Bone Damage
in Rheumatoid
Arthritis: Mechanistic
Insights and
Approaches to
Prevention
Sougata Karmakar, PhD, Jonathan Kay, MD,
Ellen M. Gravallese, MD*
Rheumatoid arthritis (RA) is a systemic disease process that is characterized by
inflammation of the synovial tissues lining the joints. Proliferation of synovial lining cells
and infiltration of inflammatory cells, including monocytes and activated leukocytes,
into the joint tissues results in formation of pannus tissue that covers the surfaces
of articular cartilage and bone1 and produces proinflammatory factors that lead to
destruction of cartilage and bone matrix.
A common and characteristic feature of RA is focal articular bone loss, or erosion,
that becomes evident early in the disease process.2 In RA, inflamed joints are also
often associated with periarticular bone loss, which precedes focal bone erosion.
Erosions occur at sites where pannus invades cortical and subchondral bone and
the adjacent marrow spaces. Ultimately, trabecular bone is also lost. Systemic bone
loss also commonly affects the appendicular and axial skeleton in RA and, with
time, results in increased fracture risk.3,4 When RA is not detected and treated early,
focal bone erosions progress rapidly and result in joint deformity and functional
disability.2,5 Thus, early treatment to prevent focal bone erosions is a critical objective
in caring for patients with RA. In this review, the authors summarize the mechanisms,
cell types, and proinflammatory factors implicated in the pathogenesis of focal
Division of Rheumatology, Department of Medicine, University of Massachusetts Medical
School, UMass Memorial Medical Center, Lazare Research Building, Suite 223, 364 Plantation
Street, Worcester, MA 01605, USA
* Corresponding author.
E-mail address: ellen.gravallese@umassmed.edu
KEYWORDS
Rheumatoid arthritis Therapy Osteoclasts Osteoblasts
Wnt cytokines
Rheum Dis Clin N Am 36 (2010) 385–404
doi:10.1016/j.rdc.2010.03.003 rheumatic.theclinics.com
0889-857X/10/$ – see front matter ª 2010 Elsevier Inc. All rights reserved.
articular bone loss and review recent advances in targeted biologic therapies to
prevent progression of structural damage in RA.
ROLE OF OSTEOCLASTS IN FOCAL ARTICULAR BONE LOSS IN RA
Osteoclasts are specialized multinucleated cells that arise from cells of monocytemacrophage lineage. After attaching to bone matrix proteins, osteoclasts secrete
proteinases and create a local acidic environment that mediates bone destruction.
During physiologic bone remodeling, bone loss mediated by osteoclasts is balanced
by bone formation mediated by osteoblasts. Focal bone loss in RA is mediated by
osteoclasts located at the pannus-bone interface and in subchondral locations.6,7
These cells express typical markers of the osteoclast lineage, including cathepsin K
and tartrate-resistant acid phosphatase (TRAP), as well as the calcitonin receptor,
a marker of terminal osteoclast differentiation.7,8
The receptor activator of NF-kB (RANK) ligand (RANKL) pathway regulates osteoclast
differentiation and function during normal physiologic bone remodeling. RANKL
promotes osteoclastogenesis by binding to its cognate receptor RANK on osteoclast
precursor cells. Osteoprotegerin (OPG) is a soluble decoy receptor for RANKL that blocks
the pro-osteoclastogenic activity of RANKL. In inflammatory arthritis, the RANK/RANKL
pathway is activated, resulting in dysregulated bone remodeling. RA synovial tissues
exhibit an increased ratio of RANKL/OPG mRNA expression, indicating that pro-osteoclastogenic conditions dominate within the microenvironment of the RA joint.9
Support for the critical role of osteoclasts in the process of bone erosion in RA is
provided by studies using mice deficient in RANKL, an essential factor for osteoclast
differentiation. These mice are devoid of osteoclasts and thus have an osteopetrotic
bone phenotype.10 Induction of serum transfer arthritis in RANKL-deficient mice
resulted in full-blown arthritis, but these osteoclast-deficient mice were protected
from focal articular bone loss.11 This finding was confirmed in transgenic mice expressing human tumor necrosis factor (hTNF.tg) that develop a spontaneous, severe, and
destructive polyarthritis resulting in extensive focal articular bone erosion.12 When
crossed with osteopetrotic c-fos–deficient mice that have no osteoclasts, the resulting
mutant mice developed TNF-dependent arthritis without focal articular bone loss,
despite developing inflammation comparable with that of littermate controls expressing c-fos.13 In a T-cell–dependent model of rat adjuvant arthritis characterized by
severe joint inflammation accompanied by bone and cartilage destruction, administration of OPG prevented bone and cartilage destruction but not inflammation, indicating
the ability of OPG to block RANK signaling and focal bone erosion.14,15 Further support
for a critical role of osteoclasts in mediating focal bone erosion in RA has arisen from
murine studies that used bisphosphosphonates to block osteoclast-mediated bone
resorption. hTNF.Tg mice treated with pamidronate or zolendronic acid showed inhibition of focal articular bone loss without any significant change in inflammation.16 This
was also shown with zolendronic acid treatment of mice with collagen-induced
arthritis,17 a model of RA that is driven by T cells and antibodies.
In addition to RANKL, macrophage colony-stimulating factor (M-CSF) is required for
osteoclastogenesis and amplifies the pool of osteoclast cellular precursors.18 RANKL
and M-CSF are synthesized mainly by cells of the osteoblast lineage in the setting of
physiologic bone remodeling.19 In human disease and animal models of RA, activated
CD41 T cells and synovial fibroblasts are alternative sources of RANKL that can drive
osteoclastogenesis.8,14
OPG, the decoy receptor for RANKL that is synthesized by osteoblast-lineage cells,
acts as a key negative regulator of osteoclastogenesis by binding soluble RANKL and
386 Karmakar et al
preventing its binding to its receptor RANK, thereby inhibiting osteoclast differentiation.10,20 In RA, the balance between RANKL and OPG expression levels is a fundamental regulator of osteoclast differentiation and function.21,22 Thus, targeting the
RANKL/RANK/OPG pathway to inhibit osteoclast differentiation represents an important potential therapeutic strategy to prevent bone erosion in inflammatory arthritis.
However, even with aggressive treatment of RA to stop the progression of articular
bone destruction, erosive lesions are not typically repaired; only a limited amount of
new bone forms at sites of erosion. This suggests that osteoblast-mediated bone
formation may also be compromised at sites where inflammation and erosion occur,
thereby contributing to the net loss of bone.
ROLE OF OSTEOBLASTS IN FOCAL ARTICULAR BONE LOSS IN RA
The role of osteoblasts in focal articular bone loss in RA has received less attention
than that of osteoclasts. Osteoblasts are important regulators of bone remodeling.
These cells produce and mineralize bone matrix and modulate osteoclast differentiation and function by producing RANKL and OPG.23 Osteoblasts arise from mesenchymal stem cells and undergo maturation and differentiation toward cells with the
capacity to produce and mineralize bone matrix.24 As osteoblasts mature toward
functional bone-forming cells, they diminish their expression of RANKL and increase
their expression of OPG, thereby creating a microenvironment that favors bone formation over bone loss.25
Although patients with RA receiving disease-modifying antirheumatic drug (DMARD)
therapy demonstrate slowing or inhibition of progression of articular bone erosions on
plain radiographs in clinical trials and long-term observational studies,26 healing of
erosions has been considered to occur only infrequently. However, in recent years,
repair of bone erosions with formation of new bone has been documented in patients
with RA treated with DMARDs.27–30 Based on results of 2 studies conducted by the
OMERACT Subcommittee on Healing of Erosions, Sharp and colleagues31 confirmed
that repair of erosions does occur in RA. Additional studies reported a strong association
between clinical remission of disease and radiographic evidence of repair of
erosions,32,33 suggesting that joint inflammation may play an important role in suppressing the functional capacity of osteoblasts to produce bone and thereby repair erosions.
These observations are supported by a recent study from our laboratory in a murine
model of arthritis that showed that mineralized bone formation was compromised at
sites of erosion in areas of bone where there was local inflammation.34 It would be
expected that bone formation rates would be increased at sites of erosion where
bone has been lost. However, using dynamic bone histomorphometry, no differences
were observed in the rate of mineralized bone formation at bone surfaces in inflamed
areas, compared with similar areas without arthritis. In addition, there was a notable
paucity of osteoblast-lineage cells expressing markers of cell maturity at bone
surfaces adjacent to inflammation, but there were abundant immature osteoblastlineage cells expressing Runx2, a marker of early osteoblast-lineage cells. Furthermore, cellular expression of alkaline phosphatase, a marker of the mineralization
phase of bone formation, was minimal at sites of bone erosion. These observations
support the hypothesis that inflammation may inhibit the functional capacity of
osteoblasts to repair erosions in RA.
WNT SIGNALING PATHWAY IN RA
To elucidate the effect of inflammation on osteoblasts and bone formation, investigators have focused on the Wingless (Wnt) family of proteins. The Wnt signaling pathway
Bone Damage in Rheumatoid Arthritis 387
plays an important regulatory role in embryonic development, tissue homeostasis and
cancer.35 Wnt proteins are secreted glycoproteins that bind and activate the receptor
complex of a 7-transmembrane domain-spanning frizzled receptor and the lowdensity lipoprotein receptor-related protein 5 and 6 (LRP5/6) coreceptors present
on mesenchymal cells. Binding of Wnt proteins to the receptor complex induces the
differentiation of osteoblast-lineage cells into mature osteoblasts, leading to bone
formation.36 The importance of the Wnt signaling pathway in bone homeostasis is
shown by the abnormal bone phenotypes that result from receptor mutations. Mutations inducing loss of function in the Lrp5 gene, which encodes the LRP5 receptor,
decrease bone formation in humans and mice.37 In contrast, gain-of-function mutations in the Lrp5 gene result in increased bone mass.38
b-Cateninplaysa criticalrole inboneremodeling.Modulationof Wntsignalingalters bcatenin expression and activity, thereby affecting bone formation and osteoclast-mediated bone resorption. Mice that do not express b-catenin in differentiated osteoblasts
develop severe osteopenia associated with increased numbers of activated osteoclasts
resulting from an increased RANKL/OPG ratio.39,40 In contrast, mice expressing constitutively active b-catenin–mediated Wnt signaling in differentiated osteoblasts have
increased OPG expression and exhibit a decrease in bone resorption and an osteopetrotic phenotype.39,40 The functional dysregulation of osteoblasts seen in mice deficient
in b-catenin signaling resembles the osteoblast phenotype found at sites of focal bone
loss in inflammatory arthritis, suggesting that inhibition of Wnt signaling may be a mechanism whereby bone formation is compromised in inflammatory states.
Several families of endogenous inhibitors of the Wnt signaling pathway have been
identified; these function by limiting the effects of Wnt signaling. The Dickkopf (DKK)
and secreted frizzled-related protein (sFRP) family members are the best characterized. Diarra and colleagues41 showed that DKK1, a member of the DKK family of
Wnt antagonists, plays a central role in regulating bone remodeling in RA. Administration of a neutralizing antibody specific to DKK1 protected against focal bone erosion in
the hTNF.Tg RA mouse model. In the setting of DKK1 blockade, new bone formation
was observed at sites where erosion would typically occur. However, these effects