Nur77 prevents excessive osteoclastogenesis by inducing ubiquitin ligase Cbl-b to mediate NFATc1 self-limitation

Osteoclasts are bone-resorbing cells essential for skeletal remodeling. However, over-active osteoclasts can cause bone-degenerative disorders. Therefore, the level of NFATc1, the master transcription factor of osteoclast, must be tightly controlled. Although the activation and amplification of NFATc1 have been extensively studied, how NFATc1 signaling is eventually resolved is unclear. Here, we uncover a novel and critical role of the orphan nuclear receptor Nur77 in mediating an NFATc1 self-limiting regulatory loop to prevent excessive osteoclastogenesis. Nur77 deletion leads to low bone mass owing to augmented osteoclast differentiation and bone resorption. Mechanistically, NFATc1 induces Nur77 expression at late stage of osteoclast differentiation; in turn, Nur77 transcriptionally up-regulates E3 ubiquitin ligase Cbl-b, which triggers NFATc1 protein degradation. These findings not only identify Nur77 as a key player in osteoprotection and a new therapeutic target for bone diseases, but also elucidate a previously unrecognized NFATc1→Nur77→Cblb—•NFATc1 feedback mechanism that confers NFATc1 signaling autoresolution. DOI: http://dx.doi.org/10.7554/eLife.07217.001


Introduction
Bone is a dynamic organ that replaces itself every 10 years in humans (Office of the Surgeon General (US), 2004). It needs to maintain an optimal density to carry out important support, movement, and protective functions in organisms. The density of bone is tightly regulated by the bone-forming osteoblasts and the bone-resorbing osteoclasts. Osteoblasts come from the mesenchymal precursor cells, whereas osteoclast precursors come from monocyte-macrophage lineage cells. Upon RANKL signaling, osteoclast precursor cells fuse and become multi-nucleated giant cells that can degrade both the organic and inorganic tissues of the bone. Osteoclast dysregulation has been associated with several human skeletal diseases such as osteoporosis, rheumatoid arthritis, and cancer metastasis to the bone (Novack and Teitelbaum, 2008). Drugs that can inhibit osteoclast differentiation, activity, or survival have been shown to be effective against these diseases (Drake et al., 2008;Body et al., 2010).
NFATc1 is a key transcriptional switch that activates osteoclastogenesis. Ectopic NFATc1 expression alone in osteoclast precursors is sufficient to produce mature osteoclasts, whereas NFATc1 deletion blocks the ability of the precursors to differentiate into osteoclasts (Takayanagi et al., 2002). NFATc1 binds to its response elements containing a consensus sequence of GGAAA, and its target genes in osteoclasts include cathepsin K, CLC-7 chloride channel, vacuolar proton pump subunit Atp6v0d2, etc (Kuroda and Matsuo, 2012). NFATc1 has been shown to auto-amplify during osteoclast differentiation, and this auto-amplification process has been suggested to be important for osteoclast lineage commitment (Asagiri et al., 2005). However, so far very few studies have dealt with whether and how NFATc1 signaling is attenuated upon its initial activation. As a result, despite the crucial functions of NFATc1 in osteoclastogenesis, the mechanisms for how NFATc1 signaling is resolved to prevent excessive osteoclast differentiation are still incompletely understood.
Nur77 (encoded by Nr4a1), also known as nerve growth factor IB (NGFIB), TR3 or NAK-1, is an orphan nuclear receptor in the Nr4a family, which also includes Nurr1 (Nr4a2) and Nor-1 (Nr4a3). Unlike other nuclear receptors whose functions are mainly modulated by their ligands, Nur77 is mostly regulated at the transcriptional and post-transcriptional levels (Zhao and Bruemmer, 2010). Nur77 can bind to NurRE or NBRE as monomer, homodimer, or heterodimer (Philips et al., 1997). Nur77 has been implicated in a variety of physiological processes, including thymocyte negative selection, hypothalamic-pituitary adrenal axis, chronic inflammation, and vascular smooth muscle cell proliferation (Hsu et al., 2004). Nonetheless, it is unknown whether Nur77 can directly regulate skeletal homeostasis or bone cell differentiation. In this study, we uncover a novel and important role of Nur77 in NFATc1 protein degradation during osteoclastogenesis and bone resorption, thus revealing a previously unrecognized mechanism that is essential for the resolution of NFATc1 signaling in which NFATc1 exerts self-limitation via an NFATc1→Nur77→Cblb-•NFATc1 negative feedback loop.

Nur77 deletion enhances osteoclast differentiation
To examine the expression pattern of Nur77 during osteoclast differentiation, we treated bone marrow-derived osteoclast precursor cells with RANKL for 4 days ( Figure 1A). A time course analysis showed that Nr4a1 mRNA started to rise on day 2 during osteoclastogenesis ( Figure 1B). The other two members of the NR4A family, Nr4a2 and Nr4a3, were either not expressed or expressed at much lower levels ( Figure 1B). To determine if there is any potential regulatory role of Nur77 in osteoclastogenesis, we compared osteoclast differentiation cultures from bone marrow eLife digest Bones are constantly remodeled in response to the stresses of everyday life. Cells called osteoclasts break down old or damaged bone and cells called osteoblasts make new bone. In healthy bones, the work of these two types of cells is well balanced. But in bone-weakening diseases like osteoporosis and certain bone cancers this balance is disturbed and the osteoclasts become overly active, leading to weak and thin bones.
Some drugs can help block the development of osteoclasts and help reduce bone loss in these diseases, but they may cause unwanted side effects. A better understanding of the processes that maintain a healthy balance of osteoblasts and osteoclasts could help scientists develop better treatments with fewer side effects. Scientists have already learned that a protein called NFATc1 turns on the production of osteoclasts. But no one knew how NFATc1 is turned off in healthy bone to prevent the excessive growth of osteoclasts and too much bone turnover. Now, Li et al. have identified a protein called Nur77 as an important regulator of NFATc1 by examining genetically engineered mice that lack Nur77. These modified mice had more osteoclasts and thinner bones than normal mice. Further experiments used radiation to wipe out the bone marrow of normal mice, who then received bone marrow transplants from mice that lacked Nur77. After the transplant, the normal mice showed bone loss. When the experiment was reversed, and Nur77-lacking mice received bone marrow from normal mice, their bone loss was alleviated. This indicates that Nur77 acts in the bone marrow cells to control osteoclasts and skeletal health.
Li et al. found that Nur77 cannot control the expression of the gene that encodes NFATc1 or directly bind to the NFATc1 protein. Instead, Nur77 increases the production of an enzyme that breaks down the NFATc1 protein. Unexpectedly, the experiments also found that NFATc1 turns on the expression of Nur77. This means that NFATc1 essentially regulates itself by increasing its own breakdown when NFATc1 levels increase. This helps to explain how osteoclast production is normally kept in check, and may suggest new strategies for treating bone diseases. hematopoietic progenitors of Nur77 knockout (Nur77 KO) mice and WT littermate controls. The results revealed an enhanced osteoclast differentiation in Nur77 KO cultures shown by the higher expression of osteoclast differentiation markers such as tartrate-resistant acid phosphatase (Trap), cathepsin K (Ctsk), calcitonin receptor (Calcr), carbonic anhydrase 2 (Car2), as well as the master osteoclastogenic transcription factor Nfatc1 ( Figure 1C). Consistent with these results, we have also observed more and larger mature osteoclasts in the differentiation cultures ( Figure 1D), as well as higher resorptive activity ( Figure 1D). The expression of pro-and anti-apoptotic genes was comparable, indicating an unaffected osteoclast apoptosis ( Figure 1E). In contrast, osteoblast differentiation from Nur77 KO bone marrow mesenchymal progenitors was unaltered, shown by the similar induction of osteoblast markers such as collagen, type I, alpha 1 (Col1a1), and Osteocalcin ( Figure 1F). These results suggest that Nur77 may specifically suppress osteoclastogenesis during bone remodeling.
ELISA analyses showed that the serum bone resorption marker C-terminal telopeptide fragments of the type I collagen (CTX-1) was 3.7-fold higher in Nur77 KO mice ( Figure 1P), whereas the serum bone formation marker N-terminal propeptide of type I procollagen (P1NP) was unchanged ( Figure 1Q). Consistent with these observations, histomorphometry of the femur showed that osteoclast surface and osteoclast number were significantly increased in Nur77 KO mice ( Figure 1R), whereas osteoblast surface and osteoblast number ( Figure 1S) were unaltered. Together, these results suggest that Nur77 deletion causes low bone mass primarily through increasing osteoclastogenesis and bone resorption.

Nur77 regulation of bone resorption is intrinsic to the Osteoclast lineage
Because Nur77 has been implicated to regulate many other cell types (Hsu et al., 2004), we next performed bone marrow transplantation experiments to examine whether Nur77 regulation of bone resorption stems from the intrinsic effects in the hematopoietic/osteoclast lineage or non-autonomous effects from other tissues or cell types such as osteoblasts, osteocytes, or the neuroendocrine system. Two complimentary sets of bone marrow transplantation were performed, and serum bone markers were assessed two months later. In the first set, we harvested donor bone marrow cells from both WT and Nur77 KO mice, and transplanted them to irradiated WT recipient mice ( Figure 2A). The results showed that WT mice receiving Nur77 KO bone marrow cells exhibited significantly higher CTX-1 levels than the control group ( Figure 2B), but unaltered P1NP levels ( Figure 2C), suggesting that Nur77 KO hematopoietic lineage was sufficient to elevate bone resorption. In the second set, we transplanted WT donor bone marrow cells into either Nur77 KO or WT control recipient mice ( Figure 2D). Nur77 KO mice receiving WT bone marrow cells showed normalized CTX-1 levels similar to the control group ( Figure 2E), with also similar P1NP levels ( Figure 2F), suggesting that WT bone marrow can completely rescue the osteoclast defects in the Nur77 KO mice and thus other tissues/cell types play only a minor role if any. The results from these two experiments indicate that Nur77 regulation of bone resorption is intrinsic to the hematopoetic lineage.
In the bone milieu, osteoblasts and osteocytes provide RANKL and the RANKL decoy receptor OPG to stimulate and inhibit osteoclast differentiation, respectively (Boyce and Xing, 2007). Thus, we assessed whether RANKL and OPG levels were different between Nur77 KO and WT control mice. We first compared osteoblast differentiation cultures derived from Nur77 KO or WT control mice and found that there was no difference in the expression of Rankl or Opg, resulting in a comparable Rankl/ Opg ratio ( Figure 2G). Recently osteocytes, the mature long living osteoblasts embedded in the bone matrix, have been shown to provide the majority of RANKL and OPG to osteoclasts (Nakashima et al., 2011;Xiong et al., 2011). Hence, we compared Rankl and Opg expression in femur shafts that contain mainly osteocytes. Our results showed that there was no difference in Rankl expression ( Figure 2H); however, Nur77 KO osteocytes express a markedly higher level of Opg ( Figure 2H), leading to a significantly lower Rankl/Opg ratio ( Figure 2H). The lower Rankl/Opg ratio in the Nur77 KO mice is unlikely to be the cause of the augmented osteoclastogenesis, but presumably an attempt of osteocytes to suppress the osteoclast over-activation. These results, coupled with unaltered in vitro osteoblast differentiation ( Figure 1F) and in vivo bone formation ( Figure 1Q,S), suggest that osteoblast/osteocyte are not major contributors to the excessive osteoclastogenesis and bone resorption in Nur77 KO mice.
Inflammatory cytokines in the bone microenvironment can also promote osteoclast differentiation (Zupan et al., 2013). Thus, we collected bone marrow cells from mouse femurs for gene expression assessment. No significant difference was found in the expression of Tnfα, IL1β, or IL6 ( Figure 2I). Moreover, it has been shown that serum levels of TNFa, IL1, and IL6 were unchanged in mice injected with either Nur77-expressing lentivirus or Nur77 siRNA (Hu et al., 2014), further suggesting that the augmented osteoclastogenesis in Nur77 KO mice is not due to differential levels of these inflammatory cytokines.
The above findings, together with the fact that Nur77 KO osteoclast precursors exhibited enhanced osteoclast differentiation independent of the bone and neuroendocrine environment ( Figure 1C-D), indicate that Nur77 regulation of osteoclastogenesis is mainly intrinsic and cellautonomous to the osteoclast lineage.

Nur77 promotes NFATc1 degradation
Given that Nur77 exerts functions within the osteoclast itself, we decided to investigate whether Nur77 affects RANKL signaling pathways. RANKL binding to RANK receptor on osteoclast precursor cells activates AP1 transcription factor via c-Jun phosphorylation, as well as NFκB transcription factor via IκBα degradation, which in turn induces and initiates the autoamplification of NFATc1, the master transcriptional switch of osteoclastogenesis (Kuroda and Matsuo, 2012). Although Nfatc1 mRNA level was 2-fold higher in Nur77 KO osteoclast differentiation cultures on day 3 ( Figure 1C), we found that NFATc1 protein level was sevenfold higher in Nur77 KO cultures compared to WT control cultures on day 3 ( Figure 3A). It has been reported that NFATc1 protein is degraded on day 3-4 during osteoclast differentiation, despite the continuously rising Nfatc1 mRNA levels (Kim et al., 2010). Indeed, our time course analysis showed that in WT cultures, NFATc1 protein was elevated on day 2 but then rapidly down-regulated on day 3 and day 4 ( Figure 3A). In contrast, in Nur77 KO cultures, the initial increase of NFATc1 protein was sustained and NFATc1 protein remained high on day 3 and day 4 ( Figure 3A). We then compared c-Jun phosphorylation and IκBα degradation in the osteoclast differentiation cultures upon RANKL stimulation, but did not observe any significant difference between Nur77 KO and WT control cultures ( Figure 3B), which is in agreement to the similar NFATc1 protein induction on day 1 and day 2 ( Figure 3A). Moreover, there is no Nur77 binding sequence in a 4-Kb region of NFATc1 promoter. These results indicate that Nur77 regulation of NFATc1 protein level resides downstream of transcription.
It has been shown that the decrease in NFATc1 protein levels at later stage of osteoclast differentiation is due to ubiquitin-mediated protein degradation; and that MG132, a proteasome inhibitor, can restore NFATc1 protein to a similar level on day 2 (Kim et al., 2010). To examine whether the differences in NFATc1 levels between Nur77 KO and WT mice were due to protein degradation, we treated osteoclast differentiation cultures with MG132 on day 3. As our result shows, MG132 treatment increased NFATc1 protein level in WT cultures to a level similar to Nur77 KO cultures; and MG132 treatment could no longer further increase NFATc1 protein level in the Nur77 KO cultures ( Figure 3C). In line with these observations, our co-IP experiments reveal that Nur77 (either endogenous or flag-tagged) does not directly interact with NFATc1 (not shown), suggesting that Nur77 does not directly modulate NFATc1's localization or activity. These results indicate that Nur77 deletion elevates NFATc1 protein levels by suppressing ubiquitin degradation pathway.
As a complementary gain-of-function approach, we tested whether Nur77 over-expression could promote NFATc1 protein degradation. We transfected HEK293 cells with NFATc1 together with Nur77 or a GFP control, and then quantified Nfatc1 mRNA and protein levels. The result shows that Nur77 over-expression significantly decreased NFATc1 protein levels ( Figure 3D, left) without altering Nfatc1 mRNA levels ( Figure 3D, right). Consistent with the lower Nur77 protein abundance, Nur77 over-expression also dosage-dependently reduced the NFATc1 transcriptional output from a luciferase reporter driven by NFATc1 response elements ( Figure 3E). This Nur77 reduction of NFATc1 activity was completely abolished by MG132 ( Figure 3F), indicating that it was mediated by ubiquitin degradation pathway. The mRNA expression of Nr4a1 and Nfatc1 in osteoclasts is comparable on day 2-3 ( Figure 1B), supporting that Nur77 regulation of NFATc1 is relevant in the osteoclast. These findings further support the notion that Nur77 promotes NFATc1 protein degradation.

Nur77 transcriptionally up-regulates E3 ligase Cbl-b
In the ubiquitin degradation pathway, E3 ligases are responsible for substrate specificity and ubiquitination regulation. We next searched for E3 ligases that could be responsible for NFATc1 degradation in osteoclasts. It has been reported that Cbl-b, an E3 ligase in the Cbl family, is a major contributor to the ubiquitin-mediated down-regulation of NFATc1 at late stage of osteoclast differentiation (Kim et al., 2010). Therefore, we tested the hypothesis that Nur77 may promote NFATc1 degradation by inducing Cbl-b. We found that Cblb expression was significantly lower in Nur77 KO osteoclast differentiation cultures compared to WT control cultures on day 2 and day 4 ( Figure 4A). Conversely, Nur77 over-expression in HEK293 cells significantly increased Cblb expression ( Figure 4B). Importantly, a truncated Nur77 mutant in which the DNA binding domain (DBD) was deleted could no longer up-regulate Cblb, suggesting that Nur77 induction of Cblb transcription depends on its DNA binding ability ( Figure 4B). The functional connection between Nur77 and Cbl-b is further supported by the similar bone phenotype in Nur77 KO mice ( Figure 1) and Cblb KO mice (Nakajima et al., 2009), including increased ex vivo osteoclast differentiation and in vivo bone resorption, but unaltered bone formation, leading to lower bone mass.
To investigate whether Cblb is a direct Nur77 target gene, we tested whether Nur77 could transcriptionally activate the Cbl-b promoter. We cloned a 1-kb segment of Cbl-b promoter upstream of a luciferase reporter and tested its expression in a transient transfection assay in HEK293 cells. The result showed that co-transfection with Nur77 significantly up-regulated the Cbl-b promoter activity by 2.2-fold ( Figure 4C), suggesting that Nur77 is able to directly activate Cblb transcription. Bioinformatic analyses revealed a pair of motifs that may comprise a Nur77 response element (NurRE) in the Cbl-b promoter at ∼600 bp upstream of the transcription start site ( Figure 4C, inset). To examine whether these putative NurRE motifs are important for Nur77 induction of Cbl-b promoter, we mutated each NurRE motif to derive mutant-1 and mutant-2 luciferase reporters ( Figure 4C, inset). Both mutant reporters exhibited a significantly compromised ability to be activated by Nur77 compared to the WT reporter ( Figure 4C), indicating that both NurRE motifs are functionally required. To determine whether Nur77 can bind to the endogenous Cbl-b NurRE, we performed chromatin immunoprecipitation (ChIP) assay. Nur77 was found to be enriched at the NurRE region, leading to transcription activation shown by the presence of H3K4Me3 histone mark at the transcription start site ( Figure 4D). These results suggest that Nur77 can directly induce Cbl-b transcription by binding to a NurRE in the Cbl-b promoter.
We next sought to elucidate whether Nur77 induction of Cbl-b is functionally required for Nur77 down-regulation of NFATc1 protein. Instead of deleting Cblb, we designed a more prudent strategy to specifically disrupt the NurRE region in the endogenous Cbl-b promoter using CRISPR/Cas9 genome editing tool, thus more precisely dissecting the functional interaction among Nur77, Cbl-b, and NFATc1 ( Figure 4E). Compared with WT control cells, the ability of Nur77 to increase Cblb mRNA ( Figure 4F), as well as to decrease NFATc1 protein level ( Figure 4G) and transcriptional output ( Figure 4H), was significantly attenuated in two independent CRISPR mutant clones. These results provide strong evidence that Nur77 promotes NFATc1 protein degradation by directly inducing the transcription of Cbl-b E3 ligase.

NFATc1 up-regulates Nur77 to form a self-limiting loop
Since Nur77 expression consistently rises during osteoclastogenesis ( Figure 1B), we hypothesize that there is an upstream regulator that induces Nur77 transcription upon RANKL signaling activation. Interestingly, we found several NFATc1 response elements in the Nur77 promoter region, suggesting that NFATc1 up-regulates Nur77 to initiate a negative feedback loop. To test whether NFATc1 itself is sufficient to increase Nur77 expression independent of other RANKL signaling pathways, we performed transfection assays to over-express NFATc1. Compared to a GFP negative control, NFATc1 over-expression significantly increased Nur77 expression in both HEK293 cells and mouse myoblast C2C12 cells ( Figure 5A). Conversely, treatment of osteoclast differentiation cultures with cyclosporin A, a calcineurin inhibitor that suppresses NFATc1 activity, dosage-dependently decreased Nur77 expression ( Figure 5B).
We next examined whether NFATc1 could directly activate Nur77 promoter. We cloned a 0.8 Kb Nur77 promoter region upstream of a luciferase reporter and tested its inducibility by NFATc1 by transient transfection. Compared to a GFP negative control, NFATc1 over-expression significantly elevated the luciferase output ( Figure 5C). Moreover, ChIP assay showed that RANKL treatment of osteoclast differentiation cultures markedly increased NFATc1 binding to the endogenous Nur77 promoter ( Figure 5D), leading to activated Nur77 transcription as shown by the higher level of H3K4Me3 histone mark at the transcription start site ( Figure 5D). Together, these results indicate that Nur77 is a direct transcriptional target of NFATc1, and thus revealing a key mechanism for how NFATc1 resolves its own signaling to prevent excessive osteoclastogenesis via an NFATc1→-Nur77→Cblb-•NFATc1 self-limiting loop ( Figure 5E).

Cbl-b deletion abolishes Nur77 regulation of osteoclastogenesis and bone resorption
To further determine whether Cbl-b is required for the anti-osteoclastogenic function of Nur77 in vivo, we conducted genetic experiments by comparing Nur77 Cblb double knockout mice (DKO) with Nur77 KO mice, Cblb KO mice, and WT littermate controls. Ex vivo bone marrow osteoclast differentiation assay showed that Nur77 deletion could no longer further enhance osteoclast differentiation in the absence of Cbl-b; furthermore, Cblb KO cultures and Nur77 Cblb DKO cultures showed a similar enhanced osteoclastogenesis as Nur77 KO cultures compared with WT control cultures ( Figure 6A-E). In accordance to this ex vivo finding, in vivo analyses showed that when Cblb is absent, Nur77 deletion could no long further elevate serum bone resorption ( Figure 6F) or reduce bone mass ( Figure 6G-K); Cblb KO mice and Nur77 Cblb DKO mice showed a similar high bone resorption and low bone mass phenotype as Nur77 KO mice compared with WT controls ( Figure 6F-K). These findings demonstrate that Cbl-b deletion fully recapitulates Nur77 deletion, and Cbl-b deletion completely abolishes Nur77 regulation of osteoclastogenesis and bone resorption. Therefore, this in vivo genetic evidence strongly supports Cbl-b as a major and essential mediator of Nur77 function in the osteoclast lineage.

Discussion
In this study, we have identified the nuclear receptor Nur77 as a critical negative regulator of osteoclastogenesis and bone resorption, revealing its novel bone protective role. Nur77 deletion causes elevated bone resorption and bone loss in mice. Moreover, we have also unraveled a previously unrecognized mechanism for how Nur77 attenuates NFATc1 signaling at late stage of osteoclast differentiation. Nur77 transcriptionally up-regulates Cbl-b E3 ligase to trigger NFATc1 protein degradation, so that NFATc1 signaling can be resolved in a timely fashion to prevent excessive osteoclastogenesis and bone resorption ( Figure 5E).
When osteoclast differentiation is enhanced in a culture, typically all RANKL-induced osteoclastogenic genes will be elevated, including Nfatc1, because the gene expression analysis is a populationbased assay ( Figure 1C). In addition to induction by RANKL-activated upstream signaling, NFATc1 also auto-amplifies itself at mRNA level (Asagiri et al., 2005), which makes it harder to discern the origin of NFATc1 regulation in osteoclast cultures. Thus, we subsequently examined the intrinsic ability of Nur77 to regulate NFATc1 via transfection in a heterologous cell type such as HEK293 cells, in the absence of the other RANKL signaling in osteoclast cultures. We show that Nur77 can decrease NFATc1 protein levels without affecting Nfatc1 mRNA levels ( Figure 3D). Also, there is no Nur77 binding sequence in NFATc1 promoter, further support that Nur77 does not directly regulate NFATc1 transcription. Moreover, our co-IP experiment show that Nur77 (either endogenous or a flag-tagged) does not directly interact with NFATc1, suggesting that Nur77 does not directly modulate NFATc1 localization or activity. As a result, we conclude that Nur77 mainly regulates NFATc1 protein degradation.
In searching for the Nur77-induced NFATc1-targeting E3 ubiquitin ligase, we have considered both of the two members in the Cbl family-Cbl and Cbl-b. Findings from our study and previous studies indicate that Cbl-b, but not Cbl, is the major regulator in osteoclast due to the following reasons: (1) previous studies and our data show that Cbl-b KO mice exhibit enhanced osteoclast differentiation, higher bone resorption and lower bone mass that is similar to the phenotype we observed in Nur77 KO mice (Nakajima et al., 2009) (Figure 6). In contrast, adult Cbl KO mice have no obvious bone phenotype (Chiusaroli et al., 2003). This supports that Cbl-b, but not Cbl, is a physiologically significant regulator of osteoclast in vivo. (2) Nur77 over-expression induces the expression of Cbl-b, but not Cbl. (3) A complete NurRE exists only in Cbl-b promoter, but not in Cbl promoter. (4) Nur77 KO osteoclast cultures show a significant lower level of Cbl-b, but not Cbl. (5) Finally, our in vivo genetic data comparing Nur77 Cblb DKO mice with Cblb KO mice show that Cbl-b deletion completely abolishes the enhanced osteoclast differentiation, elevated bone resorption, and reduced bone mass in Nur77 KO mice, demonstrating that Cbl-b is the major and essential mediator of Nur77 regulation of osteoclastogenesis, which is not compensated by Cbl.
In the process of studying the role of Nur77 in NFATc1 regulation and osteoclast differentiation, we inadvertently discovered that Nur77 is not only a regulator of NFATc1 but also a transcriptional target of NFATc1, thus revealing a negative feedback loop where NFATc1 induces its own degradation by up-regulating Nur77 and Cbl-b. This mechanism is crucial for proper cellular differentiation and function, since the resolution of signaling is just as important as its initiation and amplification. A breach in the NFATc1→Nur77→Cblb-•NFATc1 regulatory loop, exemplified by the Nur77 KO mice, will cause pathologically elevated NFATc1 levels during late stage of osteoclastogenesis and send osteoclasts into overdrive. To the best of our knowledge, we are the first group to propose an NFATc1 self-limiting regulatory mechanism. Therefore, NFATc1 exerts three functions to control osteoclastogenesis. In addition to the previously recognized roles in its auto-amplification and activating osteoclast genes to initiate the differentiation, NFATc1 also plays a role in its autoresolution to cease the differentiation ( Figure 5E). Our findings will pave the road for future investigations to examine whether this NFATc1→Nur77→Cblb-•NFATc1 negative feedback loop may be widely applicable to NFATc1 regulation of other cellular processes such as T cell activation and cancer development.
Identification of novel osteoclast signaling pathways provides insights into potential new therapeutic options to treat bone-degenerative diseases by inhibiting osteoclast activity. Current clinically approved osteoclast inhibitors, such as bisphosphonates and denosumab (anti-RANKL antibody), may cause severe side effects such as osteonecrosis of the jaw (ONJ) (Khan et al., 2009;Boquete-Castro et al., 2015). These side effects could stem from a variety of factors including target specificity. The whole body Nur77 KO mice, however, provides an example where despite the wide spread expression of a gene, it still can be targeted due to the differential sensitivity in different tissues. Although Nur77 has been implicated in numerous physiological functions in vitro, most of these functions did not hold true in vivo based on a general lack of phenotype in Nur77 KO mice (Lee et al., 1995;Chao et al., 2013). Until recently, Nur77 KO mice largely appear healthy and normal, and only exhibit clinical deficiencies under severe stress (Chao et al., 2009;Palumbo-Zerr et al., 2015) or with the deletion of an additional NR4A gene (Mullican et al., 2007). This may be partially due to functional redundancy among NR4A family members so that the compensation by Nurr1 and Nor1 masks the effects of Nur77 loss. Interestingly, we found that Nur77 is the predominant NR4A member in the osteoclast lineage with little or no Nurr1 or Nor1 expression ( Figure 1B), explaining the critical role of Nur77 in osteoclastogenesis so that the effects on bone is evident by Nur77 deletion alone. This creates an exciting opportunity for selective drug targeting and precision medicine with minimal side effects. Most recently, Tontonoz et al. have uncovered a muscle protective role of Nur77 as mice deficient in Nur77 alone exhibit reduced muscle mass and myofiber size (Tontonoz et al., 2015). Therefore, Nur77 activation may represent a promising therapeutic strategy for musculoskeletal degenerative diseases with dual benefits on muscle and bone.
The crucial regulation of NFATc1 protein degradation by Nur77 and Cbl-b suggests that it may be therapeutically beneficial to accelerate RANKL signaling resolution during osteoclastogenesis. Indeed, defects in the components of ubiquitin and proteasome system have been implicated in diseases including cancer and neurodegenerative disorders (Popovic et al., 2014). Bortezomib, a peptide inhibitor of proteasome, has been approved for clinical usage in pathological settings such as refractory multiple myeloma (Richardson et al., 2005). The discovery of pathway-specific ubiquitin-proteasome activators, however, is somewhat lagging behind. Nonetheless, oleuropein, a small molecule proteasome activator, has been shown to delay replicative senescence of human embryonic fibroblast (Katsiki et al., 2007). In addition, oleuropein treatment has been shown to inhibit osteoclast formation and suppress the loss of trabecular bone in ovariectomized mice (Hagiwara et al., 2011), giving hope that small molecules that selectively activate the protein degradation pathway may be a promising future therapeutic strategy for skeletal and other diseases.

Mice
Nur77 KO mice (Lee et al., 1995) in a C57BL/6 and 129SvJ hybrid background was originally generated by Jeffrey Milbrandt at Washington University School of Medicine and kindly provided by Orla Conneely at Baylor College of Medicine. Cblb KO mice in a mixed background were originally generated by Josef Penninger (Bachmaier et al., 2000). Mice were fed with standard chow ad libitum and kept on a 12-hr light, 12-hr dark cycle. Nur77 KO mice were bred with Cblb KO mice to generate Nur77 Cblb double heterozygous mice, which were then bred to generate littermates for Nur77 KO, Cblb KO, Nur77 Cblb DKO, and WT control. All experiments were conducted using littermates. Bone marrow transplantation was performed as described Krzeszinski et al., 2014). Briefly, bone marrow cells from 2-month-old male donor (WT or Nur77 KO) were intravenously transplanted via retro orbital injection into 2-month-old male recipients (WT or Nur77 KO) that were irradiated at lethal dose (1000 roentgen); the mice were analyzed 3 month post transplantation. Our established protocol for lethal irradiation and bone marrow transplantation achieves >95% repopulation of donor cells in the recipient mice, as measured by the percentage of CD45.1 vs CD45.2. Sample size estimate was based on power analyses performed using SAS 9.3 TS X64_7PRO platform at the UTSW Biostatistics Core. With the observed group differences and the relatively small variation of the in vivo measurements, n = 4 and n = 3 will provide >90% and >80% power at type I error rate of 0.05 (two-sided test), respectively. All protocols for mouse experiments were approved under number 2008-0324 by the Institutional Animal Care and Use Committee of UTSW.
Laboratory. Both vectors for gRNAs and the expression plasmids for hCas9 and GFP marker were cotransfected into HEK293 cells. GFP + cells were sorted into 96-well plates at 1 cell/well 48 hr later. Each clone was expanded; genomic DNA was amplified by PCR and genotyped by sequencing. Two independent clones with NurRE deletion were compared to WT control.