Spontaneous Germinal Centers and Autoimmunity

Abstract

Germinal centers (GCs) are dynamic microenvironments that form in the secondary lymphoid organs and generate somatically mutated high-affinity antibodies necessary to establish an effective humoral immune response. Tight regulation of GC responses is critical for maintaining self-tolerance. GCs can arise in the absence of purposeful immunization or overt infection (called spontaneous GCs, Spt-GCs). In autoimmune-prone mice and patients with autoimmune disease, aberrant regulation of Spt-GCs is thought to promote the development of somatically mutated pathogenic autoantibodies and the subsequent development of autoimmunity. The mechanisms that control the formation of Spt-GCs and promote systemic autoimmune diseases remain an open question and the focus of ongoing studies. Here, we discuss the most current studies on the role of Spt-GCs in autoimmunity.

Germinal Centers (GCs)

Germinal centers (GCs) are specialized microarchitectures developed in the secondary lymphoid organs to produce high-affinity antibodies against microbial antigens. Naive B cells initiate a GC reaction by responding to an antigen and obtaining help from CD4⁺ T cells at the B cell:T cell border that recognize cognate antigen (Figure 1) (1). B cells that obtain CD4⁺ T cell help seed into the B cell follicle and rapidly proliferate in the presence of Follicular Dendritic Cells (FDCs) to form early GC structures (2). After several days of early GC B cell proliferation, a subset of GC B cells express CXCR4 to migrate toward CXCL12 gradients that are maintained by Reticular Cells (3). The region that contains these rapidly proliferating B cells and Reticular Cells is named the dark zone of the GC, where B cells undergo two major maturation processes of class-switching and somatic hypermutation, as centroblasts. Class-switching is a process by which B cells express a new antibody constant region (IgA, IgE, or IgG) to acquire effector function features that are required to clear a specific type of infection.

Figure 1. Schematic of the Germinal Center (GC) Reaction.

(1.) Naive B cells initiate a response to antigen and obtain CD4⁺ T cell help at the B cell:T cell border. (2.) GC B cells rapidly proliferate as centroblasts and interact with Reticular Cells that secrete CXCL12 within the dark zone of the GC. Centroblasts differentiate into centrocytes and follow a CXCL13 gradient that is produced by follicular dendritic cells (FDCs) in the light zone. (3.) Centrocytes compete for limiting antigen within immune complexes that are trapped on the surface of FDCs. (4). Centrocytes with high-affinity BCRs process and present antigen via MHCII to limiting cognate Tfh. In turn, Tfh provide survival signals for the conjugated GC B cell to undego several different fates. (5). GC B cells can remain within the GC and undergo another round of mutation and class switching or (6.) leave the GC and mature to (7.) long-lived plasma cells or (8.) memory B cells that can persist in the periphery for extended periods of time. (9.) Centrocytes that fail to be selected will undergo apoptosis and will be readily cleared from the GC by tingible body macrophages to prevent the buildup of self-antigen.

During somatic hypermutation, B cells use activation induced cytidine deaminase (AID) to induce point mutations in the variable region of the B cell receptor gene to generate a diverse population of GC B cells from the founder B cell clone (4). Due to the random nature of this process, B cells with lower affinity, higher affinity, and novel autoreactivity are generated, necessitating strict selection of the B cells with optimal antigen affinity to maintain tolerance and select for B cells with improved fitness (5, 6). GC B cells then translocate to the light zone of the GC by using CXCR5 to follow a CXCL13 gradient that is maintained by FDCs to recruit GC B cells (centrocytes) and T cells into the light zone for selection (7-9). Within the light zone, GC B cells compete for limiting antigen complexes trapped on the surface of FDCs. B cells which acquire antigens from the FDCs process and present the antigen to follicular helper T (Tfh) cells located in the light zone of GCs in order to receive survival signals for positive selection. Well-regulated selection of GC B cells is vital for maintaining self-tolerance and several immune cell types are utilized in GC B cell selection.

The primary function of T cells within the GC is to support GC stability by costimulation and cytokine signals. To directly regulate GC B cell activity, Tfh utilize a variety of costimulatory molecules including ICOS, SLAM family molecules, CD40L and CTLA4 (10-13). Tfh also secrete cytokines that support GC B cell activity, including BAFF to promote GC B cell survival (14), IL-17 to control B and T cell migration (15), IL-21, IL-4 and IFNγ to drive GC B cell differentiation (16-20) and osteopontin to dampen BCR-mediated apoptosis of GC B cells (21) (Figure 2). Follicular regulatory T cells (Tfr) are a distinct subset of T cells that repress Tfh differentiation to regulate the size of the GC by a Bcl-6/SAP mediated mechanism (22). Tight regulation of costimulatory signaling is required for maintaining B cell tolerance. In the context of autoimmune disease, dysregulation of T cell populations within the GC also can alter GC B cell activation and differentiation by several different mechanisms to break self-tolerance (23).

Figure 2. Signaling Pathways that modulate B Cell Receptor (BCR) signaling in Spt-GCs.

Several signaling pathways have been shown to either promote or inhibit BCR signaling in the GC. BCR signaling in the GCs is triggered by BCR interaction with antigen complexes that are captured on the surface of FDCs (Fig 1). Lyn and FcgRIIb (shown in red) have been shown to limit BCR signaling and limit Spt-GC formation. However, TLRs (shown in yellow), Btk (in green), WASp (in green) and CD20 (blue) have been shown to promote BCR signaling pathways and enhance Spt-GC formation. Cytokine signals from FDCs (BAFF) and pDCs (type 1 interferon) also promote BCR signaling as well as GC B cell survival.

Spontaneous Germinal Centers (Spt-GCs)

Recently, with more sensitive techniques and methods for detecting GCs, several labs have discovered that GCs can spontaneously develop (named Spt-GCs) in the absence of overt immunization or detectable adventitious infection (18, 24-28). Spt-GCs are detected in both non-autoimmune and autoimmune strains of mice, but the size and frequency of Spt-GCs are augmented in autoimmune-prone mice, leading to the development of class-switched pathogenic antibodies that drive antibody-mediated autoimmune disease (18, 24, 25, 27, 28). Spt-GCs are reported in several autoimmune diseases in both humans and animal models.

The development of GC structures without an overt immune challenge was first observed in autoimmune-prone (NZB×NZW) F1 (NZB/W) mice and anti-CD8 monoclonal antibody-treated C3H/lpr and C3H/gld mice (26, 29). Furthermore, Mackay and colleagues described the development of “germinal centers in the absence of immunization”, in transgenic mice that overexpress BAFF, to promote systemic autoimmune disease (30). In support of the role of Spt-GCs in promoting autoimmunity, the increased GC sizes correlated with an increased number of autoantibody-producing B cells (25). Also, in type 1 diabetes patients, Spt-GC structures form and specifically recognize self-antigen (31).

Lunzina and colleagues further extended the studies of Spt-GCs to rule out the possibility that Spt-GCs form due to asymptomatic infection. In their study of Spt-GCs, several autoimmune-prone mouse lines (NOD, PN, NZB, NZB/W, MRL lpr/lpr, MRL+/+, and B6 lpr/lpr, and male BXSB mice) developed Spt-GCs while the collected blood tested negative for monitored bacterial, viral, or mycoplasma infections (24). Spt-GCs in these mice required CD40-CD40L interaction to form, and were enlarged in aged mice (24). Mice used in this study were also obtained from several different breeding colony locations to account for local endogenous infections (24).

Although overt immunogenic challenge and adventitious infections have been shown not to be causal factors for Spt-GC formation (24), further studies are required to define the role of endogenous viral populations or retroviral elements in promoting Spt-GCs. In support of the role of endogenous viral infections, transgenic expression of an EBV-encoded CD40 mimic in autoimmune-prone mice promotes a break in tolerance through Spt-GC formation and dysregulation of B cell selection (32). Further, Human Endogenous Retroviruses (HERVs) have been proposed as environmental factors that promote a loss of self-tolerance in genetically susceptible patients (33). HERVs integrate into the genome and promote the initial loss of tolerance by stimulating self-reactive B cells through molecular mimicry (33). In SLE, autoantibodies that are targeted to the 70k/U1 snRNP autoantigen, also exhibit cross-reactivity with a p30 gag protein expressed by a specific class of HERVs (33, 34). As many as 50% of SLE patients express p30 gag-reactive Ab titers while only 3.7% of healthy controls express these antibodies (34), suggesting that further studies into the role of HERVs in persistent Spt-GC responses might provide some insights into this correlation.

At the sites of autoimmune inflammation, transient lymphoid structures (called ectopic lymphoid structures, ESLs) can develop in response to inflammatory cytokine signals (35). ESLs also contain GCs called ectopic germinal centers (e-GCs) that may help generate class-switched and somatically mutated B cell populations at the site of inflammation (35). In autoimmune diseases, ESLs and e-GCs develop in the absence of overt infection to promote chronic relapsing inflammation (35-39). In rheumatoid arthritis (RA) patients, autoantibodies to several self-antigens are observed in correlation with ESLs and AID-expressing e-GCs in inflamed synovial tissue (37, 40). Correspondingly, e-GCs that form in autoimmune-prone mice are phenotypically similar to Spt-GCs in regard to induction, regulation and activity. Overall, genetic susceptibility to autoimmunity is thought to promote the loss of tolerance through Spt-GCs by driving the generation of antibodies with high-affinity to self-antigens. Several studies have implicated the roles of innate sensing, BCR signaling and costimulatory molecules in promoting Spt-GC formation. In humans, these molecules and various downstream signaling components are altered due to genetic mutations, establishing susceptibility that leads to the loss of self-tolerance.

Role of Spt-GCs in several autoimmune diseases

Systemic Lupus Erythematosus (SLE)

Systemic Lupus Erythematosus (SLE) is a progressive and multifaceted autoimmune disease that is characterized by the production of self-reactive antibodies that target nuclear antigens (called anti-nuclear antibodies or ANAs). ANAs are frequently class-switched and somatically mutated, suggesting that they are most likely derived from GCs (25, 41-44). Using tonsil tissues, Cappione and colleagues have shown that negative selection of self-reactive B cell populations in the GC is defective, allowing for self-reactive B cells to survive in the GC (45). In addition, pediatric patients with SLE exhibit increased pre-GC B cells in circulation as compared to healthy controls and e-GC formation has been reported in the thymic tissue of human SLE patients (46, 47). Spt-GCs are observed in several different SLE mouse models, which all develop enlarged Spt-GC structures.

Rheumatoid Arthritis (RA)

Rheumatoid Factor (RF) and anti-citrullinated protein antibodies associated with Rheumatoid Arthritis (RA) are detected in the sera of 50-70% RA patients and class switched autoantibodies targeted against chaperone proteins, nuclear antigens, enzymes, and components of cartilage are also found in the joint tissue (48, 49). Initial report indicated the production of class-switched and high-affinity RF autoantibodies within the synovial tissue of the inflamed joint in humans, suggesting a potential role of e-GC formation at the site of inflammation in this process (50, 51). Later studies by Weyand and Goronzy confirmed the formation of e-GC structures in the synovial tissues of 24% of RA patients, and additional 20% of the RA patients formed B cell:T cell aggregate structures that lacked FDCs (52).In mouse models of RA, several studies have reported both Spt-GC and e-GC formation that contribute to disease progression. Using the KBxN model that expresses a self-antigen-targeted TCR, two separate labs have reported the presence of Spt-GC structures that form within the spleen of these mice (53, 54).

Multiple Sclerosis (MS)

Multiple Sclerosis (MS) is an autoimmune demyelinating disease that specifically targets the central nervous system to cause progressive paralysis. To recapitulate MS in animal models, most animals require some form of immunization with a self-peptide or treatment with a chemical stimulus to develop experimental autoimmune encephalomyelitis (EAE) (55). This EAE model may not recapitulate the spontaneous nature of Spt-GC formation, making studies of the role of Spt-GCs in EAE challenging. However, some groups have characterized Spt-GC formation in animals by analyzing GC structures after the primary B cell response in EAE mice has ended or by using specialized mouse models with mutant B cell receptors (56). Using a mouse model in which B cells and T cells express receptors that are specific for myelin oligodendrocyte glycoprotein, Dang and colleagues found neither e-GCs in the brain tissue nor Spt-GCs in the secondary lymphoid organs in these BCR knock-in mice compared to wild type control mice without the knock-in gene (56). However, a specialized subset of partially activated B cells that are primed to present antigen were found within inflammatory sites (56).

Autoimmune lymphoproliferative syndrome (ALPS)

Autoimmune lymphoproliferative syndrome (ALPS) is most frequently caused by mutations in the Fas (CD9, Apo-1) gene, which is required for regulation of lymphocyte apoptosis (57, 58). Lymphocyte death via Fas-mediated apoptosis is crucial for maintaining tolerance in the GC. Mouse models with Fas or FasL deficiency develop lupus-like autoimmunity characterized by autoantibody production, lymphadenopathy, splenomegaly and enlarged Spt-GC formation in vivo (24, 59, 60). In addition, overexpression of miR146a, a regulator of the Fas pathway in GC B cells, leads to the development of ALPS-like disease which is associated with enlarged Spt-GC formation, increased autoantibody titers, and evidence of systemic pathological inflammation (61).

Sjogren's Syndrome (SS)

In Sjogren's syndrome (SS), a disease that is characterized by autoimmune inflammation of the salivary and lacrimal glands, the contributions of e-GCs are well characterized (62-65). These GC structures express AID, are regulated by FDC networks, contain BAFF-producing Tfh and produce SS-associated autoantibodies (65-67). Furthermore, in a mouse model of SS, e-GC formation in the salivary gland is also correlated with the development of Spt-GCs in the spleen (68). Current studies regarding the role of e-GCs in SS patients have focused on targeting the chemokines (CXCL1, CXCL13, and CCL21) that regulate the formation of e-GC structures for therapeutic intervention since high expression of these chemokines in patients is associated with progressive disease symptoms (69, 70).

Type 1 Diabetes (T1D)

In autoimmune diabetic disease (Type 1 diabetes, T1D), FDC-containing and AID-expressing e-GCs are formed within the pancreatic islet tissues and contribute to insulitis in non-obese diabetic (NOD) mice (71, 72). Spt-GC formation was first reported by Luzina and colleagues in NOD mouse spleen, but the connection of these two structures to autoimmune disease development was implied (24). Later studies by Wan and colleagues using an insulin-specific B cell model in NOD mice, demonstrated that insulin-binding B cells can form Spt-GCs in multiple lymphoid organs including the draining lymph node, spleen, and non-draining lymph nodes in NOD mice (73). Finally, Spt-GC structures with self-antigen reactive B cells were also observed in the spleens and lymph nodes of human T1D patients (31).

Factors that contribute to Spt-GC formation

The formation and regulation of Spt-GCs is complex. Multiple factors are known to contribute to Spt-GC formation and GC-associated autoimmunity in humans and in mice. These include factors that modulate BCR signaling, alter B:T costimulation/selection, block inhibitory/regulatory receptor function, affect complement-mediated regulation of selection, alter cytokine/chemokine signals and promote self-ligand sensing by pattern recognition receptors.

The Role of B Cell Receptor (BCR) Signaling in Spt-GC formation

A causative link between self-reactive B cell populations and autoimmunity is well established in mouse models and human patients with autoimmune disease (74). Self-reactive B cells are also detected in healthy individuals, but the loss of self-tolerance in autoimmune patients is governed by hyperactive BCR signaling as a result of mutations in BCR signaling molecules (75, 76) (Figure 3). Using a BCR deficient mouse line that expressed the Epstein Barr Virus LMP2A protein, Casola and colleagues reported that BCR signaling strength is a critical contributor to the formation of Spt-GCs (77). LMP2A influences intracellular BCR signaling in the absence of the receptor to control for the effects of antigen specificity. PNA⁺GL7⁺CD95⁺ GC structures in LMP2A expressing mice formed within the Peyer's patches and underwent somatic hypermutation, class switching, relied on T cells for formation and retained a population of FDCs. We and others have shown that BCR specificity for self-antigen is important for the loss of self-tolerance and autoimmunity (25, 78-80).

Figure 3. Signaling Pathways that regulate GC B cell and Tfh selection in Spt-GCs.

After acquiring antigen, B cells present antigen via MHC II to T cells with cognate T cell receptors (TCRs). Several co-stimulatory molecules are involved in B cell activation that leads to GC formation or in the maintenance of GC B cells. Most of the receptors and co-stimulatory molecules that are involved in Spt-GC formation regulate B cell survival, proliferation, and class switching (CD40, CD80/86, IL4R, IL21R, IFNγR, BAFFR, TLRs), but IL-17R controls the migration of Tfh and GC B cells and SLAM family molecules have co-stimulatory functions.

Bruton's Tyrosine Kinase

Bruton's Tyrosine Kinase (Btk) functions through PLC gamma and IKK to regulate B cell development and function (81, 82). Patients that exhibit defects in Btk function develop X-linked agammaglobulinemia (XLA), a disease that is characterized by susceptibility to recurrent bacterial infections due to severely reduced Immunoglobulin formation (83). Mouse models of XID exhibit mutations in Btk that recapitulate the B cell intrinsic phenotypes of the disease (76). Overexpression of Btk using a CD19 promoter increases B cell activation (overexpression of CD80, CD86 and MHCII, and increased calcium flux), B cell susceptibility to Fas-mediated apoptosis, and NF-κB activation (84). Restricting Btk overexpression to B cells is associated with increased Spt-GC and plasma cell formation in the spleen, and this Spt-GC formation requires CD40-CD40L interaction (84, 85). GC structures also fail to form in the lymph nodes of XLA patients (83). In addition, Btk expression in B cells promotes the expansion of Tfh populations in the GC to further support the GC structure (85).

WASp

The Wiskott-Aldrich syndrome protein (WASp) was first identified in association with Wiskott-Aldrich syndrome, an X-linked primary immunodeficiency that is characterized by several autoimmune manifestations including autoimmune cytopenia, arthritis, vasculitis, inflammatory bowel disease and nephropathy (86, 87). Interestingly, WAS patients also exhibit symptoms of primary immunodeficiency including susceptibility to recurrent infections, treatment-resistant eczema and lymphoid malignancies (86). WASp is required for actin polymerization and cytoskeletal organization that allows for lymphocyte migration, immunological synapse formation, regulation of surface receptor expression and intracellular signaling (86, 88). Organism-wide WASp deficiency is associated with higher titers of circulating IgA and IgE, suggesting a B cell intrinsic effect (88).

Several studies have focused on the role of WASp in B cells using different model systems. Overall, WASp in B cells is important for actin cytoskeleton rearrangement, Ca²⁺ flux after BCR stimulation, B cell development, GC B cell selection and differentiation (88, 89).A reduced number of marginal zone B cells and a proportionally greater number of GC B cells with increased IgM serum antibody titers were observed in mice in which B cells were deficient in WASp (88). Using a bone marrow chimera system, Becker-Herman and colleagues observed increased high-affinity dsDNA-specific IgG titers and enlarged Spt-GC structures when WASp was absent in B cells, suggesting that WASp is a critical negative regulator in B cells that prevents the development of autoimmune disease (75). WASp deficiency in B cells is associated with increased glomerulonephritis and decreased survival of the animals (75). Correspondingly, lentiviral expression of WASp restored B cell development in the spleen and reduced the development of dsDNA-specific IgG_2c. (90).

The role of costimulatory molecules in Spt-GC formation

CD40-CD40L

Interaction of CD40 on B cells with the cognate ligand CD40L on Tfh cells is essential for GC differentiation and function (91, 92). Specifically, CD40L-CD40 interaction rescues GC B cells from Fas-mediated apoptosis (91-94), induces GC B cell proliferation (95), initiates immunoglobulin class switching (96), and promotes antibody secretion (97, 98). Both mice and humans that lack functional CD40 or CD40L are incapable of forming antigen induced or Spt-GC structures (96, 99-103). Furthermore, treatment of mice with a CD40L-blocking antibody inhibits GC formation in response to T-dependent antigen immunization and promotes B cell maturation through GC-independent pathways (104, 105). Lunzina and colleagues also confirmed that CD40-CD40L interactions are critical for Spt-GC formation, as anti-CD40L treatment inhibits Spt-GC formation and resultant autoimmunity in mice (24).

CD80 and CD86

CD80/B7-1 and CD86/B7-2 are co-stimulatory molecules that are expressed on the surface of activated antigen presenting cells, including B cells, to interact with CD28 or CTLA-4 on the surface of CD4+ T cells (106, 107). CD80 and CD86 are rapidly upregulated on B cells upon antigen interaction or inflammatory signals to enhance T cell activation and to prevent target T cell anergy (106, 107). In B cells, activation of CD80/CD86 in conjunction with other stimuli induces class switching, B cell proliferation and survival.

Blockade of both CD80 and CD86 in mice by genetic knockout (CD80/CD86^-/- mice) or by anti-CD80/86 treatment inhibits GC formation and IgG class switching in both T-dependent antigen-challenged mice and unimmunized control mice (108, 109). Furthermore, CD80/86 deficiency blocked Spt-GC formation in mice that exhibit T cell hyperactivation as a result of a Lat^Y136F mutation (109). Correspondingly, CD28^-/- mice also failed to develop Spt-GC structures and exhibit altered class switching (110). With CTLA4 blockade, Tfh differentiation is increased, leading to enlarged Spt-GCs, autoantibody production and autoimmunity development (12). Later studies in B cell intrinsic CD80/CD86^-/- mice show formation of GCs with impaired IgG2a responses after influenza challenge (111). Using an HgCl₂-induced model of EAE, dual blockade of both CD80 and CD86 significantly reduced GC formation and EAE in Brown Norway Rats (112). Ab-mediated therapeutic blockade of CD80/86 has also protected mice in Ag-induced EAE (113). Blockade of either activation molecule individually was not sufficient to inhibit the loss of self-tolerance (112).

SLAM Family Molecules

Signaling lymphocyte activation molecule (SLAM) family proteins that regulate B cell:T cell interactions, are crucial for maintaining tolerance (25, 114-116). Deletions or point mutations of several Slamf genes have been associated with the development of autoimmune disease in mice and in humans (115-122). In human family-association studies, several single nucleotide polymorphisms (SNPs) in the promoter and coding regions of SLAMF7 and LY9 have been associated with SLE-disease susceptibility (122). Furthermore, defective engagement with SLAMF6 in human T cells is associated with autoimmune disease progression (123).

B6.Sle1b congenic mice, a mouse model for SLE in which the Slamf locus is derived from the lupus-prone NZM2410/NZW strain, develop enlarged Spt-GC structures and accumulate class-switched anti-nuclear antibodies that promote SLE-like autoimmunity (25, 114-116). Enlarged Spt-GC response in B6.Sle1b mice can be reversed and GC tolerance can be restored in B6.Sle1b mice with a B cell-intrinsic overexpression of isoforms of CD84/Slamf5 and Ly108/Slamf6 that are derived from non-autoimmune B6 mice (115), indicating the important roles played by CD84 and Ly108 in maintaining GC tolerance.

The Ly108/slamf6 gene is expressed in three separate isoforms, Ly108.1, Ly108.2 and Ly108-H1, which differentially regulate peripheral B cell and T cell tolerance. B cells that express the Ly108.1 isoform exhibit significantly reduced calcium flux upon BCR engagement and decreased cell death, causing the reduced efficiency of negative selection as compared to the autoimmune resistant isoform (Ly108.2) (116). In addition, Ly108 deficiency in autoimmune-prone mice causes reduced autoAb titers and reduced percentages of effector memory T cells that result in reduced autoimmune disease (117). Finally, overexpression of the Ly108-H1 isoform in T cells or B cells of B6.Sle1b mice reduces autoAb titers, T cell activation, and Spt-GC formation to protect against SLE disease progression (117). These observations reinforce the importance of the SLAM family molecules in regulation of both T cell and B cell-mediated self-tolerance.

The role of Toll-like receptors and Interferon signaling in Spt-GC formation

Toll-Like Receptors

Toll-Like Receptors (TLRs) are a class of pattern recognition receptors (PRRs) that were initially characterized as sensors of molecular patterns found in lipids, proteins, and nucleic acids derived from pathogens. A series of publications utilizing in vitro stimulation assays were the first to demonstrate a functional role for TLRs in self-antigen mediated B cell activation. Specifically, BCR-mediated internalization of self-RNA and self-DNA containing immune complexes (ICs) activated autoimmune rheumatoid factor (RF) transgenic B cells (124-127). Importantly, activation was dependent on synergistic BCR and TLR9 or TLR7 stimulation upon exposure to DNA and RNA-associated autoantigens, respectively. These findings suggested that weaker BCR stimulation in autoimmune B cells confers the need for synergistic TLR signaling to reach an activation threshold that can typically be achieved through the BCR alone when challenged with foreign antigen. Accordingly, a later study documented suboptimal BCR signaling in these B cells and the requirement for BCR and TLR9 synergy to activate a different set of genes than results from the activation of either receptor alone (128).

Since these in vitro studies suggested a potential role for TLRs 7 and 9 in the development of autoimmunity, multiple groups proceeded to study the impact of these receptors in autoimmune mice with a focus on how they control autoantibody production, autoimmune pathology, and cellular activation. Studies in TLR7 and TLR9-deficient MRL lpr/lpr mice demonstrated that TLRs 7 and 9 have opposing roles in SLE development (129, 130). The absence of TLR7 abrogated the production of RNA-specific autoantibodies, impaired pathogenic IgG_2a class-switching, and significantly reduced downstream kidney pathology, as a result of reduced immune activation (129). Conversely, the absence of TLR9 enhanced kidney pathology, immune activation, and class-switching (130, 131). Conflicting data in studies on anti-DNA antibody production, IgG₂ class-switching, and downstream kidney pathology in several TLR9-deficient autoimmune models including FcγRIIb ^-/-, Sle1b, WASp, B6.lpr/lpr, pristane-induced lupus, and Nba2 has emerged, indicating the complex role of TLR9 in these processes in different contexts and models (27, 129-136).

Following the generation of TLR7-deficient MRL/lpr mice, a model possessing an extra copy of TLR7 as a result of the translocation of TLR7 from the X to the Y chromosome, was characterized (137, 138). This model, termed Y autoimmune accelerator (Yaa) does not induce overt autoimmunity on a B6 background, but significantly accelerates disease on an autoimmune background (137, 138). Because the Yaa locus contains about 16 genes, including TLR7, further study strived to verify that TLR7 is the major driver of the Yaa phenotype (139, 140). Loss of TLR7 overexpression on both the FcγRIIb^-/- and Sle1 autoimmune backgrounds abrogated almost all facets of the Yaa phenotype (139, 140). The generation of transgenic mice overexpressing TLR7 phenocopied Yaa mice, further confirming the predominant role of TLR7 in driving disease (127, 139, 141). These studies demonstrate that TLR7 is a critical driver of disease, whereas TLR9 has a more complex role whereby it appears to have both promoting and regulatory functions in DNA-specific autoantibody production, immune activation, and resultant downstream pathology.

While it was clear that TLR7 is capable of driving disease, the contribution of TLR7 in specific cell-types and the mechanism by which TLR9 regulates disease development remained unclear. Molecular studies had revealed that TLR9 has a greater affinity than TLR7 for the endosomal TLR transport protein, Unc93b1, providing a potential mechanism by which TLR9 can restrain TLR7-induced activation in autoimmune mice (142). In 2010, studies in two different autoimmune models emerged that indicated TLR7-deficiency in TLR9-deficient mice abrogated accelerated pathology observed in these mice (135, 143). Additionally, several experiments suggested that TLR7 and TLR9 may function in a B cell-intrinsic manner to control autoimmune responses (135, 137). In 2012, a comprehensive study emerged that showed a B cell-intrinsic role for TLR7 in driving B and T cell activation and the production of autoantibodies in Sle1.Yaa mice, through eliminating the extra copy of TLR7 specifically in B cells (141). Later in vivo and in vitro studies utilizing self-DNA containing ICs demonstrated that B cell-intrinsic TLR9 control of TLR7 is responsible for inhibiting the development of autoimmunity and B cell activation, establishing the predominant function of TLR7 and TLR9 in controlling B cell responsiveness in autoimmunity (144, 145).

Recent studies have focused on the B cell-intrinsic role of TLRs 7 and 9 in mediating lupus-like autoimmune responses. A potential role in GC responses was apparent from an earlier study that showed increased Tfh marker expression on CD4 T cells in autoimmune Yaa mice (138). Accordingly, the impact of TLR7- and TLR9-deficiency in non-autoimmune mice and two different autoimmune mouse models, Sle1b and WASp, was comprehensively assessed (27, 136, 141). In addition to impacts on cellular activation, TLR7 was found to promote Spt-GC responses on both non-autoimmune and autoimmune backgrounds in a B cell-intrinsic manner (27, 136). Interestingly, in the same studies, TLR9 was found to negatively regulate Spt-GC responses in a B cell-intrinsic manner (27, 136). These studies suggest that the GC represents a major site whereby B cell-intrinsic TLR signaling drives SLE.

Type I IFN (T1IFN)

The type 1 interferon (T1IFN) family is a large group of diverse signaling molecules (IFN-α, -β, -ε, -κ and -ω) that all bind to the same primary signaling receptor (IFNαR1). Despite the identical primary receptor binding specificity, each member of the T1IFN family elicits distinct signaling cascades as a result of co-receptor availability (IFNαR_2a, IFNαR_2b, or IFNαR_2c) (146) and binding affinity for the receptor complex (147). Type I IFN is heavily associated with the onset and progression of human SLE, evidenced by a characteristic interferon signature (148). In human SLE, a T1IFN signature is closely correlated with autoantibody seropositivity and elevated levels of T1IFN can be observed in patient sera 2 years prior to SLE diagnosis (149-151). pDCs are believed to be the major producers of type I IFN, triggered by the sensing of RNA and DNA-containing immune complexes which stimulate endosomal TLRs (152). Depletion of pDCs in the mouse models of SLE prevents the development of Spt-GCs and autoimmunity, indicating the contribution of pDC function to Spt-GC formation and subsequent disease manifestations (152, 153).

In addition to pDCs outside the GC microenvironment, FDCs within the GC have also been proposed as producers of Type I interferon through a TLR-dependent mechanism, albeit not experimentally proven (154). In B cells, T1IFN treatment induces heightened response to BCR stimulation, proliferation and resistance to Fas-mediated apoptosis in vitro (155-158). Conversely, B cells deficient in IFNαR₁ signaling exhibit defective class switching, reduced expression of activation markers and MHCII (148, 156, 159, 160). Spt-GC responses are reduced in BXD2 autoimmune mice deficient in IFNαR₁(161). Surprisingly, the Rawlings group finds no major effects of IFNαR deficiency on autoimmune GC response using the Wiskot-Aldrich syndrome (WAS) chimera model of B cell-driven autoimmunity (20), suggesting that the function of T1 IFN in Spt-GC regulation can vary between autoimmune models. These conflicting results in animal models closely reflect similar challenges that are reported in human clinical trials showing differential outcomes among SLE patients for T1IFN-blocking therapies (162). Further studies are warranted to delineate the factors that determine the role of T1IFN in Spt-GC dysregulation.

Type II IFN

Like Type I interferon, the role of Type II interferon in autoimmune disease manifestation is well documented (163). In both human patients and mouse model of SLE, the concentration of circulating IFNγ can be an indicator of SLE disease onset and severity (150, 163-165). The role of IFNγ, however, is not consistent among all autoimmune diseases. In MS, blockade of IFNγ signaling in humans worsens MS disease onset (166), but blockade of IFNγ signaling has been shown to reduce autoantibody production and resultant autoimmunity in multiple mouse models of SLE (18, 20, 167-172).

Decrease in mRNA decay of ifng transcripts in T cells is shown to be associated with follicular T helper (Tfh) cell accumulation, increased Spt-GC formation and SLE-like autoimmunity in Roquin^san/san mice (19). Also, IFNγR1 deficiency in B cells abolishes Spt-GC formation and associated autoantibody production in two distinct mouse models of SLE (18, 20). IFNγR-mediated Spt-GC development requires STAT1 signaling (18). Jackson and colleagues confirmed the role of the JAK-STAT pathway by using JAK inhibitors to prevent in vitro Spt-GC differentiation in human B cells (20). Interestingly, IFNγR₁-deficient mice are able to form GCs upon immunization with a foreign antigen, indicating that the role of IFNγR₁ is limited to the Spt-GC response (18).

Cytokine/Chemokine Signaling in Spt-GCs

B-cell activating factor (BAFF)

The B-cell activating factor (BAFF) plays a critical role in primary B cell development (30, 173). BAFF is involved in the survival of plasmablasts (174) and GC B cells (14). BAFF also increases proliferation upon BCR crosslinking (175). In addition, treatment of mice with a BAFF decoy receptor ablated the formation of splenic GCs and IgG₁ production in mice after immunization, suggesting that BAFF is crucial for the maintenance of GC response (176, 177). In support of these findings, mice with reduced BAFF-receptor signaling formed GCs but failed to sustain these GCs upon immunization (178, 179). In SLE-prone mice, BAFF-blocking therapy did not ablate Spt-GC formation, but reduced the number of autoantibody producing cells (180). Interestingly, BAFF overexpression promotes enlarged GC formation in the Peyer's patches of the gut in the absence of immunization (30).

IL-21

The role of IL-21R signaling in initiating the GC response is well established (16, 181, 182). IL-21 is produced by both Th17 and Tfh cells and supports GC B cell survival and proliferation (15, 16, 183, 184). In B cells, IL-21R signaling activates STAT3 to induce the expression of Bcl-6, a master regulator of the GC response (182, 185). Mice deficient in IL-21, IL-21R, and B cell conditional deletion of STAT3 all exhibit hindered class-switched antibody responses to T-dependent immunization, suggesting a defect in induced GC formation (16, 181, 182, 185). However, treatment of IL-21R^-/- mice with RNA-loaded VLPs can drive GC and VLP-specific Ab responses, suggesting that IL-21R signaling can regulate BCR, but not RNA-sensing innate signaling (181).

In autoimmune prone MRL.lpr/lpr and BXD2 mice, IL-21R deficiency causes an absence of Spt-GC formation and inhibits the autoantibody- associated cutaneous SLE disease, kidney nephritis and plasma cell formation (186, 187). Our unpublished data also show absolute requirement of IL21R signaling in Spt-GC formation in B6.Sle1b mouse model of SLE (P.P.D and Z.S.M.R). In SLE-prone BXSB.Yaa mice, IL-21 expression is increased and promotes IgG₁, IgG_2b and IgG₃ class switching (188), suggesting the role of IL21 signaling in Spt-GC formation. In humans, several polymorphisms in the IL21R locus are associated with the onset of MS and SLE (189-191), and elevated concentrations of IL-21 in the sera of RA patients correlated with autoantibody production and disease severity scores (192).

Chemokines

The migration of B cells through GC structures is crucial for the separation of rapidly proliferating GC B cells that are undergoing class switching and somatic hypermutation in the dark zone from GC B cells that are competing for survival signals during selection in the light zone. In GCs, this separation is largely mediated by chemokines that regulate the movement of GC B cells in response to different chemokine stimuli (193).

CXC and CCR chemokine receptors (CXCR4, CXCR5, CCR6, CCR7, CCR9, CCR10, S1PR1, S1PR3, CNR2, and GPR183) are G-protein-coupled receptors (GPCRs) that utilize regulators of G protein signaling (RGS) proteins to modulate the magnitude of the intracellular signaling cascade (193). Several RGS proteins are expressed in B cells and modulate their migratory responses (15, 194-198). Blocking all RGS function via mutation in Gai2, a guanine exchange subunit that interacts with RGSs, causes the accumulation of poorly organized splenic structures that resemble Spt-GCs (195). Rgs1^-/- mice exhibit enhanced B cell migration in response to CXCL12 and CXCL13, which results in enhanced Spt-GC, as well as immunized GC, responses (197). Also, Rgs13^-/- mice exhibited greater formation of Peyer's patches and unrestricted enlargement of GC formation after immunization (194). Other studies of RGSs reported increased expression of Rgs2, Rgs9, and Rgs10, and decreased expression of Rgs14, Rgs18 and Rgs19 proteins in GCs (199), but the function of the entire repertoire of RGSs in the maintenance of GC is not yet fully defined. As evidenced by previous studies, multiple RGSs are likely to play a crucial role in regulating Spt-GC formation and in modulating autoimmune disease development. Future studies will likely define the role of this class of proteins in greater detail.

IL-17

IL-17R signaling has been shown to control the migratory capacity of both GC B cells and Tfh cells within the GC (200, 201). The production of IL-17 was initially observed in CD4⁺ T cells within the GC (15), and later studies by the Mountz lab further defined these IL-17 producing cells as Tfh (200). These IL-17 producing Tfh did not produce IL-21, indicating that two distinct subsets of Tfh contribute to the regulation of the GC (200). In Tfh, IL17R signaling upregulates the expression of Rgs16 to block chemotaxis of Tfh out of the GC microenvironment (200). In GC B cells, a similar pathway utilizes Rgs16 and Rgs13 to block GC B cell migration to the dark zone in response to CXCL12 (15, 201). Overexpression of IL-17 via adenoviral transduction increases GC size and T:B cell interactions whereas antibody-mediated blockade of IL-17Ra signaling inhibits Spt-GC formation (15). Furthermore, deficiency of RORγT, which is classically thought to be a master regulator of Th17 differentiation had pronounced effects on the ratio of Tfh to Tfr in GCs, and in turn, on Spt-GC formation (210). Mice deficient in RORgT have increased serum levels of BAFF, IL-21 and RORγT-independent production of IL-17 by Tfh (202). The role of IL-17 in Spt-GC formation and regulation appears to be context dependent, as Spt-GC formation in TLR7 overexpression model of SLE does not require IL-17 (10), while it contributes to Spt-GC formation and autoimmune responses in SLE-prone BXD2 mice (200).

Inhibitory Receptor Signaling and Apoptotic Cell Clearance in GC Regulation

The currently accepted model for the initiation of autoimmunity involves synergy between genetic susceptibility and a triggering event that induces the accumulation of apoptotic cells that activate autoreactive B and T cells (203-205). Tingible body macrophages located within the GC are primarily responsible for the clearance of GC B cells undergoing apoptosis as a consequence of negative selection. Subsequent secondary necrosis due to inefficient clearance results in immunogenic rather than tolerogenic responses by activating multiple cell types in the GC (203).

Apoptotic cell clearance in the GC

Tingible body macrophages clear apoptotic cells within GCs by Mer Tyrosine Kinase (MerTK) and Mfge8-integrin-mediated clearance mechanisms (205-207). MerTK dually functions to clear apoptotic cells and signal through a receptor tyrosine kinase domain to induce the production of anti-inflammatory cytokines and dampen TLR and inflammatory cytokine responses (208). We have shown that Mer-deficient mice exhibit enhanced GC responses and B cell and T cell activation that promotes the development of autoimmunity (205, 206). Hanayama et al. first showed that Mfge8-deficient mice exhibit deficient apoptotic cell clearance in GCs, leading to enhanced Spt-GC responses in aged mice (207). Kranich et al. later showed FDCs, rather than tingible body macrophages, are the major producers of Mfge8, which is trapped on the surface by tingible body macrophages (209). In human SLE patients, genetic polymorphisms are present in MerTK adaptor molecules, Gas6 and protein S, and in Mfge8 (210). Further, apoptotic cell accumulation has been observed in GCs within lymph nodes derived from SLE patients and phagocytic cells derived from lupus patients exhibit impaired clearance of apoptotic cells and their debris (211, 212).

Fc Receptors

Fc receptors bind to the Fc portion of antibodies that are complexed with antigens, cell debris and/or opsonized pathogens to enact activating or inhibitory signals within a recipient immune cell (213). Although several Fc receptors are found on the surface of immune cells in humans and mice, most of these receptors are found on the surface of myeloid cells and few are found on the surface of lymphocytes (213, 214). As these receptors can be both activating and inhibitory, several studies have linked human susceptibility to autoimmune disease with dysregulated activity of Fcg receptors (28, 213-217). Correspondingly, several FcγRs have been the focus of novel targeted therapeutics for the treatment of autoimmune diseases (214).

FcγRIIb is a low-affinity inhibitory Fc receptor that is found on the surface of B cells that negatively regulates signaling through the B cell receptor (28, 218). In humans, two different mutations in the transmembrane region of FcgRIIb are correlated with susceptibility to SLE (217, 219-224). The FcγRIIb-232T mutation exhibits high disease association among European-origin SLE patients (225), with a higher prevalence of expression among African, African-American, and southeast Asian populations with SLE (5-11% of patients) (217, 219-221, 223, 224, 226). FcγRIIb-232T-expressing human B cells exhibit defective inhibitory functions, allowing for microclustering of BCRs and resultant cellular activation, causing hyperactivity in B cells (218). In FcγRIIb^-/- mice, Spt-GC and Tfh responses are increased, causing greater SLE-associated disease pathology (increased ANA titers and kidney pathology) (28). These elevated Spt-GC, Tfh and severe autoimmune responses in FcγRIIb^-/- mice have recently been shown to be mediated by FcγRIIb deficiency and lupus-associated SLAM family genes derived from the 129 strain (28).

Lyn

Lyn is a Src family tyrosine kinase that can both activate and inhibit BCR signaling cascades in B cells (227-229). The activating functions of Lyn can be compensated by other Src family kinases within the same signaling cascade (227), whereas Lyn plays non redundant role in mediating signaling in the inhibitory feedback pathways, including recruitment of FcγRIIb signaling in B cells (228, 230). Thus, B cells that lack Lyn exhibit BCR signaling hyperactivity leading to increased proliferative responses, activation marker expression, and capacity for antigen presentation (229, 230). Lyn deficiency in B cells also promotes anti-nuclear antibody formation, antibody-mediated kidney pathology and elevated levels of pro-inflammatory cytokines that promote SLE-like disease (229, 231). The autoimmune phenotypes including Spt-GC responses in Lyn^-/- mice are mediated by MyD88 signaling in B cells (229, 231, 232). Interestingly, conditional deletion of Lyn in B cells under the control of CD79^Cre resulted in a significant reduction in Spt-GC formation, rather than expansion (232). Even though bone marrow B cell development is not affected by conditional deletion of Lyn, these mice have significantly reduced number of mature B cells, suggesting that fewer cells are available to enter the GC reactions (229). Further studies using Cre mice (i.e., CD23^Cre) that will delete Lyn at the mature B cell stage may elucidate stage-specific role of Lyn in Spt-GC development.

Complement Receptors and Proteins

Complement proteins and receptors are an integral component of the selection process of B cells within the GC. Complement receptors (CRs) on the surface of FDCs aid in trapping immune complexes on the surface of FDCs in limiting quantities to promote competition-mediated selection of B cells with high affinity BCRs. SLE patients carry mutations and SLE-associated SNPs in both Complement Receptor 1 (CR1) and Complement Receptor 2 (CR2) genes that contribute to disease (233-235). Deficiencies of CR1 and CR2 on FDCs cause deficiencies in GC maintenance, indicated by reduced efficiency of the IgG response to immunization (236).

CR2 can bind to many ligands including complement proteins (including C4 and C1q), CD23, EBV surface proteins, and T1IFN (237-243). The interaction of CR2 with these ligands has been shown to affect B cell development in the GC by regulating class switching, apoptosis, and B:T cell contact to regulate plasma cell formation (243-245). In mouse models of SLE, CR2 deficiency in B6.lpr mice contributes to increased autoantibody formation and renal disease, however MRL/Lpr mice with CR2 deficiency expressed increased autoantibodies without renal disease (246-248).

Over 75% of SLE patients are deficient in C1q and C4 proteins (249). Also, C1q and C4 knockout mice produce autoantibodies and develop autoimmune kidney disease (154, 246, 250, 251). In 564Igi mice, which contain a BCR knock-in that targets nuclear antigens, C4 deficiency results in an enlarged Spt-GC phenotype without antinuclear IgG accumulation (79, 252). However, in MRL.lpr/lpr mice, C4 deficiency exacerbates IgG autoantibody production and autoimmune disease (246). Thus, the role of complement proteins in promoting Spt-GC formation and autoimmune disease is dependent on specific genetic factors that have not yet been elucidated.

Conclusions

The role of Spt-GCs in several different autoimmune diseases is well established in the literature by multiple research groups. Intriguingly, several studies have identified factors and signaling pathways that are vital for Spt-GC responses, but are not required for the GC responses induced (iGC) by foreign antigens, suggesting that distinct signaling mechanisms may control autoimmune Spt-GC and induced GC responses. Delineating distinct cell-intrinsic pathways that promote Spt-GC versus iGC responses could uncover signaling molecules that can inform the development of targeted therapeutics for autoimmune disease, limiting the use of non-specific immunosuppression based therapies. As we have described above, several different pathways have been shown to regulate Spt-GC formation and activity by controlling GC initiation at the B cell:T cell border, by controlling GC selection through the regulation of FDC and Tfh function or by altering cell migration in the GC structure by cytokine and chemokine signaling. Although many of these pathways have been identified, few have been fully characterized, necessitating future work to investigate these pathways in greater detail while continuing to identify novel factors that regulate Spt-GCs.