Introduction

Rheumatoid Arthritis (RA) is a chronic and systemic inflammatory disease characterized by autoimmune-mediated attack of the joints . A hallmark of this disease is inflammation of the synovial membrane (synovitis) characterized by infiltration of multiple immune lineages, concomitant joint swelling and tenderness, systemic inflammation with elevation of acute phase reactants, and production of autoantibodies such as Rheumatoid Factor (RF) directed against the Fc component of IgG, and anti-citrullinated protein antibodies (ACPAs). These processes lead ultimately to destruction of joint tissues with erosion of bone by osteoclasts and degradation of cartilage by proteases.

Treatment of RA patients with Disease Modifying Anti-Rheumatic Drugs (DMARDs), comprising both targeted therapies such as anti-TNFα, as well as therapies with broader anti-inflammatory mechanisms of action such as methotrexate, have demonstrated significant clinical benefit in reducing signs and symptoms as well as decreasing joint destruction. However, these therapies are not effective in all patients, and not all patients demonstrate a satisfactory clinical outcome. Indeed, both randomized placebo-controlled clinical trials as well as open label studies of multiple distinct therapies show a proportion of patients are resistant to therapy and continue to present with synovitis and ongoing joint destruction . It is likely that underlying disease heterogeneity contributes to incomplete drug response, and underscores the importance of a deeper understanding of disease pathogenesis to better treat patients by matching them to specific therapies as well as identifying new therapeutic drug targets. Such approaches could indicate approaches to identify biomarker-defined RA patient subpopulations that could be used for clinical decision-making in terms of disease prognosis and choice of appropriate therapy. In this review, we discuss current knowledge around synovial cellular and molecular heterogeneity, clinical correlations with these disease processes, and their impact on and response to drug treatment.

Cellular and molecular subsets of synovitis

The primary manifestation of RA is autoimmune-mediated synovitis involving the large-scale infiltration of leukocytes into the synovial tissues. Importantly, however, is the emerging understanding that multiple processes precede the onset of synovitis by quite a considerable time period. It has been recognized for many years that the strongest genetic association lies within the HLA-DRB1 locus, and that alleles containing a ‘shared epitope’ within this region are strongly associated with antibody reactivity to anti-citrullinated proteins , likely through preferential binding of post translationally-modified peptides for the binding groove of this HLA. Other risk alleles lie within regions associated with T cell receptor signaling such as PTPN22 , the NF-κB pathway such as TRAF1-C5 , and regulation of TNFα and IL-12 signaling such as TNFAIP3 and STAT4 . A recent large meta-analysis of ∼30,000 RA cases and >70,000 control subjects confirmed prior identified risk loci and identified a further 42 novel loci to bring the total number of RA risk alleles with genome wide significance to over 100 . Strikingly many of these loci lie within or adjacent to immune system associated genes, indicating that genetic risk is tied to immunological perturbation. Furthermore, assessment of the joints by magnetic resonance imaging and biopsy of individuals who are seropositive for prototypical autoantibodies such as RF and ACPAs but who have not yet developed fulminant disease indicate that the synovial lining is relatively normal, with only minor infiltration of T cells being associated with subsequent development of disease . Such data suggests that a break in immune tolerance and development of auto-reactive lymphocytes has its genesis in other anatomic sites. Attention has focused on the presence of infectious agents such as Escherichia coli that may trigger the immune system and result in autoimmunity through molecular mimicry of self-antigens . In particular the presence of the micro-organism Porphyromonas gingivalis , present in periodontal disease, has been associated with RA through its production of Peptidyl Arginine Deiminase (PADI) 4, an enzyme that catalyzes citrullination of native arginine residues in host proteins . Thus, a systemic break in tolerance occurs prior to onset of pathophysiology in the joints, and magnification of this response through epitope spreading to additional self-antigens present in joints can lead to onset of synovitis .

Synovitis itself is characterized by several closely interlinked processes, and is characterized by the invasion of the synovial lining by leukocytes including B cells, T cells, macrophages and mast cells with concomitant expression of a multitude of inflammatory cytokines and chemokines. Auto reactive T cells have long been appreciated as contributors to synovitis, and are supported by abundant levels of myeloid cells and dendritic cells in the synovium that produce T cell stimulating cytokines such as IL-12, IL-18 and IL-23, MHC class II molecules that present self antigen, and co-stimulatory molecules such as CD80 and CD86 that strengthen the T cell-antigen presenting cell interaction . Indeed, the proven efficacy of the CTLA4-IgG1 Fc molecule abatacept in decreasing RA synovitis and disease activity underscores the importance of these interactions . T cells themselves are present in multiple subtypes including CD4 ⁺ and CD8 ⁺ T cells, and as both effector and regulatory cells with opposing functions. Recent studies have shown that while CD4 ⁺ T helper (Th) 1 cells, producing the hallmark cytokine IFNγ, are important contributors to disease, CD4 ⁺ Th17 cells producing IL-17A, IL-17F, IL-21, IL-22, IL-23 and TNFα, also play an important role . IL-17 is notable for its ability to synergize with TNFα to mediate activation of synovial fibroblasts and osteoclasts thus driving structural damage to the joint .

Auto reactive B cells are also hallmarks of the rheumatoid synovium. Like many inflammatory diseases, B cells exist in the site of inflammation within localized lymphoid aggregates and, in some patients, as true tertiary lymphoid follicles with associated follicular dendritic cell networks although the latter is not required for B cell auto reactivity . A multitude of growth factors are also present to support B cell infiltration and survival including B cell activating factor of the TNF family (BAFF), A proliferating-inducing ligand (APRIL), and a spectrum of chemokines . These B cell regions are also a local source of autoantibodies including ACPAs , and the resulting plasma cells are present both in synovial lining and adjacent bone marrow . The important role for B cells is underscored by the positive treatment benefit of RA patients by therapies targeting B cells such as rituximab .

The innate immune system is also widely appreciated to play a critical role in the pathophysiology of synovitis. Infiltrating macrophages of the pro-inflammatory M1 phenotype produce a wide spectrum of inflammatory cytokines including TNFα, IL-1, IL-6, IL-12, IL-15, IL-18 and IL-23, and have been shown to be central to synovitis as successful therapeutic treatment of disease and reduction of tissue inflammation reduces levels of synovial macrophages . These inflammatory cytokines themselves have shown to play critical roles in disease as, for example, therapeutic blockade of the TNFα and IL-6 pathways along with signaling molecules such as JAK1 have shown positive clinical benefits . Other innate immune system contributors to synovitis include neutrophils, largely resident in synovial fluid, that produce a spectrum of prostaglandins, proteases, and reactive oxygen species , while infiltrating mast cells have been observed to be a significant source of cytokines, chemokines, proteases and vasoactive amines .

Beyond activation of the immune system through autoimmunity, a further hallmark of synovitis is activation of synovial fibroblasts derived from mesenchymal precursors. While normally present as a thin lining layer under conditions of health, a notable feature of the RA synovium is massive proliferation of this lineage accompanied by a change in cell behavior characterized by loss of contact inhibition and anchorage dependence. Further, these cells are a significant source of proteases such as matrix metalloproteinases and inflammatory cytokines . There is evidence that these cells arise by endogenous mechanisms including epigenetic changes and regulation of micro RNAs, and that they are capable of migrating and infiltrating joints themselves . These cells thus serve both to provide a pro inflammatory environment for the infiltrating leukocytes and also as direct effector cells of cartilage destruction. These processes are accompanied by neovascularization in the context of local tissue hypoxia within the synovial lining that facilitates leukocyte infiltration, and has been linked with the production of pro-angiogenic factors such as VEGF . The end result of these multitudes of processes is joint swelling, systemic inflammation, and structural damage to the joint. Cartilage damage occurs through the production of metalloproteinases and cytokine-mediated death of chondrocytes resulting in radiographic joint space narrowing. Bone erosion, resulting in radiographically visible damage, occurs through perturbation in the balance between bone resorbing osteoclasts and bone producing osteoblasts mediated in part by amplification of osteoclast differentiation by receptor activator of NF-κB ligand (RANKL) and pro-inflammatory cytokines including TNFα and IL-17. These processes are summarized in Fig. 1 .

Major cellular processes present within the synovium of patients with Rheumatoid Arthritis. Disease pathophysiology is characterized by infiltration of immune cells including B cells, T cells, inflammatory macrophages, neutrophils and mast cells, activation of osteoclasts resulting in increased bone erosion, and activation and hyperplasia of synovial fibroblasts resulting in production of cartilage-damaging proteases.

One overarching question is whether these processes of autoimmune-mediated synovitis are present in all patients at all stages of disease, or instead are differentially active in different patients resulting in both heterogenous synovitis as well as clinically heterogenous disease. These questions have been tackled using multiple techniques over the past years in patients with varying stages of disease. In particular, the ability to sample synovial tissue either through the use of blind needle biopsies, visually guided arthroscopic biopsies and most recently ultrasound guided biopsies that combines an imaging assessment of synovial lining thickness and vascularity as well as optimally guiding the site of tissue sampling to areas of maximal inflammation . Further assessment can be done with larger tissue specimens obtaining through arthroscopic surgery and during joint replacement, although such patients are typically further advanced in their course of disease. One obvious caveat with biopsy approaches is ensuring that representative sampling occurs, as results may otherwise be misleading if key inflammatory areas are missed. The imaging guided techniques described above coupled with computer-assisted digital image analysis can mitigate some of this issue. Another is the invasive nature of the procedures, which means that of the multitude of swollen and tender joints affected in a given RA patient, typically only one joint can be biopsied at a time. One study assessed histological similarities between inflamed small joint and knee joint biopsies taken from the same RA patient, and concluded there were no major differences between the biopsies in terms of sub lining macrophages, T cells and plasma cells , although there were no correlations observed for intimal macrophages and synovial fibroblasts. Thus synovial tissue may be quite comparable across affected joints at least in terms of infiltrating immune cells. Nonetheless, caution must be taken when interpreting biopsy studies and extrapolating findings to the disease in its entirety.

In addition to the multitudinous features of synovitis described above, it was noted many years ago that infiltrating immune cells were present in a variety of histologic forms. Cells were present within the synovial sub lining in some patients as a diffuse infiltration with no higher order structure, or alternatively could be present as lymphoid aggregates or even tertiary lymphoid follicles in about 30% of patients . These aggregates and follicles were associated with the presence of neo-angiogenesis and high endothelial venules and follicular dendritic cell networks . Further, B cells within these aggregates have been shown to have specificity to self antigens , produce higher levels of autoantibodies and express activation-induced cytidine deaminase that regulated somatic hypermutation and class switch recombination . It remains unclear as to whether these structures represent clinically distinct subsets of RA, however, as comparing patients who have synovial lymphoid neo-angiogenesis with patients without these structures have not yielded significant differences in clinical activity parameters such as DAS28 scores and presence of autoantibodies, although C-reactive protein (CRP) levels, Erythrocyte sedimentation rates (ESR) and levels of peripheral leukocytes were elevated in patients with lymphoid neogenesis . Consistent with this observation, it was also shown that B cell autoimmunity was not dependent on the presence of lymphoid neogenesis and was not required for local production of autoantibodies . It is likely that defining RA patients on the basis of these histologic patterns may be oversimplifying the ongoing disease processes, as these patterns may not be absolutely distinct. Indeed, it has been observed that individual tissues may contain varying levels of both forms of inflammation, and these structures may rather reflect a continuum of inflammation rather that wholly discrete disease subsets. When histologically defined, B cells are present at maximal numbers in patients with lymphoid aggregates, while present at much lower levels in synovium with a diffuse histological phenotype, where instead there are a predominance of monocytes and T cells. It is also noteworthy that patients who are positive for ACPAs were shown to have greater levels of infiltrating synovial leukocytes and CXCL12 as compared to ACPA ⁻ patients, and also had a higher rate of radiographically-determined joint destruction . This finding is in good agreement with the well validated finding that presence of ACPAs and other autoantibodies are prognostic for development of more severe erosive RA . In connection with this, it has been reported that patients with synovial aggregates have increased presence of erosive disease, and this is intriguing since B cells have been connected with expression of TNF-like weak inducer of apoptosis (TWEAK) that induces RANKL as well as activation through MHC class 2/T cell receptor of synovial T cells that express high levels of this ligand . Importantly, a further group of patients has been observed, tentatively assigned ‘pauci-immune’, where there a very few infiltrating leukocytes despite active disease, and these patients continue to have increased fibroplasia, elevated swollen and tender joint counts and presence of autoantibodies .

Genomics technologies have allowed a further dissection into the molecular heterogeneity of RA synovitis. A large body of work has been reported in recent years using expression microarrays and quantitative PCR methodologies to study differences between RA and healthy or osteoarthritic synovial tissue, or in synoviocytes . Such studies are advantageous as they allow for comparatively unbiased assessments of disease-associated gene expression pathways (for example using unsupervised hierarchical agglomerative clustering), and as gene expression profiling technologies have evolved over time they progressively allow for deeper appreciation of molecular heterogeneity. These studies are summarized in Table 1 . As would be expected, the RA synovium is highly enriched for genes associated with both the adaptive and innate immune systems, inflammatory cytokines and their associated signaling transcription factors as well as matrix-destructive proteases . Furthermore, evidence was found of activation of the IL-7 pathway, JAK/STAT pathways and lymphoid neogenesis-associated genes such as CXCL13, CCL21, CCR7 and LTα in patients with lymphoid aggregates . Genomics studies also interrogated synovial RNA from patients at different stages of disease progression, and showed that some early RA patients resembled long standing RA, while others formed a distinct subset with differing levels of inflammatory gene expression . A further study comparing early versus long standing RA confirmed these findings by showing that distinct biological processes are differentially present between early and late stage RA including higher expression of processes pertaining to granulocytes, cell cycle, macrophages and T cells in later disease . This differential presence of inflammatory pathways in synovium is borne out in a highly characterized RA cohort where a high inflammatory phenotype, defined by genomic analysis and correlating with synovial histologic findings for T cells, B cells, macrophages and cytokines, had higher DAS28 scores, higher CRP and ESR levels, elevated platelets and decreased disease duration . A separate study showed consistent results, with a later-stage RA cohort demonstrating high inflammatory and low inflammatory subsets defined by immune gene expression that strongly correlated with synovial histology, and the high inflammatory subset had elevated CRP and ESR levels and DAS28 scores, although the DAS elevation was driven primarily by the acute phase reactants while swollen and tender joint counts were not different between the two subsets . Further dissection of subtypes of RA have been undertaken through genomic studies of synovial tissue in larger patient cohorts, which have revealed molecular subtypes of disease that agree well with prior histologic classification of patients. These studies show that there are several molecular phenotypes that primarily drive patient segregation, including the presence of B cell-rich aggregates, presence of inflammatory macrophages but with few B cells, a stromal/pauci-immune phenotype and more intermediate subset containing features of all of these phenotypes . The presence of a macrophage-rich subset is notable given the central role for inflammatory macrophages in production of key inflammatory cytokines and their correlation with tissue inflammation, joint destruction and response to therapy . It is again important to note that these genomic studies, as was noted with the previously described histologic studies, do not point to discrete subsets of disease but rather represent a continuum of disease processes. In particular, analyses encompassing principle components analysis (PCA) and partitioning around medoids (PAM) methodologies underscored this fact by showing there was comparatively little clear separation between RA synovial samples at a genome-wide level, unlike what can been seen with similar analyses in tumor samples . It is therefore likely that the spectrum of RA clinical disease activity, characterized by varying numbers of affected joints, course of disease, and levels of acute phase reactants and autoantibodies is paralleled by a spectrum of inflammation within the synovial tissue.

Cellular and molecular subsets of synovitis

The primary manifestation of RA is autoimmune-mediated synovitis involving the large-scale infiltration of leukocytes into the synovial tissues. Importantly, however, is the emerging understanding that multiple processes precede the onset of synovitis by quite a considerable time period. It has been recognized for many years that the strongest genetic association lies within the HLA-DRB1 locus, and that alleles containing a ‘shared epitope’ within this region are strongly associated with antibody reactivity to anti-citrullinated proteins , likely through preferential binding of post translationally-modified peptides for the binding groove of this HLA. Other risk alleles lie within regions associated with T cell receptor signaling such as PTPN22 , the NF-κB pathway such as TRAF1-C5 , and regulation of TNFα and IL-12 signaling such as TNFAIP3 and STAT4 . A recent large meta-analysis of ∼30,000 RA cases and >70,000 control subjects confirmed prior identified risk loci and identified a further 42 novel loci to bring the total number of RA risk alleles with genome wide significance to over 100 . Strikingly many of these loci lie within or adjacent to immune system associated genes, indicating that genetic risk is tied to immunological perturbation. Furthermore, assessment of the joints by magnetic resonance imaging and biopsy of individuals who are seropositive for prototypical autoantibodies such as RF and ACPAs but who have not yet developed fulminant disease indicate that the synovial lining is relatively normal, with only minor infiltration of T cells being associated with subsequent development of disease . Such data suggests that a break in immune tolerance and development of auto-reactive lymphocytes has its genesis in other anatomic sites. Attention has focused on the presence of infectious agents such as Escherichia coli that may trigger the immune system and result in autoimmunity through molecular mimicry of self-antigens . In particular the presence of the micro-organism Porphyromonas gingivalis , present in periodontal disease, has been associated with RA through its production of Peptidyl Arginine Deiminase (PADI) 4, an enzyme that catalyzes citrullination of native arginine residues in host proteins . Thus, a systemic break in tolerance occurs prior to onset of pathophysiology in the joints, and magnification of this response through epitope spreading to additional self-antigens present in joints can lead to onset of synovitis .

Synovitis itself is characterized by several closely interlinked processes, and is characterized by the invasion of the synovial lining by leukocytes including B cells, T cells, macrophages and mast cells with concomitant expression of a multitude of inflammatory cytokines and chemokines. Auto reactive T cells have long been appreciated as contributors to synovitis, and are supported by abundant levels of myeloid cells and dendritic cells in the synovium that produce T cell stimulating cytokines such as IL-12, IL-18 and IL-23, MHC class II molecules that present self antigen, and co-stimulatory molecules such as CD80 and CD86 that strengthen the T cell-antigen presenting cell interaction . Indeed, the proven efficacy of the CTLA4-IgG1 Fc molecule abatacept in decreasing RA synovitis and disease activity underscores the importance of these interactions . T cells themselves are present in multiple subtypes including CD4 ⁺ and CD8 ⁺ T cells, and as both effector and regulatory cells with opposing functions. Recent studies have shown that while CD4 ⁺ T helper (Th) 1 cells, producing the hallmark cytokine IFNγ, are important contributors to disease, CD4 ⁺ Th17 cells producing IL-17A, IL-17F, IL-21, IL-22, IL-23 and TNFα, also play an important role . IL-17 is notable for its ability to synergize with TNFα to mediate activation of synovial fibroblasts and osteoclasts thus driving structural damage to the joint .

Auto reactive B cells are also hallmarks of the rheumatoid synovium. Like many inflammatory diseases, B cells exist in the site of inflammation within localized lymphoid aggregates and, in some patients, as true tertiary lymphoid follicles with associated follicular dendritic cell networks although the latter is not required for B cell auto reactivity . A multitude of growth factors are also present to support B cell infiltration and survival including B cell activating factor of the TNF family (BAFF), A proliferating-inducing ligand (APRIL), and a spectrum of chemokines . These B cell regions are also a local source of autoantibodies including ACPAs , and the resulting plasma cells are present both in synovial lining and adjacent bone marrow . The important role for B cells is underscored by the positive treatment benefit of RA patients by therapies targeting B cells such as rituximab .

The innate immune system is also widely appreciated to play a critical role in the pathophysiology of synovitis. Infiltrating macrophages of the pro-inflammatory M1 phenotype produce a wide spectrum of inflammatory cytokines including TNFα, IL-1, IL-6, IL-12, IL-15, IL-18 and IL-23, and have been shown to be central to synovitis as successful therapeutic treatment of disease and reduction of tissue inflammation reduces levels of synovial macrophages . These inflammatory cytokines themselves have shown to play critical roles in disease as, for example, therapeutic blockade of the TNFα and IL-6 pathways along with signaling molecules such as JAK1 have shown positive clinical benefits . Other innate immune system contributors to synovitis include neutrophils, largely resident in synovial fluid, that produce a spectrum of prostaglandins, proteases, and reactive oxygen species , while infiltrating mast cells have been observed to be a significant source of cytokines, chemokines, proteases and vasoactive amines .

Beyond activation of the immune system through autoimmunity, a further hallmark of synovitis is activation of synovial fibroblasts derived from mesenchymal precursors. While normally present as a thin lining layer under conditions of health, a notable feature of the RA synovium is massive proliferation of this lineage accompanied by a change in cell behavior characterized by loss of contact inhibition and anchorage dependence. Further, these cells are a significant source of proteases such as matrix metalloproteinases and inflammatory cytokines . There is evidence that these cells arise by endogenous mechanisms including epigenetic changes and regulation of micro RNAs, and that they are capable of migrating and infiltrating joints themselves . These cells thus serve both to provide a pro inflammatory environment for the infiltrating leukocytes and also as direct effector cells of cartilage destruction. These processes are accompanied by neovascularization in the context of local tissue hypoxia within the synovial lining that facilitates leukocyte infiltration, and has been linked with the production of pro-angiogenic factors such as VEGF . The end result of these multitudes of processes is joint swelling, systemic inflammation, and structural damage to the joint. Cartilage damage occurs through the production of metalloproteinases and cytokine-mediated death of chondrocytes resulting in radiographic joint space narrowing. Bone erosion, resulting in radiographically visible damage, occurs through perturbation in the balance between bone resorbing osteoclasts and bone producing osteoblasts mediated in part by amplification of osteoclast differentiation by receptor activator of NF-κB ligand (RANKL) and pro-inflammatory cytokines including TNFα and IL-17. These processes are summarized in Fig. 1 .

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Related

Related posts:

Concepts of pathogenesis and emerging treatments for inflammatory arthritis How undifferentiated arthritis evolves into chronic arthritis Epigenetic changes: The missing link Small-molecule therapeutics in rheumatoid arthritis: Scientific rationale, efficacy and safety Inhibitors of angiogenesis: Ready for prime time? Seronegative rheumatoid arthritis: Pathogenetic and therapeutic aspects

Stay updated, free articles. Join our Telegram channel

Join