BoneKEy-Osteovision | Commentary

New secrets unveiled by the big ‘N FAT osteoclast



DOI:10.1138/2003071

Commentary on: Takayanagi H, Kim S, Koga T, Nishina H, Isshiki M, Yoshida H, Saiura A, Isobe M, Yokochi T, Inoue J, Wagner EF, Mak TW, Kodama T, Taniguchi T. Induction and activation of the transcription factor NFATc1 (NFAT2) integrate RANKL signaling in terminal differentiation of osteoclasts. Dev Cell. 2002 Dec;3(6):889-901.

The recent paper from Taniguchi's group () on the role of nuclear factor of activated T cells-c1 (NFATc1) in osteoclast differentiation is interesting on several accounts: (a) it shows how one can productively use microarray technology; (b) it provides novel insights into the downstream signaling of receptor activator of NF-kB ligand (RANKL) that leads to osteoclast differentiation; (c) it provides the first molecular explanation for the lack of osteoclasts in c-fos null mice; (d) it points again to intracellular calcium oscillations as a critical signaling mechanism; and (e) last, but not least, it identifies NFATc1 as a necessary and sufficient transcription factor for osteoclast differentiation and adds it to the list of regulatory factors that osteoclasts share with other hemopoietic cells. It would be hard to find more information in a single paper.

A few introductory sentences on NFAT are in order; see also the review by Crabtree and Olson (). NFAT belongs to a family of transcription factors found only in vertebrates. As part of their regulatory function, NFATs shuttle between the cytoplasm and the nucleus. Nuclear import is controlled by calcium activation of the phosphatase calcineurin, which dephosphorylates an N-terminal serine and thereby unmasks the nuclear localization sequence in NFAT, resulting in its nuclear uptake. Re-phosphorylation and export is controlled by the kinase GSK3, which phosphorylates a nuclear export sequence. As most readers know, this is the mechanism by which cyclosporin and FK506 suppress the immune response to prevent rejection of transplanted organs. These agents bind to cyclophilin and FK506 binding protein, respectively, and the binding proteins associate with calcineurin and inhibit its activity, thus inhibiting nuclear translocation of NFAT and T-cell activation ().

The other background information known to most readers is that osteoclasts differentiate from hemopoietic bone marrow precursor cells of the monocyte-macrophage lineage following ligand stimulation of receptor activator of NF-kB (RANK) by RANK ligand (RANKL); see the review by (). This is an obligatory pathway for osteoclast formation, at least in mice, because deletion of RANK or RANKL by homologous recombination results in the absence of osteoclasts. RANK is known to control transcription by signaling via TNF receptor-activated factor-6 (TRAF6) to NF-kB and to Jun N-terminal kinase (JNK). But there was little information on what follows beyond that.

Takayanagi et al. () started with a microarray screen, looking for genes whose expression is upregulated by RANKL in osteoclast-generating bone marrow cells. As expected, osteoclastic genes, such as the calcitonin receptor, tartrate-resistant acid phosphatase, cathepsin K, and carbonic anhydrase II, were induced; however, in addition, NFATc1 (NFAT2) was strongly upregulated (by more than 20-fold) and translocated to the nucleus. Macrophage colony-stimulating factor (M-CSF) alone or interleukin-1 (IL-1) did not produce similar effects. RANKL also induced expression of c-Fos and Fra-1. NFATc1 upregulation was absent in cells from TRAF6 -/-, but it was also absent in Fos-/- mice. These results indicated that RANKL upregulation of NFATc1 may require RANK activation of both pathways (TRAF6/NF-kB/JNK and c-Fos) and they link, for the first time, the lack of osteoclasts in RANK/RANKL and c-Fos null mice.

As mentioned above, it had been shown, in studies of T-cell activation and muscle differentiation (), that NFAT protein translocation to the nucleus follows calcium activation of the calcium/calmodulin-dependent serine/threonine phosphatase calcineurin. Does calcium play a similar role in the RANKL effects on NFATc1? RANKL, but not IL-1, induced sustained calcium oscillations in bone marrow osteoclast precursors, but only after exposure for 24 hours. The reason for the time lag is not known. Calcium ionophores plus M-CSF failed to mimic RANKL induction and translocation of NFATc1 or osteoclast differentiation, suggesting that the oscillations were essential for sustained NFATc1 activation. Local subcellular calcium oscillations have been previously reported as part of signal transduction pathways () and may explain how this universal second messenger achieves specificity. The process of the RANKL-induced oscillations remains to be established. Calcium chelators (BAPTA-AM), as well as calcineurin inhibitors (FK506 and cyclosporin A), suppressed both NFATc1 expression and nuclear localization, emphasizing the importance of the calcium changes in RANKL action. Another interesting signal transduction loop is NFATc1 upregulation of its own transcription by positive feedback, in which c-Fos may cooperate (see below).

A major novel finding was that NFATc1 activation was essential for osteoclast differentiation. This conclusion was supported by genetic and pharmacological evidence, as well as by direct induction of osteoclast differentiation by expression of NFATc1. Since NFATc1 null mutations are embryonic lethal, the authors used embryonic stem (ES) cells to generate osteoclast precursors and deleted NFATc1 in these cells by homologous recombination. In the presence of M-CSF, these cells differentiated into monocytes/macrophages, but unlike the wild-type controls, they failed to differentiate into osteoclasts in response to RANKL. In a complementary experiment, exogenous NFATc1 was expressed, using a viral vector, in bone marrow osteoclast precursor cells. The infected cells were identified by co-expression of the fluorescent marker green fluorescent protein (GFP). About 70% of the infected cells differentiated into osteoclasts in the absence of RANKL, indicating that NFATc1 is sufficient to induce differentiation in these cells. Together these experiments offered convincing genetic evidence that NFATc1 is necessary for osteoclast differentiation. In addition, pharmacological evidence supported the rate-limiting role of NFATc1 in osteoclast differentiation, which was blocked by the calcineurin inhibitor FK506. The exogenous NFATc1-expressing cells were partially resistant to the inhibitory effects of FK506, suggesting that the major downstream action of calcium/calcineurin in this system is NFAT induction and translocation.

Where does c-Fos fit into this scheme? Many genes expressed in osteoclasts, such as tartrate-resistant acid phosphatase (TRAP), matrix metalloproteinase 9 (MMP9), cathepsin K, calcitonin receptor, and carbonic anhydrase II, contain NFAT and AP-1 (the c-Fos target) response sequences in their promoters. NFAT and AP-1 are known to cooperate in transcription activation. Indeed, NFATc1 and c-Fos exhibited synergistic effects on the TRAP and calcitonin promoters in an in vitro transcription assay, an effect not observed for a mutated NFATc1 that does not interact with c-Fos.

Together these findings support the following scheme. RANKL exposure of osteoclastogenic precursor cells causes: (a) RANK activation of TRAF6, NF-kB, and JNK; (b) calcium oscillations that inhibit calcineurin and allow NFATc1 translocation to the nucleus and further upregulation of NFATc1 expression; and (c) increased expression of c-Fos, which synergizes with NFATc1 to promote the transcription of osteoclastic genes (TRAP, calcitonin receptor, cathepsin K, etc.) and of NFATc1 itself, possibly with the cooperation of other transcription factors like NF-kB. See Figure 1.

The rate-limiting role of NFATc1 in osteoclast differentiation is totally novel and it offers a very plausible explanation for the role of c-Fos. The effect of RANKL on calcium oscillations is also new. Among the next questions in this fascinating story are the mechanisms for the effects of RANK on intracellular calcium and c-Fos and the detailed role(s) of NF-kB. On the clinical side, these findings suggest that immunosuppressive therapy targeting calcineurin would be bone-sparing, but the opposite seems to be the case. The authors attribute this discrepancy to the effects of FK506 and cyclosporin on other cells (e.g., osteoblasts and T cells). Therapeutic applications of these new insights for suppressing osteoclast generation, suggested by the authors, may take some time, given the difficulty involved in pharmacological targeting of signal transduction pathways.

In summary, Takayanagi et al. () have made a magnificent contribution to our knowledge of osteoclasts and cellular differentiation, which should delight anyone excited about science.


Creative Commons License This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 United States License.