S. Appel PhD, Broegelmann Research Laboratory, The Gade Institute, Armauer Hansen Bldg, University of Bergen, N-5021 Bergen, Norway. E-mail: firstname.lastname@example.org
Sjögren's syndrome (SS) is a chronic inflammatory and lymphoproliferative autoimmune disease of unknown aetiology. It is characterised by progressive mononuclear cell infiltration of the salivary and lacrimal glands and a decreased glandular secretion, resulting in dryness of the mouth and eyes (xerostomia and keratoconjunctivitis sicca, respectively). Dendritic cells (DC) are considered to be the most potent antigen-presenting cells. Because of their central role in initiating an immune response while maintaining tolerance, impaired function of these cells might lead to the break of peripheral tolerance and initiation of immune responses to self-antigens. This review will focus on the possible role of DC in SS.
Sjögren's syndrome (SS) is a common systemic autoimmune disease, characterised by progressive mononuclear cell infiltration of exocrine glands, preferentially the salivary and lacrimal glands, and dryness of these glands (xerostomia and keratoconjunctivitis sicca, respectively). The inflammatory infiltrations are predominantly composed of CD4+ T cells, but CD8+ T cells, B cells and plasma cells are also present. In primary SS (pSS), the sicca symptoms occur alone, whereas they are associated with other rheumatic autoimmune diseases in secondary SS, most frequently rheumatoid arthritis (RA) and systemic lupus erythematosus (SLE). In most cases of SS, the production of autoantibodies against Sjögren's syndrome-related antigen A (SSA)/Ro and SSB/La is observed, and less frequently, anti-nuclear antibodies (ANA) and rheumatoid factor (RF) are detected .
The prevalence of pSS is approximately 0.6%, depending on the criteria used for diagnosis, of which 90% are female [1, 2]. The disease manifestations range from an organ-specific autoimmune disorder to a systemic syndrome including musculoskeletal, pulmonary, gastric, haematological, dermatological, renal and nervous system involvement. Moreover, clonal expansion of B cells primarily in the salivary glands is associated with an increased risk for SS patients to develop B-cell lymphoma .
Although it is predominantly CD4+ T cells that mediate the inflammatory infiltrations in the glands, abnormalities in B-cell activation and proliferation have been found to be pivotal during the disease course of SS . In a subset of patients, the lymphocyte infiltrates form ectopic germinal centre (GC)-like structures, aggregating proliferating T and B cells within a follicular dendritic cell (FDC) network and activated endothelial cells [4–6]. Local production of autoantibodies against SSA/Ro and SSB/La in these structures implicates their participation in the pathogenesis of the disease .
B-cell activating factor (BAFF), also called B-lymphocyte stimulator (BLyS), a member of the tumour necrosis factor (TNF) superfamily, is a regulator of B-cell proliferation, maturation and survival . BAFF expression is upregulated in patients with SS and other rheumatic autoimmune diseases [8, 9] and was shown to attenuate apoptosis of B cells from SS patients , suggesting this to be an important factor in the pathogenesis of SS [7, 8]. Most recent research efforts on pSS have therefore focused on B cells and the generation of autoantibodies .
Nevertheless, the purpose of this review is to present aspects on the possible involvement of DC in SS, as they are key players in initiating and sustaining immune responses while maintaining tolerance .
Characteristics of human DC
Dendritic cells comprise heterogeneous cell populations with two main subsets, myeloid DC (mDC), which also include Langerhans cells of the skin, and plasmacytoid DC (pDC) that are phenotypically and functionally different (Table 1). Myeloid DC are considered to be the most potent antigen-presenting cells (APC), playing a central role in the initiation and maintenance of primary immune responses. Plasmacytoid DC are considered to be the main type I interferon-producing cells [13–16]. They circulate in peripheral blood, and upon activation by viral infection, which causes upregulation of CCR7, they migrate to lymph nodes .
Table 1. Summary of the characteristics of the human DC populations.
Antigen presentation, production of various cytokines and chemokines important for T-cell stimulation
CD45R+CD123+CD303+CD304+ TLR7, TLR9
IFN-γ production, TRAIL-dependent lysis of tumour cells
FDC (several subpopulations known)
Accessory cells for B-cell responses (?), presentation of antigen–antibody complexes
Myeloid and pDC originate from bone marrow (BM)-derived progenitor cells, spread via the blood stream and can be found in almost every organ where they act as the sentinels of the immune system . Myeloid DC reside in peripheral tissues in an immature state in which they are able to take up and process antigens. Upon stimulation with inflammatory cytokines, pathogen-derived products, or T cell interactions, they undergo phenotypic and functional changes known as maturation, which enables them to migrate to the afferent lymphoid organs, where they present the antigenic peptides to naive T cells. This may result either in induction or inhibition of antigen-specific immune responses ensuring protective immunity to infectious agents and tumours while preserving self-tolerance [18–20]. Among others, the maturation process includes upregulated surface expression of the lymph node homing receptor CCR7, MHC class II and co-stimulatory molecules, such as CD40, CD80 and CD86 that are indispensable for the function of these cells. The role of pDC in antigen presentation is less clear; however, they can influence mDC via production of immune regulatory cytokines like IFN-α .
Follicular dendritic cell, the stromal cells located in GCs, have been implicated to function as accessory cells for B-cell responses [22–25]. However, recent studies question their essential requirement for B-cell function [26, 27]. Their exact origin remains to be clarified, but it is believed that FDC are cells of non-haematopoietic origin . They do not process and present antigens in the context of MHC class II molecules, but as antigen–antibody complexes on their cell surface via complement receptors (CR), such as CD21 and CD35, or Fc receptors . Remarkably, these complexes are retained for months [22, 29].
Recently, a novel DC subset called IFN-producing natural killer DC (IKDC) with distinct phenotypic and functional properties was described [30, 31]. These cells provide a link between innate and adaptive immunity, but their role in autoimmune diseases like SS still remains to be analysed.
Tolerance and immune regulation by DC
Dendritic cells play a pivotal role in the induction of central as well as peripheral tolerance. Their role in central tolerance is the presentation of self-antigens to T cells and subsequent deletion of the T cells in case of autoreactivity in the thymus [32, 33]. However, the role of DC in peripheral tolerance might be even more critical, as central tolerance was shown to be functional in mouse models of SLE, a systemic autoimmune disease closely related to SS [34, 35]. Immature DC within peripheral tissues constantly sample the surroundings for antigens, i.e. from apoptotic cells. After uptake of self-antigens and in the absence of an additional stimulus, DC remain immature and present self-peptide–MHC complexes to circulating naive T cells in peripheral lymphoid organs. In the case of autoreactivity, this will normally lead to their anergy or deletion, thereby inducing tolerance [20, 32, 33, 36, 37]. Recently, Blander and Medzhitov could show that the phagosomes, intracellular organelles that contain ingested material, determine the fate of their enclosed antigens independently within one DC, controlled by the presence or the absence of Toll-like receptor (TLR) ligands . This is in concordance with the hypothesis that peripheral tolerance is maintained by a balanced cross-talk between TLR and C-type lectin receptors .
Another important function of DC is the induction of regulatory T cells [40–43] that act as suppressor cells and help maintaining tolerance. It could be shown that the maturation state of DC is a crucial point within the system, as only immature DC are able to induce antigen-specific tolerance [43, 44]. It is therefore imaginable that dysregulation of DC results in the presentation of autoantigens on mature DC, leading to stimulation of autoreactive T cells instead of anergy, thereby inducing autoimmunity (Fig. 1). Induction of autoimmune diseases by DC have been shown in model systems [45, 46], indicating a crucial involvement of these cells in the development of autoimmune diseases.
Interestingly, Chen et al. recently reported that blocking DC apoptosis in a mouse model leads to their accumulation . The animals display classical signs of autoimmunity, including ANA production, suggesting that defects in DC apoptosis might be an important aspect of autoimmunity.
Regulatory T cells also play a central role in controlling autoimmunity in animal models [41, 48, 49]. However, Gottenberg et al. could recently show that SS patients have functionally normal regulatory T cells , making a defect of these cells in the pathology of SS rather improbable.
Dendritic cells in SS
So far, not many studies have been carried out concerning alterations in the function of DC in patients with SS. Mainly, the reports focus on FDC, which are present in the chronic inflammatory cell infiltrates in minor salivary glands [51, 52]. Network-like structures of CD35+ cells within mononuclear cell infiltrates are observed in association with T- and B-cell aggregates, proliferating cells and activated endothelial cells, indicating ectopic GC formation in minor salivary glands from patients with SS (Fig. 2; [6–8]). In minor salivary glands without GC, the CD35+ cells formed no network structures but remained scattered [6, 8]. Ectopic GCs have also been characterised by CD21, detecting a higher frequency of networks that are more extensive , possibly explained by the fact that in addition to FDC, CD21 is expressed on B cells. CD21, also known as CR2, C3d receptor or Epstein–Barr virus (EBV) receptor, is expressed as part of the B cell co-receptor complex formed by CD19, CD21 and CD81, which can greatly enhance B cell responsiveness. It is not yet known whether this enhancement is achieved by increasing B-cell signalling, inducing expression of co-stimulatory molecules on B cells or increasing the receptor-mediated uptake of antigen.
Nonetheless, tertiary lymphoid tissue – GC formation with FDC networks – and local autoantibody production in the minor salivary glands of SS patients implicate a possible risk for immune recognition and breakdown of self-tolerance.
Elevated serum IFN-α levels have long been known to be one feature of SLE [54–57], the systemic autoimmune disease closely related to SS. Recently, an activated type I interferon system has been described in salivary glands of SS patients [58, 59], although no elevated IFN-α serum levels are observed. This activated type I interferon system has moved DC more into the focus of SS research, as pDC are the main type I interferon-producing cells [13–16]. Moreover, the presence of pDC in salivary glands of SS patients but not controls  as well as the presence of IFN-α-producing cells in salivary glands of SS patients  suggest that pDC might indeed play an important role in the aetiology of SS. One possible hypothesis about the role of pDC in the aetiopathogenesis of SS is that an initial viral infection triggers pDC to secrete IFN-α via TLR stimulation, and leading to apoptosis and generation of autoantibodies to the apoptotic material . The increase in type I interferon results in further recruitment of pDC, which in turn produce more IFN-α. This persistence of an activated type I interferon system might therefore cause SS. Interestingly, viral infections have indeed been suggested to play a triggering role in the aetiology of SS, although this has never been confirmed [61–64].
On the other hand, TNF-α has also been shown to be a key player in the pathogenesis of autoimmune diseases like RA . Both, in mouse models and SS patients, overexpression of TNF-α in glandular lesions as well as increased levels in plasma have been described [66, 67]. In non-obese diabetic (NOD) mice, a model for SS, anti-TNF treatment resulted in the prevention of SS onset . Pilot studies with human SS patients showed promising results [69, 70]; however, a randomised-controlled trial with infliximab could not confirm the beneficial effect of anti-TNF therapy .
Because of the involvement of IFN-α and TNF-α in autoimmunity, Banchereau et al. have proposed a model with these cytokines as opposing vectors [33, 72]. If the usual equilibrium between type I interferons and TNF-α is disturbed, autoimmunity emerges (Fig. 1). Thus, DC might act as a major controlling factor in the initiation of autoimmune diseases.
Only few studies on mDC of SS patients exist. Ozaki et al. analysed the total number of various DC populations in peripheral blood and salivary glands of SS patients and found that the patients had reduced number of mDC in peripheral blood . Using NOD mice as a model for SS, van Blokland et al. could show that accumulation of DC in submandibular glands precedes inflammatory lymphocyte infiltration , suggesting an involvement of DC in the initiation of SS. Moreover, they could also demonstrate the presence of DC in salivary glands of SS patients .
In accordance, DC have been implicated in the development of SLE . Moreover, the concentration of circulating mDC and pDC is reduced in SLE patients [77, 78], and DC directly isolated from blood as well as monocyte-derived DC show functional impairment [77, 79, 80], all of which implies the importance of further studies concerning DC in autoimmunity.
Involvement of cytokines/chemokines in the pathogenesis of SS
The migration of lymphocytes includes a complex network of chemokines and cytokines. Therefore, dysregulation of this system might be one cause for the pathogenesis of SS. In 2002, Ogawa et al. reported that the T cell attracting chemokines CXCL9 (monokine induced by IFN-γ; Mig) and CXCL10 (IFN-γ inducible 10-kDa protein; IP-10) might be involved in the accumulation of T-cell infiltrates in salivary glands of SS patients [81, 82]. Recently, it was shown that blocking of CXCL10 by an antagonist reduces the lymphocyte infiltration in a murine model for SS .
A study by Szodoray et al. showed distinct circulating cytokine profiles in SS patients and controls . Several other chemokines and cytokines have been shown to be changed, either circulating or in salivary glands of SS patients [83–88], suggesting an involvement of these in the pathogenesis of SS, yet only few reports have concentrated on the role of DC in the cytokine profile in SS [58, 59]. Interestingly, gene-expressing profiles of minor salivary glands from SS patients revealed overexpression of several IFN-α-induced genes, which is in accordance with an activated type I IFN system .
Interestingly, the impaired function of the salivary glands in SS might be associated with a distinct cytokine profile, but is independent of the lymphocyte infiltration, as recently shown in a murine model . However, one has to distinguish between cytokines and chemokines that may be more causative and those that merely mirror the effects of chronic inflammation.
Altered functions of DC might be one main feature of the systemic autoimmune reactions in SS patients, as these cells are the key regulators of the immune response. However, not much is known about potential dysregulation of DC in SS, especially concerning their maturation status, dysfunctional apoptosis, or more general functional properties. An overview of their hypothetical involvement in the pathogenesis of SS is illustrated in Fig. 3. In order to fully understand the systemic defects leading to SS, more efforts have to be devoted towards unravelling the disease mechanisms. Only by precisely knowing the aetiology of SS will it be possible to improve existing and develop new clinical approaches for the treatment of SS.
The authors thank Roland Jonsson for valuable discussions. This work was supported by the Translational Medical Research Consortium, UiB, Norway, and the Strategic Research Program at Helse Bergen.