Professional Documents
Culture Documents
Corresponding author
removal of worn out parts of bones that have been dam-
Brendan F. Boyce, MD
Department of Pathology and Laboratory Medicine, aged through normal wear and tear. Although much is
University of Rochester Medical Center, 601 Elmwood Avenue, known about what accelerates this process in pathologic
Box 626, Rochester, NY 14642, USA. states, the molecular mechanisms that control its ini-
E-mail: brendan_boyce@urmc.rochester.edu tiation, progression, and cessation at any given site in the
Current Osteoporosis Reports 2007, 5:98–104 skeleton in normal individuals remain poorly understood.
Current Medicine Group LLC ISSN 1544-1873 All pathologic conditions that affect the skeleton affect
Copyright © 2007 by Current Medicine Group LLC
normal bone remodeling, typically because they accelerate
or inhibit bone resorption or formation. These conditions
include postmenopausal osteoporosis, periodontal disease,
Understanding of osteoclast formation and activation
metastatic cancers, multiple myeloma, and rheumatoid
has advanced considerably since the discovery of the
arthritis. Patients affected by these disorders have local or
RANKL/RANK/OPG system in the mid 1990s. Osteo-
systemic alteration in the levels of hormones or proinflam-
blasts and stromal stem cells express receptor activator
matory cytokines that are known to stimulate or inhibit
of NF-LB ligand (RANKL), which binds to its receptor,
bone resorption or formation. In the early 1980s, Rodan
RANK, on the surface of osteoclasts and their precur-
and Martin [1] proposed that osteoblasts and stromal stem
sors. This regulates the differentiation of precursors
cells regulated osteoclast formation by expressing factors
into multinucleated osteoclasts and osteoclast activa-
within the bone marrow in response to agents known to
tion and survival both normally and in most pathologic
stimulate bone resorption, such as parathyroid hormone
conditions associated with increased bone resorption.
(PTH). During the past 10 years, studies of bones from
Osteoprotegerin (OPG) is secreted by osteoblasts and
genetically altered mice and from animal models of bone
osteogenic stromal stem cells and protects the skeleton
diseases have proven this hypothesis and consequently
from excessive bone resorption by binding to RANKL
have greatly increased our knowledge of the molecular
and preventing it from interacting with RANK. The
mechanisms whereby factors that regulate the formation
RANKL/OPG ratio in bone marrow is thus an important
and activity of osteoclasts work. In particular, they led to
determinant of bone mass in normal and disease states.
the identification in the mid-to-late 1990s of a signaling
RANKL/RANK signaling also regulates lymph node
system involving receptor activator of NF-LB (RANK), its
formation and mammary gland lactational hyperplasia
ligand (RANKL), and osteoprotegerin (OPG). The iden-
in mice, and OPG protects large arteries of mice from
tification of this system was a major breakthrough that
medial calcification. This article reviews the roles of the
clarified the roles of osteoblasts and stromal cells in these
RANKL/RANK/OPG system in bone and other tissues.
processes. More recently, it has become increasingly clear
that osteoclasts are not simply bone resorptive cells; they
also have important functions in regulating osteoblasts
Introduction and immune cells in pathologic states [2].
Bone has multiple functions in vertebrates, some obvious,
others less clear. These include protection of vital organs
and hematopoietic marrow, rigid support for muscles, and Regulation of Osteoclast Formation
storage and release of vital ions, especially calcium. Bone and Activation
is continuously renewed by means of a process called Osteoclasts are multinucleated cells formed by cyto-
bone remodeling, in which packets or trenches of bone are plasmic, but not nuclear, fusion of their mononuclear
removed from the surfaces of cancellous and cortical bone precursors. In most circumstances, the number of nuclei
by osteoclasts. There are more than 1 million of these correlates positively with the bone-resorbing activity
microscopic remodeling trenches in the adult skeleton at of osteoclasts. Osteoclasts are in the myeloid lineage of
any time. These trenches are subsequently filled with new hematopoietic cells that also gives rise to macrophages.
bone laid down by osteoblasts. (Other durable structures, Expression of c-Fos in osteoclast precursors (OCPs) is
such as tendons, teeth, and cartilage, are not remodeled.) required for this switch to osteoclast differentiation. c-Fos
The main function of bone remodeling is thought to be is one of a number of transcription factors activated in
The RANKL/RANK/OPG Pathway Boyce and Xing 99
response to RANKL; mice deficient in c-Fos form more which could then be cultured in the absence of osteoblasts
macrophages but no osteoclasts. Osteoclasts attach them- and stromal stem cells. The strategy for acquiring OCPs
selves firmly to bone surfaces, with extensions of their from these sources was developed in the knowledge that
cytoplasm tightly sealed to the underlying bone matrix M-CSF expression by osteoblasts and stromal stem cells
using specialized, actin-rich podosomes. Within these was required for progenitor cells to differentiate into
sealed zones, the osteoclasts form ruffled membranes, osteoclasts, but that M-CSF on its own was unable to
which resorb bone effectively by increasing the surface complete this process. This requirement for M-CSF was
area of the cell membrane, where hydrochloric acid and based on the observation that op/op mice, which do not
cathepsin K, a proteolytic enzyme, are secreted onto the express functional M-CSF, have osteopetrosis because
bone surface [3]. By this mechanism, osteoclasts simulta- they lack osteoclasts. Indeed, since 1981, when Rodan
neously dissolve the mineral and degrade the matrix of and Martin [1] proposed their hypothesis that osteoblasts
bone, at the same time protecting neighboring cells from and stromal stem cells play a central role in regulating
harm by the sealing mechanism. RANKL and C-integrin– osteoclast formation and bone resorption, many investi-
mediated signaling from bone matrix activate osteoclasts gators have attempted to identify the osteoclast-activating
[3]. Osteoclasts work in packs within remodeling units factor that completed the differentiation of precursors
under the control of cells in the osteoblast lineage. These exposed to M-CSF.
osteoblast lineage cells express macrophage colony-
stimulating factor (M-CSF), RANKL, and a growing list
of factors that regulate the formation, fusion, activity, and Identification of RANKL, RANK, and OPG
survival of osteoclasts. Four groups using different approaches discovered the
Recent studies have suggested that osteoclasts are factors expressed by osteoblasts and stromal stem cells
likely to be involved in the recruitment of packs of bone- that regulate osteoclastogenesis. OPG was discovered
forming osteoblasts to refill the trenches formed on bone unexpectedly by researchers at Amgen in studies designed
surfaces [2]. These studies examined the mechanisms to identify tumor necrosis factor receptor (TNFR)–related
whereby intermittent PTH treatment exerts its anabolic molecules that might have therapeutic utility [8]. These
effects and showed that RANKL expression is increased researchers made transgenic mice overexpressing various
by osteoblasts and stromal cells following PTH injection, TNFR-related cDNAs and found that mice overexpress-
and this increase leads to activation of existing osteoclasts, ing one particular cDNA developed marked osteopetrosis
which release one or more factors that stimulate new because they had no osteoclasts. They named the protein
bone formation. Bisphosphonate antiresorptive treatment encoded by the gene osteoprotegerin (ie, the bone pro-
appears to attenuate, rather than enhance, the anabolic tector) because it appeared to protect the skeleton from
action of PTH, at least in some studies [4,5]. Osteoclasts low bone mass and increased risk of fracture by limit-
also appear to regulate their own formation as well as ing osteoclastic bone resorption. About the same time,
immune responses at sites of inflammation in bone. They researchers at the Snow Brand Milk Products Company
also appear to positively and negatively regulate osteoblast in Japan discovered an identical molecule by purifying a
formation and activity, although the precise mechanisms factor from human embryonic fibroblasts that inhibited
whereby they do this remain to be determined [6••,7••]. osteoclastogenesis [9]. Both groups quickly identified the
During embryonic development and until growth plate ligand of this protein, using it as a probe in expression
closure, when endochondral ossification ceases, osteoclasts cloning experiments. They called it OPG ligand (OPGL)
remove bone newly formed under growth plates. This medi- and osteoclast differentiation factor (ODF), respectively
ates formation of the bone marrow cavity and facilitates [10,11]. It turned out to be identical to a member of the
normal hematopoiesis, both of which are defective if there TNF ligand family, which had been identified the year
is a failure of osteoclast formation or activity. Such failure before as RANKL [12] and TNF-related activation-
results in osteopetrosis, the most severe forms of which induced cytokine (TRANCE) [13]. Soon after OPGL/ODF
can be lethal because of attendant immunodeficiency and was identified as a ligand for OPG, the cellular receptor
increased risk of infection and recurrent fractures. The was shown to be identical to the already known RANK,
development of osteopetrosis and its attendant failure of which researchers at Immunex had discovered while they
tooth eruption in a variety of knockout mice have identi- were sequencing cDNAs from a human bone marrow–
fied essential functions of genes in osteoclast biology that derived myeloid dendritic-cell cDNA library [12]. They
had been largely unanticipated [3]. had shown that RANK had partial homology to a portion
Understanding of the molecular mechanisms that of the extracellular domain of human CD40, a member of
regulate osteoclast formation and activation has advanced the TNFR superfamily. Intriguingly, it also was involved
rapidly since the discovery of the RANKL/RANK/OPG in the activation of T cells in the immune system. They
signaling system and following the development in the late isolated RANKL by direct expression screening and,
1980s of in vitro assays that facilitated harvesting from like Wong et al. [13], found that it increased survival of
bone marrow or spleen cells of large numbers of OCPs, RANK-expressing T cells and proliferation of dendritic
100 Epidemiology and Pathophysiology
cell–stimulated naive T cells. The American Society RANKL is also expressed in mammary epithelial cells dur-
for Bone and Mineral Research [14] recommended that ing pregnancy [25••]. Its expression is required in mice for
RANKL and OPG should be the “official” names for lactational hyperplasia of mammary epithelial cells and
these proteins. The discoveries that RANKL is involved milk production [26]. It is expressed by some malignant
in osteoclast and T-cell biology have spawned the now- tumor cells, which also express RANK and thus could have
growing field of osteoimmunology. a role in inducing tumor cell proliferation [25••]. To date,
there are no published reports of spontaneous mutations in
RANKL the RANKL gene in humans or animals.
RANKL is a type II homotrimeric transmembrane pro-
tein expressed both as a membrane-bound and a secreted RANK
protein, which is derived from the membrane form as a RANK is a type I homotrimeric transmembrane protein,
result of either proteolytic cleavage or alternative splicing which is expressed widely in normal cells [27], in mam-
[15,16]. Most of the factors known to stimulate osteoclast mary glands [26], and in some cancer cells, including
formation and activity stimulate RANKL expression in breast and prostate cancers [25••,28], two tumors with
osteoblasts and stromal cells. RANKL is highly expressed high potential for bone metastasis. Activating muta-
in lymph nodes, thymus, and lung and is expressed at tions of RANK have been reported in exon 1 of RANK
low levels in a variety of other tissues, including spleen in humans, resulting in an increase in RANK-mediated
and bone marrow. It is expressed by synovial cells and is NF-LB signaling and an increase in osteoclast formation
secreted by activated T cells in inflamed joints of patients and activity. Such mutations appear to account for the
with rheumatoid arthritis. These sources of RANKL increased osteolysis seen in some patients with familial
appear to be responsible, at least in part, for mediating Paget’s disease and have confirmed the importance of
the joint destruction in affected patients [17]. However, this system in humans [29]. A potential role for RANK
T-cell production of RANKL also induces expression of in tumor cell proliferation is being investigated [25••]; if
interferon-C by activated osteoclasts through c-Fos. This proven, this could be a future target for antitumor ther-
negatively regulates osteoclast formation [18], which can apy. No humans have yet been identified with inactivating
be further enhanced by T-cell–produced interferon-B, mutations or deletions of RANK. However, a spontaneous
which degrades TNFR-associated factor 6 (TRAF6), an deletion mutation occurred in a line of transgenic mice,
essential adapter protein recruited to RANK to mediate resulting in all of the features of mice with targeted dele-
RANK signaling, as discussed later [19]. Because of these tion of RANK and confirming the importance of RANK
competing effects of RANKL signaling, it is not clear for osteoclast formation [30].
whether it plays an important role in mediating bone
destruction in inflamed joints [20]. Recently, Zhao et al. Osteoprotegerin
[7••] reported that osteoclasts and osteoblasts directly OPG is expressed widely in many tissues besides osteo-
affect one another through ephrinB2-EphB4 bidirectional blasts, including spleen, bone marrow, heart, liver, and
signaling, in which ephrinB2 ligand reverse signals into kidney [27]. A role for OPG in protection against bone
OCPs to suppress osteoclast differentiation by inhibiting loss in humans is supported by the report of homozygous
the c-Fos–nuclear factor of activated T cell c (NFATc)1 deletions of 100 kb of OPG in two patients with juvenile
cascade, whereas EphB4-receptor forward signaling Paget’s disease, an autosomal recessive disorder charac-
enhances osteoblast differentiation. terized by increased bone remodeling, osteopenia, and
TNF also mediates joint destruction in patients with fractures [31]. In another report, an inactivating deletion
inflammatory arthritis. Its concentration is increased in the in exon 3 of OPG occurred in three siblings with idio-
joints and blood of affected patients [21]. In animal models pathic hyperphosphatasia, a crippling, autosomal recessive
of inflammatory arthritis, TNF has been shown to increase bone disorder characterized by increased bone remodel-
OCP numbers in bone marrow [22•] and to increase the ing associated with kyphosis, deformities of long bones,
number of circulating OCPs by promoting their egress from and acetabular protrusion in affected children [32]. Most
the bone marrow into the peripheral blood. In inflamed factors that induce RANKL expression by osteoblasts
joints, it promotes fusion of these cells into osteoclasts, also regulate OPG expression, some positively and oth-
along with RANKL and interleukin-1 [23]. ers negatively [27]. In general, upregulation of RANKL
RANKL also stimulates the release of immature osteo- expression by cytokines and growth factors is associated
clast progenitors into the circulation. However, it does with relative downregulation of OPG, or at least lower
not induce OCP mobilization in protein tyrosine phos- induction of OPG, so that the ratio of RANKL to OPG
phatase (PTP)F–knockout mice, which have osteoclasts changes in favor of osteoclastogenesis. Although there are
with defective bone adhesion and resorption [24••]. Thus, contradictory data, many publications support the asser-
RANKL-induced osteoclast activation may regulate pro- tion that the RANKL/OPG ratio is a major determinant of
genitor recruitment as part of homeostasis and host defense, bone mass in a variety of clinical settings [33]. Bone mass
linking bone remodeling with regulation of hematopoiesis. is also influenced directly by Wnt/C-catenin signaling in
The RANKL/RANK/OPG Pathway Boyce and Xing 101
osteoblasts. This pathway regulates commitment of mes- been produced independently; both have osteopetrosis, but,
enchymal cells to the osteoblast lineage and the amount of surprisingly, one has normal numbers of osteoclasts but the
bone laid down by osteoblasts. A recent unexpected obser- cells are inactive [39], whereas the other has no osteoclasts
vation is that Wnt/C-catenin signaling directly regulates [40]. It is not clear why these two phenotypes are different,
OPG expression by osteoblastic cells [34]. Thus, two major but presumably they represent differences in the strategies
signaling pathways—OPG/RANKL/RANK in osteo- to knock out the gene. To date, six signaling pathways have
clasts and Wnt/C-catenin in osteoblasts—appear to have a been shown to be activated by RANK-mediated protein
central role in determining bone mass. Both involve osteo- kinase signaling. The three that directly mediate osteo-
blasts, the former through indirect mechanisms, whereby clastogenesis are inhibitor of NF-LB kinase (IKK)/NF-LB,
osteoblast expression of both RANKL and OPG influence c-Jun N-terminal kinase (JNK)/activator protein-1 (AP-1),
bone resorption, and the latter through direct regulation of and calcineurin/NFATc1. Others mediate osteoclast acti-
Wnt/C-catenin–regulated gene expression in osteoblasts. vation (Src and MKK6/p38/MITF) and survival (Src and
Like RANKL, OPG also appears to have roles outside extracellular signal–regulated kinase [ERK]).
of bone. For example, it protects large blood vessels of Several adapter molecules bind to RANK along
normal mice from medial calcification and dissection. with TRAFs to mediate signaling. These include Grb-2–
This finding is based on the observation of renal and associated binder (Gab) protein 2, which recruits a num-
aortic calcification and aortic dissection occurring in ber of signaling molecules that contain Src homology-2
OPG -/- mice [35]. Furthermore, when OPG -/- mice were domains after it is tyrosine phosphorylated. Gab2-deficient
crossed with apoE -/- mice, OPG -/-/apoE -/- double-knockout mice have reduced RANKL/RANK-induced osteoclast
mice had acceleration of the calcific atherosclerosis that differentiation, decreased bone resorption, and mild
typically develops in apoE -/- mice. Although this observa- osteopetrosis, indicating that Gab2 plays an important
tion suggests that OPG may protect against calcification of role in RANKL-induced osteoclast formation. Another
atherosclerosis [36], it remains to be determined whether adapter molecule is p62, which binds to TRAF6 to medi-
OPG or RANKL signaling play important roles in car- ate RANKL-induced NF-LB activation in osteoclast
diovascular disease. This is a controversial issue [37]. For precursors. Mice deficient in p62 have normal osteoclast
example, there is an association between high OPG serum formation and bone volumes, but they have an impaired
levels and cardiovascular disease, diabetes, and chronic osteoclastogenic response to PTH-related peptide in vivo,
renal failure in humans [37]. However, patients with suggesting that it has an important role in induced osteo-
renal bone disease typically have increased, rather than clastogenesis in pathologic conditions. More importantly,
decreased, bone resorption because of secondary hyper- mutations in p62 have been linked to some forms of Pag-
parathyroidism, and they also have vascular calcification; et’s disease, which is characterized by increased osteoclast
OPG does not appear to prevent either of these complica- formation and activity [41].
tions in this setting. It is possible that OPG in the serum Several transcription factors have essential functions
of patients with chronic renal failure is rendered inactive in osteoclast formation. All were discovered unexpect-
by being bound to one or more plasma proteins or by edly after the generation of mice with targeted deletion
some other mechanism. Further studies will be required of the genes encoding them developed osteopetrosis.
to determine the significance of these observations, which These include NF-LB p50 and p52, c-Fos, and NFATc1
question the role of the RANKL/OPG ratio in serum [42]. NF-LB p50/p52 -/- mice live until 4 to 6 weeks of age
samples as an indicator of bone mass or bone resorption with severe osteopetrosis and no osteoclast formation.
in some settings [38]. Nfatc1-/- mice die early during embryonic development
from severe cardiac anomalies before the skeleton is
formed [43]. Like NF-LB, NFATc1 also has important
RANKL/RANK Signaling in Osteoclasts roles in immune cell function, and its role as the master
Following the identification of RANKL and RANK, major gene regulating osteoclastogenesis was identified in molec-
efforts were made to identify signaling pathways that ular rescue experiments. Retroviral expression of NFATc1
are activated downstream in osteoclasts and OCPs and induces osteoclast formation from M-CSF–treated NF-LB
to determine more fully the extent of RANKL’s involve- p50/p52 -/- or c-Fos -/- osteoclast precursors in the absence
ment in osteoclast biology and common bone diseases. of RANKL [44], suggesting that NFATc1 is downstream
RANK, like other TNFRs, has no intrinsic protein kinase from NF-LB and c-Fos in RANKL-induced osteoclast
activity to mediate signaling after RANKL binds to it. A differentiation (Fig. 1). However, the precise relationship
key preliminary step in TNFR downstream signaling is among NF-LB, AP-1, and NFAT family members remains
the binding of TRAFs to the cytoplasmic domain of the to be determined. We have demonstrated that overexpres-
receptors. TRAFs 2, 5, and 6 all bind to RANK, but of sion of c-Fos in NF-LB p50/p52 -/- cells induces osteoclast
these only TRAF6 appears to have an essential function in formation, but overexpression of NF-LB p50 and p52 in
osteoclasts, because only TRAF6 knockout mice develop c-Fos -/- cells has no effect, which puts c-Fos downstream
osteopetrosis. Two lines of TRAF6-deficient mice have of NF-LB (Fig. 1). Further study is needed to determine
102 Epidemiology and Pathophysiology
Osteoimmunology, Immunoreceptors,
Unknown and RANKL
RANKL ligands
It has long been recognized that proinflammatory cyto-
OSCAR
kines, which mediate immune responses, can also affect
TREM-2
osteoclasts and osteoblasts, either positively or negatively.
RANK FcRG More recent studies have determined that some of the
22.• Yao Z, Li P, Zhang Q, et al.: Tumor necrosis factor-B 36. Bennett BJ, Scatena M, Kirk EA, et al.: Osteoprotegerin
increases circulating osteoclast precursor numbers by inactivation accelerates advanced atherosclerotic lesion pro-
promoting their proliferation and differentiation in the gression and calcification in older apoE -/- mice. Arterioscler
bone marrow through up-regulation of c-Fms expression. Thromb Vasc Biol 2006, 26:2117–2124.
J Biol Chem 2006, 281:11846–11855. 37. Collin-Osdoby P: Regulation of vascular calcification by
This study demonstrates that TNF increases osteoclast precursor osteoclast regulatory factors RANKL and osteoprotegerin.
proliferation and differentiation via regulation of c-Fms expression. Circ Res 2004, 95:1046–1057.
23. Li P, Schwarz EM, O’Keefe RJ, et al.: Systemic tumor necro- 38. Rogers A, Eastell R: Circulating osteoprotegerin and recep-
sis factor B mediates an increase in peripheral CD11bhigh tor activator for nuclear factor LB ligand: clinical utility
osteoclast precursors in tumor necrosis factor B-transgenic in metabolic bone disease assessment. J Clin Endocrinol
mice. Arthritis Rheum 2004, 50:265–276. Metab 2005, 90:6323–6331.
24.•• Kollet O, Dar A, Shivtiel S, et al.: Osteoclasts degrade 39. Lomaga MA, Yeh WC, Sarosi I, et al.: TRAF6 deficiency
endosteal components and promote mobilization of hema- results in osteopetrosis and defective interleukin-1, CD40,
topoietic progenitor cells. Nat Med 2006, 12:657–664. and LPS signaling. Genes Dev 1999, 13:1015–1024.
First study to show that RANKL not only increases osteoclast 40. Naito A, Azuma S, Tanaka S, et al.: Severe osteopetro-
function but also stimulates the release of hematopoietic stem sis, defective interleukin-1 signalling and lymph node
cells from the bone marrow. Thus, osteoclasts play a key role in organogenesis in TRAF6-deficient mice. Genes Cells 1999,
hematopoietic homeostasis. 4:353–362.
25.•• Kim NS, Kim HJ, Koo BK, et al.: Receptor activator of 41. Kurihara N, Hiruma Y, Zhou H, et al.: Mutation of the
NF-KB ligand regulates the proliferation of mammary sequestosome 1 (p62) gene increases osteoclastogenesis
epithelial cells via Id2. Mol Cell Biol 2006, 26:1002–1013. but does not induce Paget disease. J Clin Invest 2007,
The authors demonstrated that mammary epithelial cells express 117:133–142.
RANK and respond to RANKL treatment. The RANKL/RANK 42. Takayanagi H, Kim S, Koga T, et al.: Induction and activa-
system also regulates mammary gland functions. tion of the transcription factor NFATc1 (NFAT2) integrate
26. Fata JE, Kong Y-Y, Li J, et al.: The osteoclast differentiation RANKL signaling in terminal differentiation of osteoclasts.
factor osteoprotegerin-ligand is essential for mammary Dev Cell 2002, 3:889–901.
gland development. Cell 2000, 103:41–50. 43. Phoon CK, Ji RP, Aristizabal O, et al.: Embryonic heart
27. Wada T, Nakashima T, Hiroshi N, Penninger JM: RANKL- failure in NFATc1-/- mice: novel mechanistic insights from in
RANK signaling in osteoclastogenesis and bone disease. utero ultrasound biomicroscopy. Circ Res 2004, 95:92–99.
Trends Mol Med 2006, 12:17–25. 44. Yao Z, Matsuo K, Nishimura R, et al.: c-Fos/NFAT1- or
28. Chen G, Sircar K, Aprikian A, et al.: Expression of 2-mediated osteoclastogenesis requires NF-LB p50/p52 expres-
RANKL/RANK/OPG in primary and metastatic human sion [abstract]. J Bone Miner Res 2005, 20(suppl 1):S145.
prostate cancer as markers of disease stage and functional 45.•• Asagiri M, Sato K, Usami T, et al.: Autoamplification of
regulation. Cancer 2006, 107:289–298. NFATc1 expression determines its essential role in bone
29. Hughes AE, Ralston SH, Marken J, et al.: Mutations in homeostasis. J Exp Med 2005, 202:1261–1269.
TNFRSF11A, affecting the signal peptide of RANK, cause This study investigated the transcriptional regulation of NFATc1
familial expansile osteolysis [letter]. Nat Genet 2000, and revealed that the transcription factors, NF-LB and c-Fos, bind
24:45–48. to the NFATc1 promoter in a temporal fashion and that NFATc1
30. Kapur RP, Yao Z, Iida MH, et al.: Malignant autosomal autoamplifies its expression.
recessive osteopetrosis caused by spontaneous mutation of 46. Nakashima K, Zhou X, Kunkel G, et al.: The novel zinc
murine Rank. J Bone Miner Res 2004, 19:1689–1697. finger–containing transcription factor osterix is required
31. Whyte MP, Obrecht SE, Finnegan PM, et al.: Osteoprote- for osteoblast differentiation and bone formation. Cell
gerin deficiency and juvenile Paget’s disease. N Engl J Med 2002, 108:17–29.
2002, 347:175–184. 47. Koga T, Matsui Y, Asagiri M, et al.: NFAT and osterix
32. Cundy T, Hegde M, Naot D, et al.: A mutation in the gene cooperatively regulate bone formation. Nat Med 2005,
TNFRSF11B encoding osteoprotegerin causes an idiopathic 11:880–885.
hyperphosphatasia phenotype. Hum Mol Genet 2002, 48. Thiebaud D, Krieg MA, Gillard-Berguer D, et al.: Cyclospo-
11:2119–2127. rine induces high bone turnover and may contribute to bone
33. Hofbauer LC, Schoppet M: Clinical implications of the loss after heart transplantation. Eur J Clin Invest 1996,
osteoprotegerin/RANKL/RANK system for bone and 26:549–555.
vascular diseases. JAMA 2004, 292:490–495. 49. Koga T, Inui M, Inoue K, et al.: Costimulatory signals
34. Boyce BF, Xing L, Chen D: Osteoprotegerin, the bone mediated by the ITAM motif cooperate with RANKL for
protector, is a surprising target for beta-catenin signaling. bone homeostasis. Nature 2004, 428:758–763.
Cell Metab 2005, 2:344–345. 50. Takayanagi H: Mechanistic insight into osteoclast differen-
35. Bucay N, Sarosi I, Dunstan CR, et al.: Osteoprotegerin- tiation in osteoimmunology. J Mol Med 2005, 83:170–179.
deficient mice develop early onset osteoporosis and arterial
calcification. Genes Dev 1998, 12:1260–1268.