (37). In summary, Nfatc1-deficient mice display a progenitor intrinsic defect in osteoclastogenesis.

Figure 3

Nfatc1^Δ/Δ mice display impaired osteoclast differentiation in vivo and in vitro. (A) TRAP stain of the femoral growth plates of 9-week-old male Nfatc1^fl/fl and Nfatc1^Δ/Δ mice. Images are representative of at least 8 mice analyzed per genotype. (B) Histomorphometric enumeration of osteoclasts within the femoral epiphysis of 5-month-old female Nfatc1^fl/fl and Nfatc1^Δ/Δ mice (n = 4/genotype). The data is the mean + SD. NOc/BS, number of osteoclasts per mm bone surface; NOc/BAr, number of osteoclasts per mm² bone area. P < 0.05 for both panels. (C) TRAP5b levels in the serum of 10- to 20-week-old female Nfatc1^fl/fl and Nfatc1^Δ/Δ mice (n = 8/genotype); P < 0.05. (D) TRAP stain and (E) TRAP assay of Nfatc1^fl/fl and Nfatc1^Δ/Δ M-CSF–primed BM cells cultured with osteoblasts (Obs) and Vitamin D₃ (E) or Vitamin D₃ and PGE₂ (D and E). The data in E are the mean + SD of triplicate wells and representative of 2 independent experiments. (F) TRAP stain of Nfatc1^fl/fl and Nfatc1^Δ/Δ M-CSF–primed BM cells cultured with M-CSF or M-CSF and RANKL. (G) c-kit versus c-fms FACS plot of Nfatc1^fl/fl and Nfatc1^Δ/Δ BM cells gated on the CD11b^lo/–B220^–CD3ε^– population. (H) Quantification of osteoclast precursors in the BM of Nfatc1^fl/fl and Nfatc1^Δ/Δ mice. The data is the mean + SD (n = 5/genotype). (I) TRAP assay and (J) TRAP stain of Nfatc1^fl/fl and Nfatc1^Δ/Δ BMOcPs cultured with M-CSF or M-CSF and RANKL. The data in I is the mean + SD of triplicate wells and representative of more than 3 independent experiments. Original magnification, ×400 (A); ×100 (D, F, and J).

Nfatc1 deletion using lysozyme–M-Cre does not result in osteopetrosis. Lysozyme–M-Cre (LysM-Cre) (38) has been used to interrogate osteoclast function in mice carrying conditional alleles (39, 40). LysM-Cre mice express the Cre recombinase from a myeloid specific promoter (38). Surprisingly, no alterations in bone density or osteoclast differentiation were observed in Nfatc1^fl/Δ, LysM-Cre mice (Supplemental Figure 2, A–C). This result was explained by the fact that LysM-Cre failed to efficiently delete the floxed allele in BM macrophage or osteoclast cultures (Supplemental Figure 2D). However, near complete deletion was observed in peritoneal macrophages (Supplemental Figure 2E). These findings are consistent with the original report of this line (38) and indicate that LysM-Cre may be of limited value to the study of conditional alleles in osteoclasts.

Microarray analysis of FACS-sorted osteoclast progenitors reveals 2 families of NFATc1-regulated genes in the developing osteoclast. NFATc1 is necessary for osteoclast differentiation in vivo and in vitro (Figures 2 and 3) (22). However, the universe of NFATc1-regulated transcripts in osteoclasts is unknown. A protocol to isolate BMOcPs was developed (Supplemental Figure 3, A and B) to facilitate a high fidelity comparison of the mRNA profiles of WT and NFATc1-deficient developing osteoclasts. In response to RANKL, WT BMOcPs (a) displayed an approximately 33-fold higher osteoclast forming capacity on a per cell basis compared with crude BM cells, (b) upregulated Nfatc1 mRNA and NFATc1 protein, and (c) expressed genes associated with terminal osteoclast differentiation (Supplemental Figure 3, C–F). mRNA profiling, by DNA microarray and qRT-PCR, of Nfatc1^fl/fl and Nfatc1^Δ/Δ M-CSF– and RANKL-stimulated BMOcPs (MROcPs) revealed 2 differentially regulated sets of osteoclast genes: (a) NFATc1-dependent genes (Figure 4, A and B) and (b) “NFATc1-augmented” genes (Figure 4, C and D). Oscar, Calcr, and Itgb3, all well-characterized osteoclast genes (6, 19), were absolutely dependent on NFATc1 for expression (Figure 4, A and B). Likewise, Myo1d and Mst1r, potential regulators of bone resorption (41, 42), depended on NFATc1. Lastly, Rcan2, a regulator of the NFAT pathway and homolog of Rcan1, an NFATc1 target gene in the heart (43), was only induced in NFATc1-sufficient cells. In contrast, the following genes, whose function in osteoclasts is established (35, 44, 45), could be partially induced by RANKL even in the absence of NFATc1: Mmp9, Acp5, Ctsk, Car2, Mmp14, and Clcn7 (Figure 4, C and D). However, NFATc1 did augment the expression of all 6 genes. These data support NFATc1 as a broad regulator of osteoclast gene expression but suggest that NFATc1 is more essential to the control of some genetic loci than others.

Figure 4

Comparative microarray analysis reveals 2 sets of NFATc1-regulated genes in osteoclasts. (A) Microarray signal intensities for selected NFATc1-dependent mRNAs from Nfatc1^fl/fl and Nfatc1^Δ/Δ MROcPs. (B) qRT-PCR analysis for the expression of the NFATc1-dependent mRNAs identified in A in Nfatc1^fl/fl and Nfatc1^Δ/Δ BMOcPs stimulated with M-CSF or M-CSF and RANKL for 3 days. (C) Microarray signal intensities for selected NFATc1-augmented mRNAs from Nfatc1^fl/fl and Nfatc1^Δ/Δ MROcPs. (D) qRT-PCR analysis for the expression of the NFATc1-augmented mRNAs identified in C in Nfatc1^fl/fl and Nfatc1^Δ/Δ BMOcPs stimulated with M-CSF or M-CSF and RANKL for 3 days. (E) qRT-PCR analysis for the expression of the Nfatc1 mRNA isoforms in Nfatc1^fl/fl and Nfatc1^Δ/Δ BMOcPs stimulated with M-CSF or M-CSF and RANKL for 3 days. PCR primers targeted the deleted exon (Nfatc1ex3) or the Nfatc1/A isoform (see Table 1). Note that the qRT-PCR analyses presented in B, D, and E were performed on the same mRNA samples and are representative of at least 2 independent experiments. (F) qRT-PCR analysis for the expression of Nfatc1/A mRNA in NFATc2 sufficient (Nfatc2^+/+) and deficient (Nfatc2^–/–) Nfatc1^fl/fl and Nfatc1^Δ/Δ BMOcPs stimulated with M-CSF or M-CSF and RANKL for 4 days. The data in F are representative of at least 2 independent experiments.

Surprisingly, a microarray probe set (1417621_at) to Nfatc1 revealed expression in Nfatc1^Δ/Δ MROcPs (Figure 4C). This probe set targets the 3′ untranslated region of the Nfatc1/A isoform, which is preferentially induced during osteoclastogenesis and suggested to be a target of autoamplification (22). Since neither full-length Nfatc1 mRNA (Figure 4E, primer set Nfatc1ex3, Table 1) nor NFATc1 protein could be detected in Nfatc1^Δ/Δ MROcPs (Supplemental Figure 3E), this suggested that an untranslated, mutant Nfatc1/A mRNA is induced by RANKL in these cells. qRT-PCR confirmed the expression of Nfatc1/A mRNA in Nfatc1^Δ/Δ BMOcPs stimulated with RANKL (Figure 4E). Thus, the endogenous Nfatc1 locus can still be stimulated by RANKL in the absence of NFATc1. However, NFATc1 was necessary for optimal expression, since higher levels of Nfatc1/A mRNA were detected in Nfatc1^fl/fl BMOcPs stimulated with RANKL compared with Nfatc1^Δ/Δ cells (Figure 4E), consistent with autoamplification (22). Thus, Nfatc1/A is an NFATc1 augmented transcript. Since NFAT family members often display redundant functions (46) and NFATc2 is recruited to the Nfatc1 promoter (22), the induction of Nfatc1/A mRNA by RANKL was investigated in NFATc2-deficient and NFATc1/NFATc2 doubly deficient BMOcPs. WT (Nfatc1^fl/flNfatc2^+/+) and NFATc2-deficient (Nfatc1^fl/flNfatc2^–/–) BMOcPs strongly upregulated Nfatc1/A in response to RANKL treatment for 4 days (Figure 4F). Interestingly, the induction of Nfatc1/A mRNA was not further attenuated in doubly deficient (Nfatc1^Δ/ΔNfatc2^–/–) BMOcPs compared with those lacking NFATc1 alone (Nfatc1^Δ/ΔNfatc2^+/+) (Figure 4F). These data suggest that NFATc2 does not play a dominant or compensatory role in the induction of Nfatc1/A mRNA during osteoclastogenesis.

Table 1

qRT-PCR primers used in this study

Tnfrsf11b is upregulated in osteoclast precursors in the absence of NFATc1. Although the expression and function of OPG by osteoblasts has been well studied (47), OPG expression by osteoclast-like cells has only been reported, to our knowledge, in giant cell tumors (24, 48). Remarkably, 2 probe sets showed expression of Tnfrsf11b in Nfatc1^Δ/Δ MROcPs but not in Nfatc1^fl/fl cells (Figure 5A). In comparison, Tnfrsf11a expression levels were similar (Figure 5A). The expression of Tnfrsf11b mRNA by Nfatc1^Δ/Δ MROcPs was similar to levels found in WT osteoblasts (Figure 5B), increased over time in culture (Figure 5C), and was dependent on RANKL, as cells cultured in M-CSF alone did not express Tnfrsf11b (Figure 5D). In addition, OPG was detected in the culture supernatants of Nfatc1^Δ/Δ MROcPs (Figure 5E). Treatment of Nfatc1^fl/fl MROcPs with cyclosporine A, an inhibitor of the NFAT activating phosphatase calcineurin (10), reciprocally upregulated Tnfrsf11b and downregulated the NFATc1-dependent gene, Itgb3 (Figure 5, F and G). Despite the dramatic upregulation of Tnfrsf11b mRNA by Nfatc1^Δ/Δ MROcPs, no differences in serum OPG levels were observed between Nfatc1^fl/fl and Nfatc1^Δ/Δ mice (Supplemental Figure 4). To exclude the possibility that cells of mesenchymal origin (i.e., osteoblasts) were contaminating our BMOcP preparations from Nfatc1^Δ/Δ mice and producing the OPG signal observed in these experiments, osteoclast precursors were derived from spleen cells. Similar to FACS-sorted BMOcPs, Nfatc1^Δ/Δ spleen cells both failed to differentiate into osteoclasts and upregulated Tnfrsf11b upon stimulation with RANKL (Supplemental Figure 5, A–D). Interestingly, a putative NFAT binding site was identified in a conserved region of the human TNFRSF11B and mouse Tnfrsf11b promoters, located approximately 2 kb and 3 kb from the transcription start site (TSS), respectively (Supplemental Figure 6A). After 3 days of RANKL treatment, NFATc1 was recruited to the Tnfrsf11b promoter, as determined by a ChIP assay using primers flanking this conserved sequence (Figure 5H and Supplemental Figure 6A). In contrast, NFATc1 binding to the Itgb3 promoter peaked at 1.5 days of RANKL treatment and was almost undetectable by 3 days (Supplemental Figure 6B), consistent with previous observations (18). Lastly, a constitutively active version of NFATc1 (caNFATc1) repressed the basal activity of a Tnfrsf11b reporter plasmid (pOPG 3.6-luc) (Figure 5I). Appropriately, caNFATc1 had little effect on an NF-κb reporter plasmid (pBII-luc) and strongly activated an NFAT luciferase reporter construct (pNFAT-luc) (Supplemental Figure 6, C and D). Taken together, these results uncover a previously unrecognized regulatory loop downstream of RANKL: calcineurin/NFATc1-mediated repression of OPG expression.

Figure 5

RANKL induces Tnfrsf11b expression in the absence of NFATc1. (A) Microarray signal intensities for Tnfrsf11b (upper panel) and Tnfrsf11a (lower panel) mRNAs from Nfatc1^fl/fl (black bars) and Nfatc1^Δ/Δ (white bars) MROcPs. (B–D) qRT-PCR analysis for Tnfrsf11b mRNA in (B) Nfatc1^fl/fl and Nfatc1^Δ/Δ MROcPs or WT calvarial osteoblasts with or without Vitamin _D3 and parathyroid hormone (PTH); (C) Nfatc1^fl/fl (diamonds) and Nfatc1^Δ/Δ (circles) BMOcPs stimulated with M-CSF and RANKL for 1, 2, or 4 days; and (D) Nfatc1^fl/fl and Nfatc1^Δ/Δ BMOcPs stimulated for 3 days with M-CSF or M-CSF and RANKL. (E) ELISA for OPG in the supernatants of Nfatc1^fl/fl and Nfatc1^Δ/Δ BMOcPs stimulated for 4 days with M-CSF or M-CSF and RANKL and for 2 days with M-CSF only. b.d., below detection. The data are the mean + SD of 4 independent wells. (F) qRT-PCR analysis for Tnfrsf11b and (G) Itgb3 mRNA in Nfatc1^fl/fl BMOcPs stimulated with M-CSF and RANKL in the absence or presence of increasing concentrations of cyclosporine A (CsA) (62.5, 125, or 250 ng/ml CsA). The triangle under the x axis refers to increasing CsA concentrations. (H) NFATc1 ChIP of osteoclast precursors incubated with RANKL for 0, 1.5, or 3 days. Immunoprecipitated chromatin was analyzed by qRT-PCR for Tnfrsf11b promoter DNA, which was normalized to input. Data is the mean + SD of 2 independent experiments (*P < 0.05, **P = 0.053). (I) Relative luciferase (luc) activity of 293T cells transfected with pOPG 3.6-luc and increasing amounts of pMSCV-caNfatc1 (0, 8, 40, and 200 ng/transfection). The triangle under the x axis refers to increasing amounts of pMSCV-caNfatc1 per transfection. Data are the mean + SD of transfections performed in triplicate. The data in C and B, D–G, and I are representative of 2 and 3 similar experiments, respectively.

NFATc1 dissociates systemic bone loss and osteoclastogenesis from inflammation in cherubism. Our conditional knockout strain represents a unique tool to evaluate the role of NFATc1 in disease. Cherubism is a pediatric, craniofacial, inflammatory disorder, characterized by bone loss in the jaw that is caused by mutations in an adapter molecule SH3BP2 (3, 25). To clarify the role of NFATc1 in cherubism, Mx1-Cre was used to ablate Nfatc1 in mice carrying 1 (KI/+) or 2 (KI/KI) cherubism alleles (3). This approach was taken since the cherubism phenotype is intrinsic to myeloid cells (3) and deletion of Nfatc1^fl/fl by Mx1-Cre is complete in the BM (Figure 1, B and D). Deletion of Nfatc1 reversed the diffuse bone loss and focal cystic changes characteristic of cherubism mice (Figure 6A). In contrast, Nfatc1^Δ/ΔKI/KI mice still developed widespread inflammation in the liver and lungs (Figure 6, B and C) as well as lymph nodes, spleen, and stomach (A.O. Aliprantis and Y. Ueki, data not shown). The systemic bone loss and inflammatory phenotypes of cherubism mice are dependent on TNF-α derived from myeloid cells, which express abnormally high levels of Tnf mRNA (3). Consistent with our observation that deletion of Nfatc1 had no effect on inflammation in vivo, BM macrophage cultures from KI/+ and KI/KI mice expressed higher levels of Tnf mRNA than WT cells irrespective of their Nfatc1 genotype (Figure 6D). Thus, NFATc1 uncouples diffuse osteopenia from inflammation in this TNF-α driven, clinically relevant mouse model of a human disease. Furthermore, the enhanced osteoclastogenesis conferred by mutant SH3BP2 in vitro, which occurs in a gene dosage-dependent manner, was completely abolished in BM cells lacking NFATc1, as assessed by TRAP staining (Figure 7A) or the expression of NFATc1-regulated genes (Figure 7B). Accordingly, mutant SH3BP2 augmented RANKL induced expression of Nfatc1/A in proportion to gene dosage (Figure 7C) in NFATc1-sufficient cells. Conversely, RANKL did not increase the expression of the mutant Nfatc1/A isoform in either Nfatc1^Δ/ΔKI/+ or Nfatc1^Δ/ΔKI/KI BM cells, indicating that NFATc1 protein is necessary for SH3BP2 to mediate this effect (Figure 7C). Western blotting confirmed increased NFATc1 expression in Nfatc1^fl/flKI/KI BM cells treated with RANKL compared with WT cells (Figure 7D). Moreover, NFATc1 was not detected in Nfatc1^Δ/Δ cells, irrespective of their Sh3bp2 genotype (+/+ or KI/KI) or RANKL treatment (Figure 7D). These observations strongly suggest that SH3BP2 promotes osteoclastogenesis by augmenting NFATc1 autoamplification. Since TNF-α has been reported to induce osteoclast differentiation independent of RANKL (49) and Sh3bp2 mutant mice have elevated levels of TNF-α (3), osteoclast differentiation was assessed in vivo in cherubism mice lacking NFATc1. Consistent with our observations of increased bone density in Nfatc1^Δ/ΔKI/KI mice (Figure 6A) and failure of BM cells from these mice to differentiate into osteoclasts in vitro (Figure 7, A and B), a dramatic reduction in TRAP-positive cells was seen at the proximal humerus (Figure 7E) and distal femur (A.O. Aliprantis and Y. Ueki, data not shown) of Nfatc1^Δ/ΔKI/KI compared with Nfatc1^fl/flKI/KI mice. These data imply that TNF-α in this model cannot drive NFATc1-independent osteoclast differentiation.

Figure 6

NFATc1 uncouples bone loss from inflammation in a mouse model of cherubism. (A) Anteroposterior digital radiograph of the femurs and tibias of 12-week-old female mice. White arrow denotes cystic changes at the distal tibia of Nfatc1^fl/flKI/KI mice. Two other Nfatc1^Δ/ΔKI/KI mice (1 mouse each at 12 and 17 weeks old) had a similar radiographic appearance to the Nfatc1^Δ/ΔKI/KI sample displayed. (B) H&E stain of the liver and (C) lung of 12-week-old female mice. Black arrows in B and C indicate examples of inflammatory infiltrates. The original magnifications used to obtain the images are indicated above each column. The organ histology of Nfatc1^fl/flKI/KI mice was identical to that previously described for mice bearing 2 KI alleles (3). A second Nfatc1^Δ/ΔKI/KI mouse had similar histology to the Nfatc1^Δ/ΔKI/KI sample displayed. (D) qRT-PCR analysis for the expression of Tnf mRNA in BM cells cultured for 5 days with 50 ng/ml M-CSF.

Figure 7

NFATc1 is downstream of SH3BP2 in osteoclastogenesis in vitro and in vivo. (A) TRAP stain (original magnification, ×100) of BM cells incubated with 50 ng/ml M-CSF and 25 ng/ml RANKL for 4 days. (B and C) qRT-PCR analysis for the expression of the indicated genes in BM cells incubated with 50 ng/ml M-CSF and 25 ng/ml RANKL for 3 days. The data in A–C is representative of 2 independent experiments. (D) Western blot analysis for NFATc1 and β-actin in BM cells incubated with 50 ng/ml M-CSF and 25 ng/ml RANKL for 3 days. (E) TRAP stain (original magnification, ×100) of the proximal humerus of 12-week-old female mice. The histology in E is representative of at least 3 Nfatc1^Δ/ΔKI/KI mice.

Evidence for the role of NFATc1 in osteoclastogenesis in vivo has been provided by 2 other studies. The first showed that Nfatc1-deficient fetal liver or embryonic stem cells do not rescue the osteopetrotic phenotype of c-fos^–/– mice (22). The second study targeted NFATc1 expression to the developing heart to rescue embryonic lethality. Nfatc1^–/–,Tie2-Nfatc1 mice showed osteopetrosis with absent osteoclasts (23). These studies examined mice at 6 and 3 weeks, respectively, and provided important evidence for the function of NFATc1 in specifying osteoclast differentiation during development. However, the requirement for NFATc1 in adult skeletal remodeling remained unresolved. Here Mx1-Cre was used to delete Nfatc1 in mice at 10 days of age as was previously done to investigate the role of Iκβ kinase β in osteoclast survival (50). Nfatc1^Δ/Δ mice developed osteoclast-poor osteopetrosis that was most prominent in the long bones and mandibles (Figures 2 and 3). Our data complement these previous studies (22, 23) and extend the requirement for NFATc1 to remodeling of the growing and adult skeleton. Moreover, unlike osteopetrotic, M-CSF–deficient op/op mice, which resolve their phenotype with age (51), Nfatc1^Δ/Δ mice up to the age of 5 months still displayed osteopetrosis. Interestingly, TRAP5b levels were reduced but not absent in Nfatc1^Δ/Δ mice (Figure 3C). This observation may be due to variability in Nfatc1^fl/fl deletion by Mx1-cre at anatomic sites not examined in the current study (e.g., the vertebral skeleton). In our experience, within the bone marrow of long bones, Mx1-cre–mediated deletion of Nfatc1^fl/fl was nearly complete at the level of DNA recombination or mRNA expression (Figure 1, B and D).

NFATc1 binds the promoters of many osteoclast-specific genes, including Nfatc1 itself (13–19, 22). However, a source of NFATc1-deficient osteoclast progenitors from mouse BM was not available prior to this study. Therefore, the relative contribution of NFATc1 to the expression of the osteoclast transcriptome could not be determined using loss of function experiments. To facilitate this analysis, a strategy to isolate a pure population of BMOcPs was devised (Supplemental Figure 3). Transcriptional profiling revealed NFATc1 as a broad regulator of osteoclast gene expression (Figure 4). Two sets of NFATc1-regulated genes were identified: those that are augmented by NFATc1 (Figure 4, C and D) but still induced to some extent by RANKL in NFATc1-deficient cells and those that absolutely depend on NFATc1 for expression (Figure 4, A and B). The temporal recruitment of NFATc1 to different genetic loci may underlie this observation. Recently, ChIP analyses have demonstrated that NFATc1 binds the Itgb3, Oscar, Acp5, and Ctsk promoters during osteoclastogenesis (16–18, 52). Interestingly, whereas NFATc1 was recruited to the Itgb3 and Oscar promoters (NFATc1-dependent genes) within 12–48 hours of RANKL stimulation (17, 18, 52), NFATc1 was only detected at the Acp5 and Ctsk promoters (NFATc1-augmented genes) at day 5 after RANKL stimulation (16). The fact that we see differences in expression of Acp5 and Ctsk at day 3 (Figure 4D) between NFATc1-sufficient and -deficient cells is likely due to our use of highly osteoclastogenic BMOcPs. We hypothesize that NFATc1-dependent genes need early NFATc1 recruitment to initiate transcription, whereas NFATc1-augmented genes employ NFATc1 at later time points to effect optimal expression. Other possible reasons these 2 sets of genes differentially require NFATc1 include compensatory use of other osteoclast transcription factors, like AP-1, c-Fos, Pu.1, and Mitf (19, 53), and indirect effects of NFATc1 deficiency.

The most intriguing finding from our profiling experiments was the dramatic upregulation of Tnfrsf11b mRNA and protein by RANKL in NFATc1-deficient BMOcPs (Figure 5). OPG, a decoy receptor for RANKL, is a potent negative regulator of osteoclastogenesis. Hence, OPG-deficient mice are osteoporotic (54) while OPG transgenic mice, like RANK knockout mice, display increased bone mass (7, 55). Osteoblasts are thought to be the physiologic source of OPG in the skeleton (47). However, contaminating osteoblasts are very unlikely to be the source of OPG in our experiments, since a pure population of BMOcPs, isolated by flow cytometry, was used. Moreover, osteoclast precursors derived from Nfatc1^Δ/Δ spleen cells also upregulated Tnfrsf11b mRNA in response to RANKL. Although, the physiologic significance of our observation is presently unknown, we suggest that NFATc1 dampens the induction of OPG that occurs as a consequence of signaling pathways shared by RANK and other TNF receptor family members. The RANKL/RANK/OPG axis has a well-described role outside of the skeleton in regulating adaptive immune responses (9). In human B cells and dendritic cells, CD40 ligation induces OPG expression (56, 57). CD40, like RANK, is a member of the TNF receptor superfamily. Both receptors utilize TNF receptor-associated factor 6 as a proximal signaling component (58, 59). However, RANK is unique in its ability to promote calcium oscillations and upregulate NFATc1 (59). Consistent with these observations, WT BMOcPs stimulated with RANKL in the presence of cyclosporine A, which blocks calcium-mediated activation of the calcineurin/NFAT pathway, express more Tnfrsf11b mRNA (Figure 5F). Furthermore, it is likely that NFATc1 directly, negatively regulates the Tnfrsf11b promoter (Figure 5, H and I), which is similar to the observation in previous reports that NFAT family members repress Cdk4 expression (12, 60). Taken together, these observations suggest that suppression of OPG expression downstream of RANKL is a previously unrecognized function of NFATc1 in osteoclastogenesis. Interestingly, giant cell tumors of bone, which contain multinucleated giant cells with features of osteoclasts, have been shown to express OPG (24, 48). This intriguing clinical correlate suggests that dysregulation of OPG expression by cells of the osteoclast lineage may occur in pathologic conditions.

Our Nfatc1 conditional knockout mouse provides a genetic system to study the role of this transcription factor in disease. Here, a mouse model of the human disorder, cherubism, was investigated. Cherubism mice carry a gain-of-function mutation in Sh3bp2 and develop systemic inflammation and bone loss due to overproduction of TNF-α by myeloid cells (3). NFATc1-deficient mice homozygous for the cherubism mutation still developed inflammation in multiple organs (Figure 6, B and C). In addition, macrophages grown from the BM of Nfatc1^Δ/ΔKI/KI mice had similarly elevated levels of Tnf transcripts compared with their NFATc1-sufficient counterparts (Figure 6D). These results indicate that NFATc1 is not downstream of SH3BP2 for TNF-α production and are consistent with a recent report that NFATc1 does not promote TNF-α expression in T cells (30). However, the inflammatory phenotype of cherubism may depend on other NFAT family members, like NFATc2 (30, 61, 62). Genetic ablation of Nfatc1 did protect cherubism mice from systemic bone loss (Figure 6A) and completely abrogated the enhanced osteoclastogenesis observed in vitro in BM cells carrying mutant alleles of Sh3bp2 (Figure 7, A and B). Mutant Sh3bp2 cells also displayed increased induction of the Nfatc1/A isoform by RANKL in a manner proportional to gene dosage. This effect was not observed in NFATc1-deficient cells (Figure 7C). Taken together, these observations situate SH3BP2 upstream of NFATc1 in osteoclastogenesis, likely within the NFATc1 autoamplification loop. Cherubism mice also develop TNF-α–dependent arthritis with cortical erosions (3). Whether NFATc1 deficiency will protect from local bone loss around inflamed joints, as has been described for osteoclast-deficient mice in other arthritis models (63, 64), is under investigation.

In conclusion, this study reveals that NFATc1 is an important regulator of osteoclastogenesis in growing and adult mice consistent with previous reports exploring its function in embryonic and early postnatal development (22, 23). NFATc1 deficiency also protects from systemic bone loss in a mouse model of the human disease, cherubism, without inhibiting the inflammatory aspects of the disease. A gene profiling approach revealed that most genes classically associated with terminal osteoclast differentiation require NFATc1 for expression. Intriguingly, we found that RANKL promoted a dramatic induction of OPG in NFATc1-deficient cells. We propose that pharmacologic perturbation of the NFAT pathway in osteoclast precursors may transform these cells into potent OPG producers, which could inhibit resorption in the bone microenvironment and lead to novel treatments for osteolysis.

View Supplemental data

The authors especially thank Stanford Peng, Geancarlo Lugo-Villarino, and Arlene Sharpe for their early participation in the NFATc1 conditional knockout project. We also acknowledge Dorothy Zhang, Kazuki Ueno, David Burr, and Keith Condon for their preparation and analysis of histologic sections; Deneen Kozoriz for cell sorting; Landy Kangaloo, Steven Swartz, Pamela Okerholm, and Jennifer Donovan for technical assistance; Zie Skobe, Roberto Fajardo, and Pat Dunning for help with radiology; Kyleigh Nevis for manuscript preparation; and Dallas Jones, Ann-Hwee Lee, Marc Wein, Jing Wu, Ellen Gravallese, Kevin McHugh, Tania Crotti, and Nicole Walsh for guidance and experimental advice. We are indebted to the staff at the Harvard Medical School Center for Animal Resources and Comparative Medicine for their excellent care of our mice. This work was supported by grants from the Charles H. Hood Foundation (to Y. Ueki), NIH AR36819 (to B.R. Olsen), NIH AI31541 (to L.H. Glimcher), and NIH AR0447129 (to M.C. Ostrowski). A.O. Aliprantis is the recipient of the Abbott Scholar Award in Rheumatology Research and the 2007 American Society for Clinical Investigation Young Investigator Award. R. Sulyanto is a recipient of the Harvard Medical School Carl W. Walter Fellowship.

Address correspondence to: Laurie H. Glimcher, Harvard School of Public Health, 651 Huntington Avenue, FXB 205, Boston, Massachusetts 02115, USA. Phone: (617) 432-0622; Fax: (617) 432-0084; E-mail: lglimche@hsph.harvard.edu.

Nonstandard abbreviations used: BMOcP, BM-derived CD11b^lo/–CD3e^–B220^–c-kit⁺c-fms⁺ osteoclast precursor; FACS, fluorescence-activated cell sorting; KI, knock-in; LysM-Cre, lysozyme–M-Cre; MROcP, M-CSF– and RANKL-stimulated BMOcP; NFAT, nuclear factor of activated T cells; OPG, osteoprotegerin; poly I:C, polyinosinic-polycytidylic acid; Sh3bp2, SH3-domain binding protein 2; TRAP, tartrate-resistant acid phosphatase.

Conflict of interest: Laurie H. Glimcher has equity in and is on the corporate board of directors of the Bristol-Myers Squibb Company.

Reference information: J. Clin. Invest.118:3775–3789 (2008). doi:10.1172/JCI35711

Yasuyoshi Ueki’s present address is: Department of Oral Biology, School of Dentistry, University of Missouri-Kansas City, Kansas City, Missouri, USA.

Yasuyoshi Ueki and Rosalyn Sulyanto contributed equally to this work.

1. United States Department of Health and Human Services. 2004. Bone health and osteoporosis: a report of the Surgeon General. http://www.surgeongeneral.gov/library/bonehealth/content.html .
2. Woolf, A.D., Pfleger, B. 2003. Burden of major musculoskeletal conditions. Bull. World Health Organ. 81:646-656.
   View this article via: PubMed Google Scholar
3. Ueki, Y., et al. 2007. Increased myeloid cell responses to M-CSF and RANKL cause bone loss and inflammation in SH3BP2 “cherubism” mice. Cell. 128:71-83.
   View this article via: CrossRef PubMed Google Scholar
4. Goldring, S.R., Gravallese, E.M. 2000. Pathogenesis of bone erosions in rheumatoid arthritis. Curr. Opin. Rheumatol. 12:195-199.
   View this article via: CrossRef PubMed Google Scholar
5. Wada, T., Nakashima, T., Hiroshi, N., Penninger, J.M. 2006. RANKL-RANK signaling in osteoclastogenesis and bone disease. Trends Mol. Med. 12:17-25.
   View this article via: CrossRef PubMed Google Scholar
6. Teitelbaum, S.L., Ross, F.P. 2003. Genetic regulation of osteoclast development and function. Nat. Rev. Genet. 4:638-649.
   View this article via: CrossRef PubMed Google Scholar
7. Simonet, W.S., et al. 1997. Osteoprotegerin: a novel secreted protein involved in the regulation of bone density. Cell. 89:309-319.
   View this article via: CrossRef PubMed Google Scholar
8. Bucay, N., et al. 1998. Osteoprotegerin-deficient mice develop early onset osteoporosis and arterial calcification. Genes Dev. 12:1260-1268.
   View this article via: CrossRef PubMed Google Scholar
9. Boyce, B.F., Xing, L. 2007. Biology of RANK, RANKL, and osteoprotegerin. Arthritis Res. Ther. 9(Suppl. 1):S1.
   View this article via: PubMed Google Scholar
10. Hogan, P.G., Chen, L., Nardone, J., Rao, A. 2003. Transcriptional regulation by calcium, calcineurin, and NFAT. Genes Dev. 17:2205-2232.
    View this article via: CrossRef PubMed Google Scholar
11. Graef, I.A., Chen, F., Crabtree, G.R. 2001. NFAT signaling in vertebrate development. Curr. Opin. Genet. Dev. 11:505-512.
    View this article via: CrossRef PubMed Google Scholar
12. Horsley, V., Aliprantis, A.O., Polak, L., Glimcher, L.H., Fuchs, E. 2008. NFATc1 balances quiescence and proliferation of skin stem cells. Cell. 132:299-310.
    View this article via: CrossRef PubMed Google Scholar
13. Ishida, N., et al. 2002. Large scale gene expression analysis of osteoclastogenesis in vitro and elucidation of NFAT2 as a key regulator. J. Biol. Chem. 277:41147-41156.
    View this article via: CrossRef PubMed Google Scholar
14. Takayanagi, H., et al. 2002. Induction and activation of the transcription factor NFATc1 (NFAT2) integrate RANKL signaling in terminal differentiation of osteoclasts. Dev. Cell. 3:889-901.
    View this article via: CrossRef PubMed Google Scholar
15. Kim, K., Lee, S.H., Ha Kim, J., Choi, Y., Kim, N. 2008. NFATc1 induces osteoclast fusion via up-regulation of Atp6v0d2 and the dendritic cell-specific transmembrane protein (DC-STAMP). Mol. Endocrinol. 22:176-185.
    View this article via: CrossRef PubMed Google Scholar
16. Sharma, S.M., et al. 2007. MITF and PU.1 recruit p38 MAPK and NFATc1 to target genes during osteoclast differentiation. J. Biol. Chem. 282:15921-15929.
    View this article via: CrossRef PubMed Google Scholar
17. Kim, Y., et al. 2005. Contribution of nuclear factor of activated T cells c1 to the transcriptional control of immunoreceptor osteoclast-associated receptor but not triggering receptor expressed by myeloid cells-2 during osteoclastogenesis. J. Biol. Chem. 280:32905-32913.
    View this article via: CrossRef PubMed Google Scholar
18. Crotti, T.N., et al. 2008. PU.1 and NFATc1 mediate osteoclastic induction of the mouse beta3 integrin promoter. J. Cell. Physiol. 215:636-644.
    View this article via: CrossRef PubMed Google Scholar
19. Sharma, S.M., et al. 2006. Genetics and genomics of osteoclast differentiation: integrating cell signaling pathways and gene networks. Crit. Rev. Eukaryot. Gene Expr. 16:253-277.
    View this article via: PubMed Google Scholar
20. Ranger, A.M., et al. 1998. The transcription factor NF-ATc is essential for cardiac valve formation. Nature. 392:186-190.
    View this article via: CrossRef PubMed Google Scholar
21. Luis de la Pompa, J., et al. 1998. Role of the NF-ATc transcription factor in morphogenesis of cardiac valves and septum. Nature. 392:182-186.
    View this article via: CrossRef PubMed Google Scholar
22. Asagiri, M., et al. 2005. Autoamplification of NFATc1 expression determines its essential role in bone homeostasis. J. Exp. Med. 202:1261-1269.
    View this article via: CrossRef PubMed Google Scholar
23. Winslow, M.M., et al. 2006. Calcineurin/NFAT signaling in osteoblasts regulates bone mass. Dev. Cell. 10:771-782.
    View this article via: CrossRef PubMed Google Scholar
24. Liu, B., Yu, S.F., Li, T.J. 2003. Multinucleated giant cells in various forms of giant cell containing lesions of the jaws express features of osteoclasts. J. Oral Pathol. Med. 32:367-375.
    View this article via: CrossRef PubMed Google Scholar
25. Ueki, Y., et al. 2001. Mutations in the gene encoding c-Abl-binding protein SH3BP2 cause cherubism. Nat. Genet. 28:125-126.
    View this article via: CrossRef PubMed Google Scholar
26. Tiziani, V., et al. 1999. The gene for cherubism maps to chromosome 4p16. Am. J. Hum. Genet. 65:158-166.
    View this article via: CrossRef PubMed Google Scholar
27. Lietman, S.A., Kalinchinko, N., Deng, X., Kohanski, R., Levine, M.A. 2006. Identification of a novel mutation of SH3BP2 in cherubism and demonstration that SH3BP2 mutations lead to increased NFAT activation. Hum. Mutat. 27:717-718.
    View this article via: PubMed Google Scholar
28. Weitzmann, M.N., Pacifici, R. 2005. The role of T lymphocytes in bone metabolism. Immunol. Rev. 208:154-168.
    View this article via: CrossRef PubMed Google Scholar
29. Peng, S.L., Gerth, A.J., Ranger, A.M., Glimcher, L.H. 2001. NFATc1 and NFATc2 together control both T and B cell activation and differentiation. Immunity. 14:13-20.
    View this article via: CrossRef PubMed Google Scholar
30. Kaminuma, O., et al. 2008. Differential contribution of NFATc2 and NFATc1 to TNF-alpha gene expression in T cells. J. Immunol. 180:319-326.
    View this article via: PubMed Google Scholar
31. Kuhn, R., Torres, R.M. 2002. Cre/loxP recombination system and gene targeting. Methods Mol. Biol. 180:175-204.
    View this article via: PubMed Google Scholar
32. Kuhn, R., Schwenk, F., Aguet, M., Rajewsky, K. 1995. Inducible gene targeting in mice. Science. 269:1427-1429.
    View this article via: CrossRef PubMed Google Scholar
33. Rodda, S.J., McMahon, A.P. 2006. Distinct roles for Hedgehog and canonical Wnt signaling in specification, differentiation and maintenance of osteoblast progenitors. Development. 133:3231-3244.
    View this article via: CrossRef PubMed Google Scholar
34. Kawata, T., et al. 1999. Recruitment of osteoclasts in the mandibular condyle of growing osteopetrotic (op/op) mice after a single injection of macrophage colony-stimulating factor. Arch. Oral Biol. 44:81-88.
    View this article via: CrossRef PubMed Google Scholar
35. Tolar, J., Teitelbaum, S.L., Orchard, P.J. 2004. Osteopetrosis. N. Engl. J. Med. 351:2839-2849.
    View this article via: CrossRef PubMed Google Scholar
36. Jacquin, C., Gran, D.E., Lee, S.K., Lorenzo, J.A., Aguila, H.L. 2006. Identification of multiple osteoclast precursor populations in murine bone marrow. J. Bone Miner. Res. 21:67-77.
    View this article via: CrossRef PubMed Google Scholar
37. Leenen, P.J., de Bruijn, M.F., Voerman, J.S., Campbell, P.A., van Ewijk, W. 1994. Markers of mouse macrophage development detected by monoclonal antibodies. J. Immunol. Methods. 174:5-19.
    View this article via: CrossRef PubMed Google Scholar
38. Clausen, B.E., Burkhardt, C., Reith, W., Renkawitz, R., Forster, I. 1999. Conditional gene targeting in macrophages and granulocytes using LysMcre mice. Transgenic Res. 8:265-277.
    View this article via: CrossRef PubMed Google Scholar
39. Kenner, L., et al. 2004. Mice lacking JunB are osteopenic due to cell-autonomous osteoblast and osteoclast defects. J. Cell Biol. 164:613-623.
    View this article via: CrossRef PubMed Google Scholar
40. Kim, H.J., et al. 2006. Glucocorticoids suppress bone formation via the osteoclast. J. Clin. Invest. 116:2152-2160.
    View this article via: JCI CrossRef PubMed Google Scholar
41. Kurihara, N., Tatsumi, J., Arai, F., Iwama, A., Suda, T. 1998. Macrophage-stimulating protein (MSP) and its receptor, RON, stimulate human osteoclast activity but not proliferation: effect of MSP distinct from that of hepatocyte growth factor. Exp. Hematol. 26:1080-1085.
    View this article via: PubMed Google Scholar
42. Sato, M., Grasser, W. 1990. Myosin II antibodies inhibit the resorption activity of isolated rat osteoclasts. Cell. Motil. Cytoskeleton. 17:250-263.
    View this article via: CrossRef PubMed Google Scholar
43. Wu, H., et al. 2007. Down syndrome critical region-1 is a transcriptional target of nuclear factor of activated T cells-c1 within the endocardium during heart development. J. Biol. Chem. 282:30673-30679.
    View this article via: CrossRef PubMed Google Scholar
44. Sundaram, K., et al. 2007. RANK ligand signaling modulates the matrix metalloproteinase-9 gene expression during osteoclast differentiation. Exp. Cell Res. 313:168-178.
    View this article via: CrossRef PubMed Google Scholar
45. Sato, T., et al. 1997. Identification of the membrane-type matrix metalloproteinase MT1-MMP in osteoclasts. J. Cell Sci. 110:589-596.
    View this article via: PubMed Google Scholar
46. Wu, H., Peisley, A., Graef, I.A., Crabtree, G.R. 2007. NFAT signaling and the invention of vertebrates. Trends Cell Biol. 17:251-260.
    View this article via: CrossRef PubMed Google Scholar
47. Kieslinger, M., et al. 2005. EBF2 regulates osteoblast-dependent differentiation of osteoclasts. Dev. Cell. 9:757-767.
    View this article via: CrossRef PubMed Google Scholar
48. Atkins, G.J., et al. 2001. Osteoprotegerin inhibits osteoclast formation and bone resorbing activity in giant cell tumors of bone. Bone. 28:370-377.
    View this article via: CrossRef PubMed Google Scholar
49. Kim, N., et al. 2005. Osteoclast differentiation independent of the TRANCE-RANK-TRAF6 axis. J. Exp. Med. 202:589-595.
    View this article via: CrossRef PubMed Google Scholar
50. Ruocco, M.G., et al. 2005. I{kappa}B kinase (IKK){beta}, but not IKK{alpha}, is a critical mediator of osteoclast survival and is required for inflammation-induced bone loss. J. Exp. Med. 201:1677-1687.
    View this article via: CrossRef PubMed Google Scholar
51. Begg, S.K., et al. 1993. Delayed hematopoietic development in osteopetrotic (op/op) mice. J. Exp. Med. 177:237-242.
    View this article via: CrossRef PubMed Google Scholar
52. Kim, K., et al. 2005. Nuclear factor of activated T cells c1 induces osteoclast-associated receptor gene expression during tumor necrosis factor-related activation-induced cytokine-mediated osteoclastogenesis. J. Biol. Chem. 280:35209-35216.
    View this article via: CrossRef PubMed Google Scholar
53. Asagiri, M., Takayanagi, H. 2007. The molecular understanding of osteoclast differentiation. Bone. 40:251-264.
    View this article via: CrossRef PubMed Google Scholar
54. Mizuno, A., et al. 1998. Severe osteoporosis in mice lacking osteoclastogenesis inhibitory factor/osteoprotegerin. Biochem. Biophys. Res Commun. 247:610-615.
    View this article via: CrossRef PubMed Google Scholar
55. Li, J., et al. 2000. RANK is the intrinsic hematopoietic cell surface receptor that controls osteoclastogenesis and regulation of bone mass and calcium metabolism. Proc. Natl. Acad. Sci. U. S. A. 97:1566-1571.
    View this article via: CrossRef PubMed Google Scholar
56. Yun, T.J., et al. 1998. OPG/FDCR-1, a TNF receptor family member, is expressed in lymphoid cells and is up-regulated by ligating CD40. J. Immunol. 161:6113-6121.
    View this article via: PubMed Google Scholar
57. Schoppet, M., et al. 2007. Osteoprotegerin expression in dendritic cells increases with maturation and is NF-kappaB-dependent. J. Cell. Biochem. 100:1430-1439.
    View this article via: CrossRef PubMed Google Scholar
58. Naito, A., et al. 1999. Severe osteopetrosis, defective interleukin-1 signalling and lymph node organogenesis in TRAF6-deficient mice. Genes Cells. 4:353-362.
    View this article via: CrossRef PubMed Google Scholar
59. Gohda, J., et al. 2005. RANK-mediated amplification of TRAF6 signaling leads to NFATc1 induction during osteoclastogenesis. EMBO J. 24:790-799.
    View this article via: CrossRef PubMed Google Scholar
60. Baksh, S., et al. 2002. NFATc2-mediated repression of cyclin-dependent kinase 4 expression. Mol. Cell. 10:1071-1081.
    View this article via: CrossRef PubMed Google Scholar
61. Hodge, M.R., et al. 1996. Hyperproliferation and dysregulation of IL-4 expression in NF-ATp-deficient mice. Immunity. 4:397-405.
    View this article via: CrossRef PubMed Google Scholar
62. Tsytsykova, A.V., Goldfeld, A.E. 2000. Nuclear factor of activated T cells transcription factor NFATp controls superantigen-induced lethal shock. J. Exp. Med. 192:581-586.
    View this article via: CrossRef PubMed Google Scholar
63. Pettit, A.R., et al. 2001. TRANCE/RANKL knockout mice are protected from bone erosion in a serum transfer model of arthritis. Am. J. Pathol. 159:1689-1699.
    View this article via: PubMed Google Scholar
64. Redlich, K., et al. 2002. Osteoclasts are essential for TNF-alpha-mediated joint destruction. J. Clin. Invest. 110:1419-1427.
    View this article via: JCI PubMed Google Scholar
65. Xu, X., et al. 2001. Direct removal in the mouse of a floxed neo gene from a three-loxP conditional knockout allele by two novel approaches. Genesis. 30:1-6.
    View this article via: CrossRef PubMed Google Scholar
66. Xanthoudakis, S., et al. 1996. An enhanced immune response in mice lacking the transcription factor NFAT1. Science. 272:892-895.
    View this article via: CrossRef PubMed Google Scholar
67. Lugo-Villarino, G., Maldonado-Lopez, R., Possemato, R., Penaranda, C., Glimcher, L.H. 2003. T-bet is required for optimal production of IFN-gamma and antigen-specific T cell activation by dendritic cells. Proc. Natl. Acad. Sci. U. S. A. 100:7749-7754.
    View this article via: CrossRef PubMed Google Scholar
68. Erlebacher, A., Derynck, R. 1996. Increased expression of TGF-beta 2 in osteoblasts results in an osteoporosis-like phenotype. J. Cell Biol. 132:195-210.
    View this article via: CrossRef PubMed Google Scholar
69. Takahashi, N., Udagawa, N., Tanaka, S., Suda, T. 2003. Generating murine osteoclasts from bone marrow. Methods Mol. Med. 80:129-144.
    View this article via: PubMed Google Scholar
70. Fajardo, R.J., Muller, R., Ketcham, R.A., Colbert, M. 2007. Nonhuman anthropoid primate femoral neck trabecular architecture and its relationship to locomotor mode. Anat. Rec. (Hoboken). 290:422-436.
    View this article via: PubMed Google Scholar
71. McHugh, K.P., et al. 2000. Mice lacking beta3 integrins are osteosclerotic because of dysfunctional osteoclasts. J. Clin. Invest. 105:433-440.
    View this article via: JCI CrossRef PubMed Google Scholar
72. Jones, D.C., et al. 2006. Regulation of adult bone mass by the zinc finger adapter protein Schnurri-3. Science. 312:1223-1227.
    View this article via: CrossRef PubMed Google Scholar
73. Janckila, A.J., Takahashi, K., Sun, S.Z., Yam, L.T. 2001. Naphthol-ASBI phosphate as a preferred substrate for tartrate-resistant acid phosphatase isoform 5b. J. Bone Miner. Res. 16:788-793.
    View this article via: CrossRef PubMed Google Scholar
74. Smith, C.W., Goldman, A.S. 1972. Effects of concanavalin A and pokeweed mitogen in vivo on mouse peritoneal macrophages. Exp. Cell Res. 73:394-398.
    View this article via: CrossRef PubMed Google Scholar
75. Loots, G., Ovcharenko, I. 2007. ECRbase: database of evolutionary conserved regions, promoters, and transcription factor binding sites in vertebrate genomes. Bioinformatics. 23:122-124.
    View this article via: CrossRef PubMed Google Scholar
76. Glass (2nd), D.A., et al. 2005. Canonical Wnt signaling in differentiated osteoblasts controls osteoclast differentiation. Dev. Cell. 8:751-764.
    View this article via: CrossRef PubMed Google Scholar
77. Hirotani, H., Tuohy, N.A., Woo, J.T., Stern, P.H., Clipstone, N.A. 2004. The calcineurin/nuclear factor of activated T cells signaling pathway regulates osteoclastogenesis in RAW264.7 cells. J. Biol. Chem. 279:13984-13992.
    View this article via: CrossRef PubMed Google Scholar
78. Wang, X., Seed, B. 2003. A PCR primer bank for quantitative gene expression analysis. Nucleic Acids Res. 31:e154.
    View this article via: PubMed Google Scholar