Cell-cell Interactions
in Rheumatoid
Arthritis Synovium
David A. Fox, MDa,*, Alison Gizinski, MDa, Rachel Morgan, BScb,
Steven K. Lundy, PhDb
Although the cause of rheumatoid arthritis (RA) remains unknown, insights into the
pathogenesis of RA have been achieved by careful study of inflammatory, immune,
and tissue-destructive processes that take place in synovial tissue. Successful
approaches have included ex vivo analysis of RA synovium, experiments in animal
model systems, and use of cultured cell lines, especially fibroblast-like synoviocytes,
derived from patients’ synovial tissue. These insights have led to remarkable advances
in the treatment of RA and other diseases, such as the use of tumor necrosis factor
(TNF)-blocking biologics.
None of the molecular targets of medications currently used in the treatment of RA
are expressed uniquely in synovial tissue. Instead, all of them are of great importance
in host defense as well. Recent research is revealing important molecules and pathways pertinent to joint inflammation and damage that may be less central to host
defense compared with currently targeted molecules. A conceptual framework for
such investigations is the realization that although RA synovium can display some
features of an immune organ and is justifiably regarded as a tertiary lymphoid structure, it also contains cells that are distinct from those found in primary or secondary
lymphoid tissue, namely, the intrinsic structural cells of the joint such as fibroblastlike synoviocytes.
Grant support: David Fox: NIH RO-1 AR38477; Alison Gizinski: American College of
Rheumatology Research and Education Foundation; Rachel Morgan: Immunology Training
Grant and Rackham Pre-doctoral Merit Fellowship: Steven Lundy: Arthritis Foundation Arthritis
Investigator Award and NIH K Award.
a Division of Rheumatology and Rheumatic Diseases Research Core Center, 3918 Taubman
Center, 1500 East Medical Center Drive, The University of Michigan, Ann Arbor, MI 48109, USA
b Division of Rheumatology and Rheumatic Diseases Research Core Center, 109 Zina Pitcher
Place, The University of Michigan, Ann Arbor, MI 48109, USA
* Corresponding author.
E-mail address: dfox@umich.edu
KEYWORDS
Synovial fibroblasts T lymphocytes B lymphocytes
Antigen-presenting cells Endothelial cells Cytokines
Rheum Dis Clin N Am 36 (2010) 311–323
doi:10.1016/j.rdc.2010.02.004 rheumatic.theclinics.com
0889-857X/10/$ – see front matter ª 2010 Elsevier Inc. All rights reserved.
It is clear that no one cell type explains the pathologic behavior of RA synovial
tissue. It is the interactions between these cells that define the disease. The 3 most
abundant cell populations in RA synovium are the monocyte/macrophage synoviocytes (type A), the fibroblast-like synoviocytes (FLS) (type B), and T lymphocytes
(which are strikingly heterogeneous). Other critically important cells of the RA synovium include B lymphocytes, plasma cells, dendritic cells, mast cells, endothelial cells,
osteoclasts, and adjacent chondrocytes. These various cell types can interact in 2
general ways: first through secreted mediators, notably inflammatory cytokines
such as TNF, interleukin (IL)-6, IL-17, and many others; and second through direct
cell-cell contact that is mediated by cell surface receptors and ligands, including
some membrane-anchored cytokines.
This article focuses on selected cell-cell interactions that may be important in the
pathogenesis of RA. It is hoped that the relative molecular specificity of some of these
interactions for events in the joint compared with the systemic immune response will
provide more specific targets (Table 1) for a new generation of biologic and nonbiologic therapeutics.
HOMOTYPIC AND AUTOCRINE INTERACTIONS OF FLS
Cadherin-11
During the course of RA, the cells of the synovial lining undergo extensive hyperplasia
to form the synovial pannus that invades and destroys cartilage and bone. Recent
discoveries have highlighted a critical role for cadherin-11 in these events. Cadherin-11 has been identified as a strongly expressed intercellular adhesion molecule
on human and mouse FLS.1,2 Transfection of L cells with cadherin-11 led to the formation of sheet-like structures with an organization similar to that seen in synovial lining1;
furthermore, cadherin-11 localized to cell-to-cell junctions between FLS.2 Joints in
cadherin-11 null mice had an underdeveloped synovial lining and decreased extracellular matrix.2 These data indicate that cadherin-11 plays a vital role in the formation of
the synovial lining layer by mediating FLS/FLS connections.
Cadherin-11 staining correlated strongly with cellular infiltration of macrophages
and T lymphocytes in RA synovium.3 Cadherin-11 staining also correlated with erythrocyte sedimentation rate and C-reactive protein level although not as strongly.3 Cadherin-11 expression in synovium is not specific to RA, because cadherin-11 staining
on synovial biopsies was similar in inflamed joints from RA, osteoarthritis (OA), and
psoriatic arthritis (PsA).3 There was also a positive correlation between cadherin-11
staining of lung tissue from patients with RA-associated interstitial pneumonitis (IP)
and CD41 T-cell infiltration of the lung.3 Cadherin-11 null mice showed an average
of 50% reduction in clinical arthritis activity in the K/BxN serum transfer model.2
Table 1
Important cell-cell interactions in RA synovium
Cell Types Potential New Molecular Targets
FLS-FLS Cadherin-11, fractalkine
FLS-T B7-H3, IL-15
T-APC OX40, CCL20, IL-7
B-T CXCL13, ICOS, OX40, BAFF
B-FLS BAFF, osteopontin
Leucocyte-endothelial Ley/H
Please see text for explanation of abbreviations.
312 Fox et al
Cadherin-11 has been explored as a possible therapeutic target using the same
mouse serum transfer model. Cadherin-11-Fc and an anticadherin-11 mAb ameliorated clinical arthritis when administered with arthritogenic serum.2 More significant
for potential treatment of human disease, anticadherin-11 mAb ameliorated established arthritis in a modified K/BxN serum transfer model.2
Fractalkine and its Receptor
Fractalkine (FKN) is a potent chemoattractant and adhesion molecule that is found in
increased levels in RA synovium. RA FLS secrete FKN and express its receptor,
CX3CR1.4 Soluble FKN induced proliferation of FLS that was blocked by addition of
anti-CX3CR1.5 Even in the absence of sFKN, the antibody was able to decrease
FLS proliferation, revealing an autocrine growth loop.5 FKN also induced migration
of RA FLS and caused significant reorganization of F-actin within FLS.4 It is likely
that FKN could act in an autocrine fashion to aid in pannus invasion of the bone
and cartilage through FLS growth and migration.
FLS/T CELL INTERACTIONS
In RA, FLS and other cells produce chemokines that attract T cells to the joint. FLS and
T cells then interact in the synovium through secreted factors and direct cell-to-cell
interactions, resulting in activation of both cell types. FLS proliferate when cocultured
with CD41 T cells, especially when RA T cells are used.5
Synovial T cells in RA patients include an expanded population of CD41CD28 cells
and this subset of cells in particular greatly enhances FLS proliferation. CD41CD28
cells, which are sometimes considered to be senescent, aberrantly express CX3CR1,
and anti-CX3CR1 decreases the FLS growth-promoting activity of these cells without
a significant effect on CD41CD281 T cells.5 Stimulation of CD41CD28 T cells by FKN
through CX3CR1 increases TNFa production, and TNFa can then act on FLS to
increase growth, FKN secretion, and CX3CR1 expression.4,5 These data suggest an
important relationship between TNFa and FKN/CX3CR1 in FLS/T cell interactions in
the RA joint, in which production and action of these molecules occur in linked paracrine and autocrine loops involving T cells and FLS.
IL-15
IL-15 is constitutively expressed on FLS and is a potent T-cell growth factor that can
cause activation/proliferation of effector T cells (Teff) and regulatory T cells (Treg) in Tcell/FLS cocultures.6 Moreover, IL-15 can decrease apoptosis of various cell types
including FLS and T cells.7,8 The IL-15 receptor is a trimer (IL-15Ra,b,gc) and subunits
of IL-15R are expressed by various cell types, including FLS and T cells.7 IL-15 can
function as a secreted or membrane-bound cytokine, with signaling similar to other
cytokines through the full trimeric IL-15R, or through dimeric IL-15Rbgc receptor (cis
presentation).7 IL-15 can also signal through a unique trans signaling form, in which
distinct subunits of IL-15R are expressed on the surface of interacting cells. IL-15
can be recycled by the cell and presented on the cell surface by IL-15Ra, a pathway
that allows for persistence of an IL-15 signal even when soluble IL-15 is no longer
available.7 In cultures containing Teff and Treg cocultured with RA FLS, proliferation
and function of both T cell subsets were stimulated, with a net proinflammatory effect.
These effects were not observed when OA FLS or dermal fibroblasts were used and
were dependent on cell-to-cell contact. Neutralizing IL-15 during T-cell/FLS cocultures significantly attenuated the proliferation of Teff and Treg, Teff production of
TNFa and IFNg, and Treg inhibition of Teff.6 The overall proinflammatory effect of
Cell-cell Interactions in Rheumatoid Arthritis Synovium 313
IL-15 makes it a potential target for RA therapy. A proof-of-concept study has been
conducted using a human immunoglobulin (Ig)G1 anti-IL-15 monoclonal antibody in
RA. This antibody suppressed effects of IL-15 in vitro, and showed meaningful efficacy
in phase I and II trials.8
FLS as Antigen-presenting Cells
FLS can also act as antigen-presenting cells (APCs) in the initiation of T-lymphocyte
responses. For example, FLS induced secretion of IL-2 by class II MHC-restricted
CD41 T-cell hybridomas specific for arthritogenic autoantigens, specifically human
cartilage gp-39 (HC gp-39) and human type II collagen (CII).9 The T-cell hybridomas
in these experiments were developed from HLA-DR4 transgenic mice, and therefore
respond to peptide antigens that are loaded onto and presented by HLA-DR4. Activation of the hybridomas required reexpression of class II on the FLS, which occurs in
vivo in RA and is reinduced in vitro by IFNg. The T-cell response to peptide antigen
presented by FLS was exquisitely MHC-restricted, identical to experiments in which
professional APCs were used,9 and blocking antibodies to human class II MHC or
murine CD4 prevented IL-2 production. Because T-cell hybridomas do not require
a second signal to respond to peptide antigen, this system was not useful for defining
potential costimulatory ligands on FLS. FLS do not express the classic APC costimulatory molecules CD80 or CD86 (B7-1, B7-2) at functionally relevant levels.10
B7-H3
Of the molecules that belong to the B7 family, B7-H3 is expressed strongly and constitutively on FLS in vitro.10 Moreover, immunostaining of RA synovium showed broad
B7-H3 expression nearly identical in distribution to the FLS marker cadherin-11.
Dual-color immunohistochemical analysis showed CD31 T cells in close proximity
to FLS expressing B7-H3.10 Furthermore, in coculture experiments, B7-H3 localized
to the contact point between FLS and cytokine-activated T cells (Tck) or during FLS
presentation of superantigen to T cells.10 Previous studies had indicated that B7-H3
can have either stimulatory or inhibitory effects on T cells, and results with FLS/Tcell cocultures were consistent with a dual role for B7-H3. RNAi knockdown of B7-
H3 in FLS decreased production of TNFa, IFNg, and IL-2 by cocultured Tck but
increased production of these cytokines by resting T cells.10 The T-cell ligand or
ligands for B7-H3 have not yet been defined. B7-H3, in contrast to B7-1 and B7-2,
is expressed on human solid tumors,11 in part controlled by the microRNA miR-
29.12 In early-phase human clinical trials, the B7-H3–specific mAb 8H9 was reported
to prolong survival in patients with solid tumors and central nervous system metastasis.12 Further trials of anti-B7-H3 in cancer could yield safety information pertinent
to consideration of therapeutic trials of this antibody in RA.
T CELLS AND APCS IN THE RA SYNOVIUM
The following sections highlight examples of selected cytokine-mediated and cognate
cell-cell contact-driven interactions between T cells and synovial APCs that are
current or potential therapeutic targets for the treatment of RA. Interruption of T-cell
costimulation between T cells and APCs in the synovium is a worthwhile approach
given the clinical success of blocking T-cell costimulation with CTLA-4Ig in RA.
Resting T cells require at least 2 signals to differentiate into effector T cells: the first
signal through engagement of the T-cell antigen receptor by the antigen-MHC
complex on the APC, and the second signal by engagement of costimulatory molecules such as CD28, on T cells by ligands such as CD80/86 on APCs. Effector
314 Fox et al
T-cell differentiation leads to the expression of additional surface molecules. These
inducible structures may have stimulatory (ICOS, OX40) or inhibitory (CTLA-4)
potential.
CTLA-4 and Indoleamine Dioxygenase
CD28 is the prototypic T-cell receptor (TCR) for costimulatory signals. The ligation of
CD28 by CD80/86 (B7-1 and B7-2) sends activating signals into the T cell and the
APC. CTLA-4 (CD152) is up-regulated on activated T cells and binds to the CD28
ligands CD80/86. The ligation of CD80/86 by CTLA-4 sends inhibitory signals directly
into the T cell. The ligation of CD80/86 by CTLA-4 can also deliver regulatory signals to
the APC. The interaction of CD80/86 with CTLA-4 leads to the induction of indoleamine dioxygenase (IDO) in APCs.13 IDO is believed to be critical in inducing anergy
in T cells, because IDO depletes tryptophan that is necessary for T-cell activation.
Blocking the activity of IDO in a mouse model of arthritis led to increased severity of
arthritis and accumulation of Th1 and Th17 cells in the inflamed joints.14,15 Conversely,
administration of L-kynurenine, a metabolite of L-tryptophan, resulted in amelioration
of arthritis. These findings suggest manipulation of tryptophan degradation as a therapeutic target in RA.15
OX40/OX40L
OX40 (CD134) is predominantly expressed on activated CD4 and CD8 T cells following
stimulation via TCR and CD28. Proinflammatory cytokines, IL-1, IL-2, and TNFa can
further augment the expression of OX40. The ligand of OX40 and OX40L (CD252) is
expressed on APCs including dendritic cells, B cells, and macrophages.16 OX40L is
induced on APCs after stimulation via CD40.17 There is bidirectional activation of T
cells and APCs via the OX40/OX40L pathway. Stimulation of OX40 on T cells induces
proliferation and cytokine secretion, and signaling via OX40L into APCs induces the
secretion of proinflammatory cytokines by APCs.17
In RA, there is increased expression of OX40 on peripheral blood CD4 T cells with
a trend toward a positive correlation with serum C-reactive protein levels.18 OX40 is
expressed on CD4- and CD8-positive synovial T cells, more so on CD4-positive T
cells, and levels of expression correlate with disease activity.19–21 OX40 and OX40L
expressing cells are also present in the RA synovium.19 Although there is increased
expression of OX40 and OX40L in the RA synovium, the role of OX40/OX40L in mediating immune events in RA is unclear. However, preclinical studies in animal models
blocking the interaction of OX40/OX40L have shown promise.19
CCL20
Chemokines mediate inflammatory responses by stimulating the recruitment of leucocytes. The chemokine CCL20 (MIP-3a) and its receptor CCR6 play a role in the migration of different cell types to the RA synovium. CCL20 is a CC-chemokine expressed
on macrophages, dendritic cells, and lymphocytes. CCR6, the receptor for CCL20, is
expressed on Th17 cells, immature dendritic cells, and B cells. CCL20 is chemotactic
for Th17 cells and dendritic cells, which express CCR6. Th17 cells secrete CCL20 and
recruit other CCR6-expressing Th17 cells to the site of Th17 cell-mediated damage.22
Increased levels of CCL20 have been observed in the synovial fluid of RA patients
compared with OA patients, and protein concentrations of CCL20 are increased in
the peripheral blood of patients with RA.23,24 Expression of CCL20 is induced in
FLS by the synergistic interaction of proinflammatory cytokines, including TNFa, IL-
1b and IL-17, and cytokine-stimulated FLS can recruit mononuclear cells in
a CCL20/CC6–dependent manner.25,26 The results suggest that CCL20 produced
Cell-cell Interactions in Rheumatoid Arthritis Synovium 315
by FLS recruits monocytes and Th17 T cells to the synovium and is an important chemokine in the pathogenesis of RA. Moreover, blockade of CCL20 binding to CCR6
with a neutralizing antibody was effective in treating arthritis in a T-cell transfer model
in mice.22 Further supporting the role of CCL20 as an important chemokine in RA,
a recent study demonstrated that treatment with infliximab, etanercept, or tocilizumab
reduced serum levels of CCL20 in patients with RA.27
IL-7
IL-7 is a member of the IL-2 family. IL-7 is associated with endothelial cells, FLS, and
macrophages in the RA synovium and colocalizes with deposits of extracellular matrix
collagen IV.28 IL-7R is composed of IL-7Ra and IL-2Rg and is expressed on CD41 and
CD81 T cells, NK T cells, and monocytes. IL-7 levels are increased in RA compared
with OA synovial fluid. Serum IL-7 levels correlate with disease activity in RA.29 In
RA patients who are poor responders to anti-TNF therapy, persistently increased
serum levels of IL-7 are seen.29 Moreover, reduced levels of serum IL-7 were observed
in patients with early RA treated with methotrexate, and the reduction correlated with
disease suppression.29
Within RA synovial biopsies, samples with lymphoid follicles demonstrated consistent
IL-7 staining, and gene expression analysis of RA synovial samples revealed increased
expression of genes involved in IL-7 signal transduction.28 IL-7 may be responsible for
generation of tertiary lymphoid follicles observed in RA synovium, because IL7 is known
to be crucial for the development of lymphoid tissue. Enhanced expression of IL-7Ra and
IL-7 in patients with RA may contribute to joint inflammation by activating T cells, B cells,
and macrophages because treatment with soluble human IL-7Ra inhibited IL-7R–mediated immune activation in vitro.30 These research findings suggest that IL-7 and its
receptor are potential targets for immune modulation in RA therapy.
B-CELL INTERACTIONS WITH SYNOVIAL T CELLS
Recent successes of therapeutic interventions using B-cell–depleting reagents have
highlighted the importance of B cells in the pathogenesis of RA.31 These current therapies, which target CD20-positive B cells, and several others that are in the pipeline
that target other B-cell markers, are designed to eliminate or disrupt the activation
of a large percentage of the entire population of B cells. Although effective as treatment for RA in the short-term, each of these reagents has the potential to have
long-term deleterious effects on the immune response to common infectious microorganisms. Therefore, it is desirable to continue pursuing B-cell–directed treatments
that are more specific to the pathogenic subset of B cells that mediate joint inflammation. The following is a summary of some recently recognized cell-cell interactions
involving synovial B cells that point to emerging potential targets of interest.
Germinal Centers in RA Synovium
It has been established that structures resembling lymph node germinal centers can
be found in the inflamed synovium of a subset of RA patients. The recently described
expression of activation-induced cytidine deaminase (AID) within these ectopic
germinal centers supports the hypothesis that maturation of the antibody response
through somatic hypermutation and class-switch recombination could occur within
the RA synovium.32 This study went on to show that germinal centers remained functionally active following implantation of RA synovial tissue in SCID mice.32 Analysis of
a panel of cytokines and chemokines that are believed to contribute to germinal center
formation showed correlations between the presence of germinal centers and TNFa,
316 Fox et al
lymphotoxin-b, APRIL, and B-lymphocyte chemoattractant (BLC, CXCL13). Another
study by the same researchers demonstrated that mononuclear cells from RA synovium had 400-fold higher expression of CXCL13 than cells isolated from the peripheral
blood of the same patients.33 Most synovial CXCL13 was produced by CD45RO1/
CD41/CD31 T lymphocytes, suggesting that this T-cell subset was of an activated
phenotype. Further analysis of surface markers on these T cells suggested a phenotype that was distinct from T follicular helper cells found in other lymphoid structures.
Although systemic blockade of CXCL13 or its receptor CXCR5 may be expected to
interrupt T cell–B cell interactions that are important to host defenses, the findings
concerning CXCL13 expression within the RA synovium open the possibility of
intra-articular immune therapy directed at CXCL13/CXCR5 or other contributors to
germinal center formation. Localized neutralization of other cytokines or cytokine
receptors involved in germinal center formation including lymphotoxin, IL-21, APRIL,
and B-cell–activating factor (BAFF) may also provide therapeutic benefit in some
patients.
ICOS and Its Ligand
As described earlier, B cells constitute an important APC population for T-lymphocyte
activation. Therapies directed at OX40/OX40L, CD80, and CD86 are expected to act
on B-cell antigen presentation as well as dendritic cells and macrophages. Another
notable interaction between CD41 T helper cells and B cells that has gained interest
recently is that of T-cell–expressed inducible costimulator (ICOS, CD278) with its
ligand on B cells, B7RP-1 (ICOS-L, CD275). These molecules are also involved in
the formation of germinal centers and are critical to the initiation and further development of antigen-specific antibody responses. Disruption of interactions between ICOS
and its ligand has been shown to diminish arthritis incidence and severity in 2 independent studies.34,35 In the latter study, treatment with a blocking antibody decreased the
number of T follicular helper cells and germinal center B cells in the draining lymph
nodes and spleens of mice in the collagen-induced arthritis model. This led to an overall reduction in T-cell cytokine production and titers of anticollagen antibodies in the
serum. ICOS/ICOS-L blockade was also effective in the NZB/NZW F1 model of lupus
nephritis. As described earlier for CXCL13, systemic blockade of ICOS/ICOS-L interactions may not prove to be desirable because of adverse side effects, but local inhibition of ICOS/ICOS-L in the synovium may be a viable treatment option.
Regulatory B Cells and NK T Cells
Not all interactions between B cells and T cells result in increased disease pathogenesis in arthritis. Regulatory B cells that express IL-10 have been studied for many
years, but have recently gained prominence through the demonstration that they
have unique surface markers (CD1dhiCD51) and may represent a subset of B cells
distinct from those that are pathogenic in autoimmune diseases.36 The high expression of CD1d by these regulatory B cells suggests that they may present antigens
to invariant NK T cells. The role of NK T cells in arthritis has been controversial with
almost an equal number of articles suggesting a pathogenic or suppressive phenotype.37 The conflicting roles of NK T cells in arthritis might be explained by differences
in the B-cell populations that are presenting antigen to them. A recent study has
demonstrated the increased presence of invariant NK T cells with regulatory phenotype following depletion of B cells with anti-CD20 therapy.38 This study suggests
a novel mechanism contributing to the success of B-cell depletion therapy, and
demonstrates the potential efficacy of targeting the interaction of B cells with iNK T
cells.
Cell-cell Interactions in Rheumatoid Arthritis Synovium 317
B-CELL INTERACTIONS WITH OTHER SYNOVIAL CELL POPULATIONS
Although B-cell interactions with T cells in the RA synovium may present important
and obvious targets for future therapy, it is critical to understand that signaling
between B cells and other synovial cell populations may also contribute to joint
pathology.
BAFF
A recent study has shown that BAFF is expressed on the cell surface of RA FLS
but not on OA FLS.39 This RA FLS-associated BAFF acted as a first signal for the
transcription of recombinase-activating genes (RAG) in B cells that are involved in
maturation of the antibody response. RAG expression in B cells cocultured with
RA FLS was blocked by interference with BAFF gene translation or by neutralization of IL-6. Forced expression of membrane-associated BAFF in OA FLS did not
result in RAG gene transcription unless IL-6 was also present in the coculture.
These results highlight that RA FLS can have a direct effect on B-cell activation
through the combined expression of cell surface BAFF and IL-6. Evidence for
the importance of local synovial BAFF expression to the pathogenesis of arthritis
comes from a study in which intra-articular injection of a lentiviral vector containing a BAFF-silencing reagent provided long-term protection from joint inflammation
in the collagen-induced arthritis model.40 This protective effect of BAFF gene
silencing within the joint was accompanied by decreases in production of anticollagen antibodies and IL-17 by cells of the joint and draining lymph node, but not in
the spleen of treated mice. This study demonstrates another potential approach to
localized treatment of arthritic inflammation.
Osteopontin
Another notable interaction between RA FLS and B cells that may contribute to joint
inflammation was found to be associated with the expression of an RA-associated isoform of osteopontin (OPN).41 This larger form of OPN was preferentially detected in RA
FLS on Western blot analysis, and was found to form a unique macrocomplex with
fibrinogen on the cell surface of RA FLS. Cocultures of purified B cells with OPN-fibrinogen complex–positive FLS led to increased expression of IL-6, an effect that was
blocked by specific inhibition of OPN production. Immunohistochemistry of RA synovial tissue demonstrated the colocalization of OPN, IL-6, and B cells to highly inflamed
areas of the joint.
RANK-L on B Cells
Provocative new data suggest that B cells can express the receptor activator of
nuclear factor-kB ligand (RANK-L).42 The expression of RANK-L in the synovium
drives differentiation of osteoclasts that mediate the erosion of bone in RA. The study
by Han and colleagues42 demonstrated that RANK-L expression was stimulated on rat
splenic B cells by culture with the inactivated bacterium, Aggregatibacter actinomycetemcomitans, a pathogenic microbe associated with human periodontal disease. Purified splenic B cells from rats infected with A actinomycetemcomitans stimulated
a reporter cell line to differentiate toward osteoclasts.42 Thus, oral infection has
systemic effects on RANK-L expression by B cells and may be a partial explanation
of the association between RA and periodontal disease. Although the expression of
RANK-L by synovial B cells has not yet been reported, it will be interesting to determine whether B cells contribute to RA tissue damage through this mechanism, and
what local or systemic signals might be involved in B cell expression of RANK-L.
318 Fox et al
LEUCOCYTE-ENDOTHELIAL INTERACTIONS IN RA SYNOVIUM
Cell migration into inflammatory lesions is a tightly regulated and complex process
that involves adhesion receptors on leucocytes and vascular endothelial cells, multiple
chemokines, and retention signals within target tissues. Space does not permit
a comprehensive description of the extensive knowledge concerning these processes
in RA synovium; they have been recently reviewed elsewhere.43 Relevant adhesion
molecules include selectins, integrins, and members of the immunoglobulin gene
superfamily.43 These molecules are also used in various combinations in leucocyte
efflux into nonsynovial tissues, but it seems likely that in some respects, the synovial
vasculature and the cell homing mechanisms to synovium are unique.
Neutrophil Recruitment
Intravital microscopy has emerged as an elegant technique for assessing leucocyte
ingress into synovium, at least in animal model systems.44 Such studies have revealed
an unexpected prominence of neutrophil versus lymphocyte recruitment that, in RA,
could be reflected by the cell composition of synovial fluid more than synovial tissue.
In the mouse model of proteoglycan-induced arthritis, neutrophil expression of CD44
and CD62L (L-selectin) is crucial for entry into the joint.45 In human RA, in vitro studies
suggest that FLS control neutrophil recruitment indirectly through their influence on
endothelial cells.46 Coculture of endothelial cells with RA FLS but not RA dermal fibroblasts markedly augmented adherence of neutrophils to the endothelial cells,
a process that depended on IL-6 production by the FLS. A role for FLS IL-6 was
also identified in a similar coculture system that measured lymphocyte recruitment
by endothelial cells.47 In these experiments, OA FLS and RA FLS behaved similarly,
but FLS from injured but otherwise normal joints or dermal fibroblasts inhibited
lymphocyte-endothelial adhesion. Although limited by the use of umbilical vein rather
than synovial microvascular endothelial cells, these interesting studies provide striking
evidence for a role of FLS in the control of endothelial cell function.
Permeability of the Synovial Vasculature
Ingress of leucocytes into synovium may also be facilitated by unique permeability
properties of the synovial microvasculature that have been revealed by experiments
in a murine immune-complex model of acute joint inflammation that is induced by
serum transfer (as described earlier). Intravital microscopy was used to show that
joints destined to become inflamed developed a unique degree of vascular leakage
on systemic administration of arthritogenic sera or even nonarthritogenic immune
complexes, through a process dependent on mast cells and vasoactive amines.48
Such studies may point to hitherto unappreciated physiologic functions of normal
synovium, and explain the high propensity for systemic inflammatory processes to
target the joints. The relationship of this model system to chronic arthritis, particularly
RA, is still difficult to define.
Blood Vessel Growth in RA Synovium
Vasculogenesis and angiogenesis are distinct but related events that occur in RA
synovium.43 Vasculogenesis, the de novo formation of blood vessels from circulating endothelial cell precursors (EPCs), is believed to occur in RA synovium
and in corresponding animal models.43,49 Coculture studies have implicated cellcell contact between RA EPCs and RA FLS, mediated by binding of vascular
cell adhesion molecule 1 (VCAM-1) to the integrin VLA-4, as critical to such interactions, which could retain EPCs in synovial tissue long enough to allow
Cell-cell Interactions in Rheumatoid Arthritis Synovium 319
vasculogenesis to occur.49 New blood vessel formation in RA synovium also
depends on TNF and can be reversed by TNF blockade.50 TNF and other proinflammatory cytokines can induce molecules that are involved in angiogenesis, leucocyte-endothelial adhesion, or both. One example is the group of molecules
termed Ley/H (identified by a carbohydrate-binding antibody termed 4A11) or its
analog H-2 g.51 These molecules not only promote angiogenesis and expression
of adhesion ligands by endothelial cells, but also stimulate monocyte migration.
This latter effect was observed in vitro and in vivo in assays of monocyte recruitment to human RA synovial tissues implanted into SCID mice.51
Lymphocyte Egress
Little is known about the control of lymphocyte egress versus retention in RA synovium. It is possible that some current treatment approaches, such as TNF blockade,
do stimulate cell egress, although this is difficult to prove. Some data exist to suggest
that expression of chemokine receptors on T lymphocytes is altered on entry into RA
synovium, and that chemokine gradients in synovial lymphatics, endothelium, and
stromal cells are distorted to discourage lymphocyte egress.52 Readjustment of these
gradients to deplete synovial tissue of inflammatory cells is a particularly compelling
approach to resolving joint inflammation.
SUMMARY
Analysis of cell-cell interactions in RA synovium is providing a vivid and detailed understanding of RA pathogenesis. Current biologic and nonbiologic therapies are likely to
be already targeting some of these interactions in ways that are still poorly understood. Several new possibilities for more targeted therapies have been identified
based on interactions that are unique, or uniquely important, in the inflamed joint.
REFERENCES
1. Valencia X, Higgins JM, Kiener HP, et al. Cadherin-11 provides specific cellular
adhesion between fibroblast-like synoviocytes. J Exp Med 2004;200:1673.
2. Lee DM, Kiener HP, Agarwal SK, et al. Cadherin-11 in synovial lining formation
and pathology in arthritis. Science 2007;315:1006.
3. Vandooren B, Cantaert T, ter Borg M, et al. Tumor necrosis factor alpha drives
cadherin 11 expression in rheumatoid inflammation. Arthritis Rheum 2008;58:
3051.
4. Volin MV, Huynh N, Klosowska K, et al. Fractalkine is a novel chemoattractant for
rheumatoid arthritis fibroblast-like synoviocyte signaling through MAP kinases
and Akt. Arthritis Rheum 2007;56:2512.
5. Sawai H, Park YW, He X, et al. Fractalkine mediates T cell-dependent proliferation
of synovial fibroblasts in rheumatoid arthritis. Arthritis Rheum 2007;56:3215.
6. Benito-Miguel M, Garcia-Carmona Y, Balsa A, et al. A dual action of rheumatoid
arthritis synovial fibroblast IL-15 expression on the equilibrium between
CD41CD251 regulatory T cells and CD41CD25 responder T cells. J Immunol
2009;183:8268.
7. Dubois S, Mariner J, Waldmann TA, et al. IL-15Ralpha recycles and presents
IL-15 in trans to neighboring cells. Immunity 2002;17:537.
8. Baslund B, Tvede N, Danneskiold-Samsoe B, et al. Targeting interleukin-15 in
patients with rheumatoid arthritis: a proof-of-concept study. Arthritis Rheum
2005;52:2686.
320 Fox et al
9. Tran CN, Davis MJ, Tesmer LA, et al. Presentation of arthritogenic peptide to
antigen-specific T cells by fibroblast-like synoviocytes. Arthritis Rheum 2007;56:
1497.
10. Tran CN, Thacker SG, Louie DM, et al. Interactions of T cells with fibroblast-like
synoviocytes: role of the B7 family costimulatory ligand B7-H3. J Immunol
2008;180:2989.
11. Modak S, Kramer K, Gultekin SH, et al. Monoclonal antibody 8H9 targets a novel
cell surface antigen expressed by a wide spectrum of human solid tumors.
Cancer Res 2001;61:4048.
12. Xu H, Cheung IY, Guo HF, et al. MicroRNA miR-29 modulates expression of immunoinhibitory molecule B7-H3: potential implications for immune based therapy of
human solid tumors. Cancer Res 2009;69:6275.
13. Munn DH, Sharma MD, Mellor AL. Ligation of B7-1/B7-2 by human CD4 1 T cells
triggers indoleamine 2,3-dioxygenase activity in dendritic cells. J Immunol 2004;
172:4100.
14. Szanto S, Koreny T, Mikecz K, et al. Inhibition of indoleamine 2,3-dioxygenasemediated tryptophan catabolism accelerates collagen-induced arthritis in mice.
Arthritis Res Ther 2007;9:R50.
15. Criado G, Simelyte E, Inglis JJ, et al. Indoleamine 2,3 dioxygenase-mediated
tryptophan catabolism regulates accumulation of Th1/Th17 cells in the joint in
collagen-induced arthritis. Arthritis Rheum 2009;60:1342.
16. Zola H, editor. Leucocyte and stromal cell molecules: the CD markers. Hoboken
(NJ): Wiley-Liss; 2007. p. 412.
17. Ohshima Y, Tanaka Y, Tozawa H, et al. Expression and function of OX40 ligand on
human dendritic cells. J Immunol 1997;159:3838.
18. Brugnoni D, Bettinardi A, Malacarne F, et al. CD134/OX40 expression by synovial
fluid CD41 T lymphocytes in chronic synovitis. Br J Rheumatol 1998;37:584.
19. Yoshioka T, Nakajima A, Akiba H, et al. Contribution of OX40/OX40 ligand
interaction to the pathogenesis of rheumatoid arthritis. Eur J Immunol 2000;
30:2815.
20. Giacomelli R, Passacantando A, Perricone R, et al. T lymphocytes in the synovial
fluid of patients with active rheumatoid arthritis display CD134-OX40 surface
antigen. Clin Exp Rheumatol 2001;19:317.
21. Passacantando A, Parzanese I, Rascente M, et al. Synovial fluid OX40T lymphocytes of patients with rheumatoid arthritis display a Th2/Th0 polarization. Int J
Immunopathol Pharmacol 2006;19:499.
22. Hirota K, Yoshitomi H, Hashimoto M, et al. Preferential recruitment of CCR6-
expressing Th17 cells to inflamed joints via CCL20 in rheumatoid arthritis and
its animal model. J Exp Med 2007;204:2803.
23. Ruth JH, Shahrara S, Park CC, et al. Role of macrophage inflammatory protein-
3alpha and its ligand CCR6 in rheumatoid arthritis. Lab Invest 2003;83:579.
24. Kageyama Y, Ichikawa T, Nagafusa T, et al. Etanercept reduces the serum levels
of interleukin-23 and macrophage inflammatory protein-3 alpha in patients with
rheumatoid arthritis. Rheumatol Int 2007;28:137.
25. Chabaud M, Page G, Miossec P. Enhancing effect of IL-1, IL-17, and TNFalpha on macrophage inflammatory protein-3alpha production in rheumatoid
arthritis: regulation by soluble receptors and Th2 cytokines. J Immunol 2001;
167:6015.
26. Tanida S, Yoshitomi H, Nishitani K, et al. CCL20 produced in the cytokine network
of rheumatoid arthritis recruits CCR61 mononuclear cells and enhances the
production of IL-6. Cytokine 2009;47:112.
Cell-cell Interactions in Rheumatoid Arthritis Synovium 321
27. Kawashiri SY, Kawakami A, Iwamoto N, et al. Proinflammatory cytokines synergistically enhance the production of chemokine ligand 20 (CCL20) from rheumatoid fibroblast-like synovial cells in vitro and serum CCL20 is reduced in
vivo by biologic disease-modifying antirheumatic drugs. J Rheumatol 2009;
36:2397.
28. Timmer TC, Baltus B, Vondenhoff M, et al. Inflammation and ectopic lymphoid
structures in rheumatoid arthritis synovial tissues dissected by genomics technology: identification of the interleukin-7 signaling pathway in tissues with
lymphoid neogenesis. Arthritis Rheum 2007;56:2492.
29. van Roon JA, Jacobs K, Verstappen S, et al. Reduction of serum interleukin 7
levels upon methotrexate therapy in early rheumatoid arthritis correlates with
disease suppression. Ann Rheum Dis 2008;67:1054.
30. Hartgring SA, van Roon JA, Wenting-van Wijk M, et al. Elevated expression of
interleukin-7 receptor in inflamed joints mediates interleukin-7-induced immune
activation in rheumatoid arthritis. Arthritis Rheum 2009;60:2595.
31. Levesque MC. Translational mini-review series on B cell-directed therapies:
recent advances in B cell-directed biological therapies for autoimmune disorders. Clin Exp Immunol 2009;157:198.
32. Humby F, Bombardieri M, Manzo A, et al. Ectopic lymphoid structures support
ongoing production of class-switched autoantibodies in rheumatoid synovium.
PLoS Med 2009;6:e1.
33. Manzo A, Vitolo B, Humby F, et al. Mature antigen-experienced T helper cells
synthesize and secrete the B cell chemoattractant CXCL13 in the inflammatory
environment of the rheumatoid joint. Arthritis Rheum 2008;58:3377.
34. Iwai H, Kozono Y, Hirose S, et al. Amelioration of collagen-induced arthritis by
blockade of inducible costimulator-B7 homologous protein costimulation. J Immunol 2002;169:4332.
35. Hu YL, Metz DP, Chung J, et al. B7RP-1 blockade ameliorates autoimmunity
through regulation of follicular helper T cells. J Immunol 2009;182:1421.
36. Matsushita T, Yanaba K, Bouaziz JD, et al. Regulatory B cells inhibit EAE initiation
in mice while other B cells promote disease progression. J Clin Invest 2008;118:
3420.
37. Coppieters K, Dewint P, Van Beneden K, et al. NKT cells: manipulable managers
of joint inflammation. Rheumatology (Oxford) 2007;46:565.
38. Parietti V, Chifflot H, Sibilia J, et al. Rituximab treatment overcomes reduction
of regulatory iNKT cells in patients with rheumatoid arthritis. Clin Immunol
2009;134:331–9.
39. Rochas C, Hillion S, Saraux A, et al. Transmembrane BAFF from rheumatoid
synoviocytes requires interleukin-6 to induce the expression of recombinationactivating gene in B lymphocytes. Arthritis Rheum 2009;60:1261.
40. Lai Kwan Lam Q, King Hung Ko O, Zheng BJ, et al. Local BAFF gene silencing
suppresses Th17-cell generation and ameliorates autoimmune arthritis. Proc Natl
Acad Sci U S A 2008;105:14993.
41. Take Y, Nakata K, Hashimoto J, et al. Specifically modified osteopontin in rheumatoid arthritis fibroblast-like synoviocytes supports interaction with B cells and
enhances production of interleukin-6. Arthritis Rheum 2009;60:3591.
42. Han X, Lin X, Seliger AR, et al. Expression of receptor activator of nuclear factorkappaB ligand by B cells in response to oral bacteria. Oral Microbiol Immunol
2009;24:190.
43. Szekanecz Z, Koch AE. Vascular involvement in rheumatic diseases: ‘vascular
rheumatology’. Arthritis Res Ther 2008;10:224.
322 Fox et al
44. Gal I, Bajnok E, Szanto S, et al. Visualization and in situ analysis of leukocyte trafficking into the ankle joint in a systemic murine model of rheumatoid arthritis.
Arthritis Rheum 2005;52:3269.
45. Sarraj B, Ludanyi K, Glant TT, et al. Expression of CD44 and L-selectin in the
innate immune system is required for severe joint inflammation in the proteoglycan-induced murine model of rheumatoid arthritis. J Immunol 1932;177:2006.
46. Lally F, Smith E, Filer A, et al. A novel mechanism of neutrophil recruitment in
a coculture model of the rheumatoid synovium. Arthritis Rheum 2005;52:3460.
47. McGettrick HM, Smith E, Filer A, et al. Fibroblasts from different sites may
promote or inhibit recruitment of flowing lymphocytes by endothelial cells. Eur J
Immunol 2009;39:113.
48. Binstadt BA, Patel PR, Alencar H, et al. Particularities of the vasculature can
promote the organ specificity of autoimmune attack. Nat Immunol 2006;7:284.
49. Silverman MD, Haas CS, Rad AM, et al. The role of vascular cell adhesion molecule 1/very late activation antigen 4 in endothelial progenitor cell recruitment to
rheumatoid arthritis synovium. Arthritis Rheum 2007;56:1817.
50. Izquierdo E, Canete JD, Celis R, et al. Immature blood vessels in rheumatoid
synovium are selectively depleted in response to anti-TNF therapy. PLoS One
2009;4:e8131.
51. Amin MA, Ruth JH, Haas CS, et al. H-2g, a glucose analog of blood group H
antigen, mediates mononuclear cell recruitment via Src and phosphatidylinositol
3-kinase pathways. Arthritis Rheum 2008;58:689.
52. Burman A, Haworth O, Hardie DL, et al. A chemokine-dependent stromal induction mechanism for aberrant lymphocyte accumulation and compromised
lymphatic return in rheumatoid arthritis. J Immunol 2005;174:1693.
Cell-cell Interactions in Rheumatoid Arthritis Synovium 323