Introduction

The mast cell has long been known to mediate important manifestations of allergic disease. Crosslinking of surface-bound IgE results in the immediate release of granule contents, including histamine, and the more gradual elaboration of other proinflammatory mediators. Clinical manifestations can range from seasonal allergic rhinitis to life-threatening anaphylaxis.

However, research over the past two decades has revealed that the role of mast cells is not limited to IgE-mediated immune responses. Mast cells express surface receptors for IgG, complement, and specific pathogen-associated molecular patterns. Mast cells are capable of phagocytosis, intracellular killing, and antigen presentation. Correspondingly, mice deficient in mast cells have been found to exhibit striking susceptibility to death from certain types of bacterial infection. Beyond the acute phase of the immune response, mast cells may participate in the response of tissue to injury by means of mediators that promote angiogenesis and fibrosis.

Recently, several laboratories have established that mast cells have a critical role in the pathogenesis of synovitis in a murine system with considerable similarity to rheumatoid arthritis (RA) [1, 2]. This finding has renewed interest in older histological data documenting prominent mast cell infiltrates in the rheumatoid synovium. We review here the functions of mast cells as a prelude to the discussion of the current state of knowledge about the role of mast cells in murine and human inflammatory arthritis.

Basic biology of mast cells

Mast cells are found principally in mucosae and in connective tissue, generally clustered at epithelial surfaces and around nerves and blood vessels [3]. They originate in bone marrow and circulate as CD34+ committed progenitor cells, differentiating into mature mast cells only after entry into the tissue [4, 5]. These mature cells may divide further. Tissue mast cells are highly heterogeneous, with great variability in size, granule contents, cytokine production and receptor expression; both in vitro experience and in vivo data suggest that this heterogeneity represents an exquisite developmental sensitivity to local signals [3]. Similarly, the maintenance of mast cells within tissues is controlled by the local environment, in particular the production of stem cell factor (SCF, c-kit ligand) by stromal cells [6]. Mature mast cells are also capable of trafficking, as shown by their recruitment to chemotactic stimuli such as RANTES and their efflux from tissue through lymphatic channels and possibly blood vessels [79].

Functions of mast cells

IgE-mediated activation

Mast cells express the high-affinity IgE receptor FcεR1, a tetrameric complex of an α chain (to which IgE binds), a β chain and a dimer of γ chains [10]. The γ chain is shared with other stimulatory receptors, including the high-affinity IgG receptor FcγR1 and the low-affinity immune complex receptor FcγR3a. On crosslinking of the IgE receptor by multivalent antigen, the immunoreceptor tyrosine-based activation motifs (ITAMs) on the β and γ chains become phosphorylated and initiate a signaling cascade, resulting in three distinct pathways of mediator production: explosive release of preformed mediators, elaboration of eicosanoids, and de novo synthesis of cytokines and chemokines.

Explosive release of preformed mediators

Within seconds to minutes of IgE crosslinking, granules in the cytoplasm of the mast cell fuse with each other and with the cell surface membrane, ejecting their contents into the extracellular milieu. The contents of the granules depend on the conditions under which the mast cell has matured, but include histamine, proteoglycans (for example heparin), and a series of neutral proteases broadly grouped into tryptases, chymases, and carboxy-peptidases. Histamine promotes vascular permeability; proteoglycans provide a scaffold within the granule that allows the packaging of proteases; and the neutral proteases cleave proteins from matrix and plasma in addition to activating propeptides such as the precursors for interleukin-1β (IL-1β) and angiotensin II. The tryptase mMCP6 (murine mast cell protease 6) also contributes potently to neutrophil chemotaxis [11]. Certain subsets of mast cells store tumor necrosis factor (TNF) within the granules as well, representing the body's only source of TNF available for immediate release [12].

Elaboration of eicosanoids

Within minutes of IgE-mediated activation, mast cells begin to generate eicosanoids derived from cleavage of arachidonic acid from membrane phospholipids [13]. Important arachidonic acid metabolites include the leukotrienes (leukotriene B4 and the cysteinyl leukotrienes), which increase vascular permeability, induce vasoconstriction and recruit leukocytes, and prostaglandins including the neutrophil chemoattractant and vasoactive mediator prostaglandin D2.

De novosynthesis of cytokines and chemokines

Within hours, a later phase of mast cell activation through IgE becomes evident with the induction of new gene transcription and translation, generating a host of cytokines and chemokines (Table 1). The mix of cytokines generated by a particular mast cell depends on its individual state of differentiation.

Table 1 Selected mast cell mediators and their potential roles in arthritis

The importance of IgE-mediated mast cell activation to the health of the organism is still incompletely defined. The preservation of this system under evolutionary pressure, despite allergic diseases and anaphylaxis, is strong suggestive evidence that there is benefit to the host. One likely candidate function is resistance to parasitic disease, because mice deficient in IgE exhibit impaired defense against the helminths Schistosoma mansoni and Trichinella spiralis [70].

At the cellular level, the concept of the mast cell as immune sentinel led to the hypothesis that this lineage might participate pathogenically in autoantibody-driven K/BxN serum transfer arthritis. Expressing receptors for both immune complexes and complement, synovial mast cells would be well positioned to initiate the tissue response to K/BxN serum. Consistent with this hypothesis is the observation that mice deficient in mast cells are highly resistant to arthritis, whereas reconstitution with normal mast cells restores the wild-type phenotype (Fig. 2). Furthermore, degranulation of mast cells in the synovium is the first event observed histologically, occurring within 1–2 hours of administration of K/BxN serum [1]. Thus, as in antibody-mediated peritonitis, synovial mast cells seem to act as early responders, mobilizing the inflammatory response against a perceived insult. In their absence, no other cell constitutively resident within the synovium or present in the circulation seems to have the capacity to initiate the recruitment of inflammatory cells to the joint that characterizes arthritis in the wild-type animal. However, details of the mechanisms of mast cell activation as well as the relevant mast cell effector functions in this model remain to be defined.

Figure 2
figure 2

Mast cells constitute a critical pathogenic link in K/BxN serum transfer arthritis. Compared with wild-type controls, mast-cell-deficient W/Wv mice injected with K/BxN arthritogenic serum are resistant to the development of arthritis. After reconstitution with cultured wild-type mast cells, but not sham reconstitution, normal susceptibility is restored. Error bars = SEM. (Adapted from reference [1], with permission.)

Mast cells and the initiation of human synovitis

The involvement of mast cells in the earliest phases of human synovitis remains a subject for conjecture. As described previously, mast cells can be triggered by IgG immune complexes, complement, TLR ligands, and microbial antigens. Each of these stimulatory pathways may be of relevance to human arthritis. Immune complexes are thought to cause the arthritis of serum sickness and cryoglobulinemia but have also been documented in the serum, synovial fluid, synovium, and cartilage of patients with RA and are once again a field of active investigation in the pathogenesis of RA [7174]. Complement activation has similarly been well documented within rheumatoid synovium [75]. Infection with bacteria or viruses could trigger mast cell activation by means of TLRs and specific pathogen receptors. Even in the absence of infection, mast cells could be stimulated via TLRs by synovial constituents with TLR ligand activity, including heat shock protein 60 and breakdown products of hyaluronan, potentially amplifying any inflammatory process within the joint [76]. Mast cell IgE receptors might also have a role in a small subset of patients, because IgE rheumatoid factors and IgE-containing immune complexes have been documented in some patients with RA [77, 78].

Once activated, mast cells in the synovium would be expected to initiate inflammation through several mechanisms; a limited number of candidate pathways are outlined in Fig. 3. Vasoactive mediators such as histamine, prostaglandin D2, and the leukotrienes increase vascular permeability, whereas TNF, IL-1, and histamine promote the expression of the adhesion molecules P-selectin, E-selectin, ICAM-1, and VCAM-1 on the endothelial surface [79, 80]. Circulating leukocytes bearing appropriate counter-receptors, such as leukocyte function-associated antigen-1 (LFA-1) (itself of heightened affinity under the influence of proinflammatory cytokines through 'inside-out' regulation), could then be recruited into the synovium along gradients of chemotactic mast cell products such as leukotriene B4, monocyte chemoattractant protein-1, tryptases (for example mMCP6), and IL-8. Activation of resident synovial macrophages and arriving monocytes and neutrophils by means of interferon-γ, IL-6 and TNF would be expected to result in further amplification of leukocyte recruitment and an enhanced output of proinflammatory cytokines.

Figure 3
figure 3

Candidate proinflammatory functions of mast cells in synovitis. Mast cell effector functions suggest their participation in diverse pathogenic pathways in inflammatory arthritis, including leukocyte recruitment and activation, synovial fibroblast activation and hyperplasia, angiogenesis, and cartilage and bone destruction. Activated mast cells elaborate mediators potently capable of enhancing vasopermeability, inducing endothelial expression of adhesion molecules, recruiting circulating leukocytes, and activating infiltrating leukocytes as well as resident macrophages, thereby contributing to the early phases of inflammatory arthritis. In chronic synovitis, mast cells synthesize mitogens and cytokines that activate synovial fibroblasts, recruit macrophages, and promote the growth of new blood vessels, implicating them in synovial lining hyperplasia and pannus formation. Further, mast cells may participate in joint destruction by the induction of matrix metalloproteinases (MMPs) from fibroblasts, by activation of chondrocytes, and by direct and indirect promotion of osteoclast differentiation and activation. Because activated synovial fibroblasts demonstrate enhanced stem cell factor (SCF) expression, a potentially important positive feedback loop is established in which SCF promotes mast cell survival and proliferation, leading to the mastocytosis described in inflamed synovium. Note that the importance of these candidate pathways in vivo remains to be established. See text for details and references. bFGF, basic fibroblast growth factor; IFN, interferon; IL, interleukin; MCP, monocyte chemoattractant protein; M-CSF, macrophage colony-stimulating factor; MIP, macrophage inflammatory protein; PDGF, platelet-derived growth factor; PMN, polymorphonuclear cell; RANK-L, receptor activator of NF-κB ligand; TNF, tumor necrosis factor. (Graphic design by Steve Moskowitz.)

Beyond the 'jump start': a role for mast cells in chronic synovitis in mouse and humans?

In some murine models of bacterial and antibody-induced disease, the physiological role of mast cells can largely be replaced by a single administration of neutrophils or neutrophil chemoattractants [17, 31, 35, 38]. This observation suggests that mast cells have no substantial continuing role in these pathologic states. In K/BxN arthritis, and potentially in human arthritis, is there a role for the synovial mast cells beyond the initiation of synovitis?

An initial observation applies. In K/BxN serum transfer arthritis, two serum injections are followed within 1–3 days by an intense synovitis. This reaction peaks over the course of 2 weeks but is ultimately self-limiting, resolving within 6 weeks. Although some human joint diseases run such a self-limited course (such as serum sickness and postviral arthritis), many human arthritides are chronic. In such chronic conditions, any factors inducing mast cell activation might well be persistent. This is so in K/BxN mice, which exhibit a progressive erosive arthritis in the setting of persistently high levels of autoantibodies in the serum. 'Chronicity' can be mimicked in wild-type mice by means of a repeated transfer of K/BxN serum. In this setting, synovial mast cells can undergo repetitive cycles of activation and thus participate in ongoing disease much more substantially than has been observed in models of peritonitis and skin disease. Indeed, degranulating synovial mast cells are readily observed in established K/BxN arthritis [1]. Yet a functional contribution of mast cells to continuing inflammation remains to be experimentally determined.

In humans, given the expanded numbers of mast cells within the joint and their enormous capacity for the production of cytokines and chemokines, it would be surprising indeed if they were of no consequence to the chronic inflammatory response. The broad range of mast cell effector functions includes the elaboration of mediators with bioactivity directed at marrow-derived leukocytes as well as mesenchymal tissue elements (Fig. 3). Because the pathogenic state of inflammatory arthritis displays prominent responses by both infiltrating leukocytes and mesenchymal cells, in particular synovial fibroblasts, we will examine the potential influence of mast cells on both compartments in arthritis.

Mast cells and synovial leukocytes

The rheumatoid synovium is thick with infiltrating leukocytes. These include T lymphocytes, B lymphocytes, macrophages, mast cells and scattered neutrophils. Ongoing recruitment of these cells results from the upregulation of selectins and integrins on synovial endothelium, allowing migration up chemotactic gradients into the joint. The composition of inflammatory cells recruited in a continuing fashion by mast cells, including the degree of skewing of lymphocytes toward Th1 versus Th2 responses, might be an important determinant of the ultimate outcome of inflammation. The production of anti-inflammatory mediators by mast cells remains uncharacterized [81].

Prominent within the rheumatoid synovium is a greatly expanded population of synovial macrophages. These cells do not proliferate locally but instead are recruited from circulating monocytes [82]. Mast cells are potent sources of chemokines that mediate this recruitment, including IL-8, monocyte chemoattractant protein-1, MIP-1α, and RANTES [3]. Mast cells might also contribute to the activation of these macrophages through the production of interferon-γ and IL-6. Because macrophages are major sources of the proinflammatory cytokines TNF and IL-1 within the joint, mast cell effects on the size and activation state of the synovial macrophage population might functionally modulate the course of inflammatory arthritis.

Mast cells and the synovial mesenchyme

The synovial mesenchyme, consisting principally of synovial fibroblasts, is prominently involved in joint inflammation. Fibroblasts increase greatly in numbers and assume a histological appearance suggestive of increased synthetic activity, with expansion of the endoplasmic reticulum and increased numbers of granules in the cytoplasm [83]. Indeed, synovial fibroblasts make up the shroud-like pannus characteristic of the rheumatoid joint and are an important source of multiple mediators implicated in arthritis. These include degradative enzymes such as collagenase and stromelysin and proinflammatory molecules including IL-1, IL-6, and prostaglandin E2 (reviewed in [84]). They contribute to the differentiation and activation of osteoclasts, the effector cell responsible for bone erosions, through the production of macrophage colony-stimulating factor (M-CSF) and receptor activator of NF-κB ligand (RANKL) [85, 86].

Mast cells may potently influence synovial fibroblast biology in RA. Consistent with a proposed role in wound healing and in multiple fibrotic disease states, mast cells produce a range of mediators with powerful effects on fibroblasts (Table 1) [87]. Further, synovial mast cells are often noted in close physical proximity to synovial fibroblasts [50]. Mast cell tryptase promotes chemotaxis and collagen synthesis in fibroblasts, and histamine stimulates fibroblast proliferation [8890]. Other fibroblast mitogens produced by mast cells include nerve growth factor, basic fibroblast growth factor, platelet-derived growth factor, vascular endothelial growth factor (VEGF), and transforming growth factor-β (TGF-β) [91]. The cytokine IL-4, produced predominantly by mast cells of a tryptase–chymase phenotype, induces proliferation and collagen production by fibroblasts [92], and indeed, as noted above, MCTC cells tend to reside in more fibrotic areas of the inflamed joint. Because leukotriene C4 seems to have antifibrotic effects, it remains possible that mast cells can limit as well as promote fibrosis, although scattered foci of fibrosis associated with mast cell infiltrates in systemic mastocytosis suggest a net profibrotic effect [91, 93, 94].

Mast cells may also potentiate mediator production by synovial fibroblasts through the elaboration of cytokines such as TNF and IL-1. IL-1 induces the elaboration of collagenase and prostaglandin E2, and TNF elicits similar responses while also inducing synovial fibroblasts to generate IL-1 [9597]. Indeed, the production of collagenase and other inflammatory products of fibroblasts has been noted to localize to the immediate environment of activated mast cells [98].

This communication between mast cells and synovial fibroblasts is bidirectional. Mast cells require stimulation by SCF for differentiation in situ as well as activation [6]. Fibroblasts in inflamed or healing tissues express higher levels of SCF, and upregulation of SCF expression has been noted in synovial specimens exposed to TNF [99101]. Indeed, such surface expression seems to be of particular importance to mast cell development, because Sl/Sld mice unable to display surface-bound SCF lack tissue mast cells despite an intact production of soluble SCF [102, 103]. Further, transwell experiments demonstrate that physical contact is required for certain stimulatory effects of fibroblasts on mast cells [104, 105]. Fibroblasts might also promote the survival of mast cells by means of SCF-independent pathways yet to be fully defined [106].

In addition to fibroblasts, the synovial mesenchyme also contains blood vessels. As would be expected, the expanded cellular population in the inflamed synovium requires an enhanced blood supply, and neoangiogenesis has an important pathophysiological function in RA. Mast cell mediators implicated in the promotion of angiogenesis include heparin, vascular endothelial growth factor, TGF-β, TNF, IL-1, and IL-18 [42, 107]. Further, TNF can induce synovial fibroblast production of another pro-angiogenic factor, angiopoietin-1 [108]. Though the ultimate importance of mast cells in synovial angiogenesis remains unclear, the association of mast cells with blood vessels, including newly develo** blood vessels, makes the promotion of angiogenesis a plausible role for mast cells in vivo (reviewed in [109]).

Finally, some data suggest that mast cell mediators might exert a direct effect on cartilage and bone. Thus, whereas the coculture of chondrocytes with inactive mast cells tends to promote the synthesis of proteoglycans, the activation of mast cells in this context favors proteoglycan degradation [110]. Further, the activation of chondrocytes via IL-1, TNF, and histamine might induce the production of matrix metalloproteinases and prostaglandins [111, 112]. Finally, mast cell mediators including histamine and MIP-1α might directly promote the differentiation and activation of osteoclasts, the final common pathway of bone destruction in inflammatory arthritis [113115]. Corroboration in vivo will be required to establish the importance of these in vitro findings.

Conclusions

Mast cells are a normal cell population within the human synovium, and in line with their role as sentinels they likely have an important physiological role as an 'early warning system' for infection within the vulnerable joint cavity. Data from the K/BxN mouse model now show that mast cells also have a critical role in the pathogenesis of inflammatory arthritis, in particular in arthritis induced by autoantibody-containing immune complexes. Although a similar mechanism remains unproven for human joint inflammation, markers of mast cell activation are observed in joint fluid from patients with chronic arthritis and mast cell numbers are often greatly expanded within the inflamed synovium. Equipped with an impressive array of mediators, mast cells can promote synovitis by recruiting inflammatory cells from the blood, inducing synovial fibroblast hyperplasia and mediator production, and fostering angiogenesis. Although much remains to be learned about the role of the mast cell in arthritis, such a role now seems highly likely, offering a potential new target for therapeutic agents in the treatment of RA and other inflammatory diseases of the joints.