Review: Abnormal B Cell Development in Systemic Lupus Erythematosus: What the Genetics Tell Us

Systemic lupus erythematosus (SLE) is an autoimmune disease characterized by B cell dysfunction, production of autoantibodies directed toward cellular and nuclear components, and multiorgan damage caused by immune complex deposition and inflammation within affected tissues 1. It largely affects women of childbearing age (the third and fourth decades of life) and is associated with significant morbidity and mortality. In healthy individuals, B cells with autoreactive receptors are selected out during B cell maturation, starting at the initial stages of B cell receptor (BCR) development in the bone marrow and continuing through to the fine tuning that occurs in activated mature B cells in secondary lymphoid tissue. Studies in lupus patients as well as mouse models indicate that these processes are altered in SLE. The etiology of the disease is complex and its phenotype is highly heterogeneous, but genetic susceptibility is thought to contribute as much as 60% of disease risk 2. Although rare monogenic causes do exist, heredity in SLE is complex, with multiple common variants contributing to disease, with patients having to achieve a certain “genetic threshold” for disease risk. This genetic risk, in combination with environmental factors (exposure to ultraviolet sunlight, smoking, and infections including Epstein-Barr virus have all been implicated), leads to development of the disease 1. In this review, we summarize some of the B cell anomalies in SLE and incorporate evidence from studies in humans and mouse models, together with data from genetic association studies, to explain the mechanisms behind B cell dysregulation in SLE. The crucial role of B cells in SLE pathogenesis is well recognized, from producing autoantibodies to abnormal regulation of immune responses 3, 4. Various abnormalities have been noted in SLE B cells. First, there is an imbalance of B cell subtype numbers, with an increase in class-switched memory B cells relative to naive B cells 3. Second, B cells from SLE patients have exaggerated BCR responses, with receptor crosslinking leading to increased calcium influx and tyrosine phosphorylation of downstream signaling molecules 3. Increased memory B cell numbers confer significant disease risk as these have a lower activation threshold, allowing autoreactive B cells to thrive with minimal antigen contact, while enhanced receptor activation contributes to the steady-state active phenotype seen in SLE 3, 5. B cells contribute to disease mainly by producing autoantibodies targeting nuclear components including DNA (anti–double-stranded DNA anti-dsDNA), RNP particles (anti-Ro, anti-La, and anti-Sm), histones, and nonhistone chromatin proteins. These are present in >90% of patients and contribute to disease progression via immune complex formation 6. Titers of these autoantibodies (especially anti-dsDNA) correlate positively with increased disease activity, and serial measurements are used to monitor patients for disease flares 6. There is also evidence that autoantibodies cross-react with cellular components other than nuclear targets 7. For example, anti-dsDNA antibodies bind to major glycosaminoglycan components in the glomerular basement membrane, suggesting a possible direct role in nephritis 7. In mouse models, transfer of autoantibodies from diseased to unaffected animals leads to development of typical immune complex–mediated nephritis 8. Moreover, in MRL/lpr mice (which develop lupus-like disease spontaneously), disease severity can be attenuated and mortality reduced by ~50% if antibody secretion is blocked, providing robust evidence that autoantibodies are more than spectators in disease etiology 9. A recent explosion in genome-wide association studies (GWAS) has identified >80 potential risk loci across multiple immunopathologic pathways 10. In this review, we discuss how genetic variants affect the development of B cells, allowing them to overcome several checkpoints to break self tolerance, and how they contribute to the abnormal active phenotype observed in SLE. We examine how these genes alter both early developmental pathways in the bone marrow and late maturation processes to cause B cell dysregulation. Normal B cell development starts in the bone marrow, where the first round of negative selection of autoreactive B cells (termed central tolerance) occurs. This process is summarized in Figure 1. Many potential abnormalities in central tolerance have been implicated in SLE, including failure of adequate negative selection of autoreactive B cells and inadequate receptor editing (steps 6 and 3, respectively, in Figure 1), both of which are critical steps in maintaining tolerance to self 11. The molecular mechanisms by which SLE autoreactive B cells evade central tolerance have yet to be fully elucidated. There are some clues, however, from various genetic studies 12. Patients with single-gene mutations resulting in primary immunodeficiencies frequently develop a wide range of autoimmune diseases in addition to increased susceptibility to infections 12. Their mutations teach us that central tolerance is largely dependent on adequate BCR signaling in the bone marrow 12 (steps 2 and 3 in Figure 1). In X-linked agammaglobulinemia, a defect in the gene for Bruton's tyrosine kinase (needed for downstream BCR signaling) results in increased frequency of autoreactive B cells 12. One possible mechanism is that binding to self antigen does not induce a strong enough response in BCR signaling to trigger clonal deletion 12. Conversely, deficiency in Wiscott-Aldrich syndrome protein (WASP; a negative regulator of BCR signaling) results in more stringent central control mechanisms, with WASP-knockout mice showing a much lower proportion of autoreactive B cells being released from the bone marrow. Their mature B cells also show abnormal peripheral tolerance and hyperactive phenotype, possibly driven by T cell abnormalities 12. Outside of single-gene mutations, GWAS have expanded our knowledge of the molecular basis of B cell developmental anomalies in SLE 13, 14. These GWAS have identified several SLE susceptibility loci near genes known to be important for early B cell development and BCR signaling (see Table 1 for summary). Variants affecting BCR signaling are discussed in more detail later in this review 13, 14. In early B cell development, several stages have been described which are associated with distinct genetic and molecular events. Two of these genes have been identified as risk loci and are discussed in the next section. Commitment of the multipotent progenitor cells in the bone marrow to lymphocyte development depends on the expression of several transcription factors, including Ikaros (encoded by IKZF1) and Aiolos (encoded by IKZF3), among others 13, 15 (step 1 in Figure 1). Ikaros in particular is known to be crucial for early commitment to B cell lineage. Its exact role in the early multipotent progenitor cells is unknown, although we know that mice deficient in Ikaros fail to develop any common lymphocyte precursors, with arrest of B cell development before lineage commitment to the B cell–biased lymphoid progenitor 16. Low expression of Ikaros allows generation of some B cells, but overall numbers remain low and differentiation is impaired at all stages 16. The pre–B cell stage is characterized by the expression of the pre-BCR, and successful signaling through the pre-BCR arrests recombination of the IgH chain and the initiation of expression of the Ig light chains of the final BCR 17 (step 2 in Figure 1). Ikaros and its closely related family member Aiolos are both induced on engagement of the pre-BCR and help terminate signaling through the pre-BCR (steps 2 and 3 in Figure 1), promote exit from the cell cycle, and allow rearrangement of the Ig light-chain genes. Ikaros also induces expression of recombination-activating gene 1 (RAG-1) and RAG-2 and is required for IgH VH gene recombination, allowing the pro–B cell to progress to the large pre–B cell stage (18). Both the Ikaros and Aiolos variants (single-nucleotide polymorphism SNP rs4917014, which lies within the 3′-untranslated region 3′-UTR of the IKZF1 gene encoding Ikaros P = 2.7 × 10–23, and SNP rs2941509, which lies within the 5′-UTR of the IKZF3 gene encoding Aiolos P = 3.198 × 10−6) have been associated with increased transcription of their respective genes in whole blood 13, 15 (Table 1). In models of overexpression in pre–B cell lines, increases in Ikaros and/or Aiolos induce termination of IgH recombination and stop signaling through pre-BCRs. However, high levels of expression are required to induce this process; therefore, it is plausible that both of these variants are promoting early transition to the small pre–B cell and may be contributing to the inadequate receptor editing observed in SLE 11, 16, 19. Humans with germline mutations in IKZF1 have an early block in common lymphocyte precursor development, with reduced pro–B cell numbers and normal pre–B cell numbers 20. Approximately half of the reported patients also developed autoimmune disease, including 1 who had SLE, suggesting that dysfunction and abnormalities of early B cell development can result in both immunodeficiency and autoimmunity 20. Many abnormalities in peripheral tolerance have been identified in SLE, from problems with somatic hypermutation to memory B cell dysfunction (see Figure 2 for summary). First, SLE patients show aberrant and raised RAG expression in peripheral B cells (step 6 in Figure 2), raising the possibility that some autoreactive B cells arise as a result of mutation of a “healthy” BCR into one that recognizes self antigen 21. This hypothesis is supported by analysis of genetic variation in the Ig produced by autoimmune mice, showing that point mutations can render a previously self-tolerant BCR autoreactive 22. These data have been replicated in the analysis of anti-dsDNA antibodies from humans with SLE 23. Second, autoantibodies in SLE often evolve over the duration of the disease, recognizing different epitopes of their respective antigens and frequently achieving a higher affinity 23, 24. The raised levels of interleukin-6 (IL-6) seen in the disease may contribute to this, as IL-6 is a known up-regulator of RAG gene expression 25. The other major player in this process of secondary maturation and in the maintenance of tolerance is the T cell–B cell interaction (step 5 in Figure 2). This is supported by evidence from single-gene immunodeficiency disorders, mouse models, and GWAS 12, 14, 26. Patients with single-gene mutations in CD40, CD40L, and major histocompatibility complex (MHC) class II develop a significant proportion of autoreactive B cells, including those with receptors recognizing nuclear components (including antinuclear antibodies ANAs) 26. This is despite their having normal central tolerance processes in the bone marrow 12, 26. In the Sle1 murine model, disease risk is inherited via a region on chromosome 1, carrying polymorphisms in genes encoding multiple receptors needed for T cell–B cell interactions (including Slam, Ly108, Cd84, Cracc, and Ly9) 27. In the germinal centers (GCs) of these mice, transient short-term contact between B cells and T cells allows rare autoreactive B cells and T cells to “sample” many different cells in the GC, increasing chances of interaction for positive costimulation 27. Shorter contact times between immune cells are also known to alter their function (e.g., poorer Treg cell ability to induce tolerance in target cells due to shorter contact times) 28. It is possible that shorter T cell–B cell contact in the GC may lead to chronic, low-grade activation, allowing autoreactive B cells to survive due to background low-quality contact, without receiving adequate signals to become anergic. Patients and mice with deficiencies in CD40, CD40L, and MHC class II who have poor T cell–B cell interaction overcome lack of B cell stimulation by up-regulating BAFF, a stimulatory cytokine that promotes B cell survival and proliferation 26. This has important consequences for B cell activation and murine models in which mice that overexpress BAFF develop lupus-like disease with ANAs and anti-dsDNA 29. These findings suggest that in the absence of specific and controlled BCR activation, more generic signals (such as BAFF) promote indiscriminate B cell activation and survival of self antigens recognizing B cells and normal foreign antigens recognizing cells equally (29). Variants in loci near or within OX40L/TNFSF4, MHC class II, and CD80 have all been implicated as associated with risk of SLE in various GWAS 13, 14, 30, 31 (Table 1). The SLE-associated in the of the gene results in of expression of the transcription in whole blood from humans Studies in mice that are both deficient in and an autoreactive BCR indicate that while central tolerance is deficiency in results in impaired responses B cells from mice produced autoantibodies despite receiving signals and to them in the absence of interaction with antigen of B cell (such as reduced tyrosine phosphorylation and be induced in deficient cells stimulation with however, this not production The gene (which for also known as on chromosome a risk SNP = × within its that leads to increased expression and protein levels in B cells not in other immune cells as T activation through increased chain complex–mediated and of an of levels of to steady-state activation and responses to stimulatory signals This in higher numbers of both and cells and response to stimulatory signals as and (Table 1 and 5 in Figure in of the risk Variants affecting B cell and pre–B cell signaling affect both central and peripheral tolerance, and many genes in both BCR and pre-BCR signaling have been identified from GWAS 13, 14. These variants contribute to the generation of autoreactive B cells and the activated phenotype identified in peripheral B cells in SLE 13, (Table 1). Variants resulting in impaired BCR signaling are thought to contribute to autoimmunity in a mechanism to that noted in the monogenic immunodeficiencies discussed in which inadequate pre-BCR signaling leads to failure in the for clonal deletion or BCR rearrangement 12, Variants contributing to increased signaling may promote the peripheral active B cell phenotype of SLE variants identified in SLE and 13, 14, (step 3 in Figure 1 and in Figure 2). The risk = × for which the tyrosine kinase is associated with impaired BCR signaling and is thought to contribute to autoreactive B cell survival through various mechanisms (Table 1). This risk is associated with central tolerance and failure to autoreactive B cells in the bone marrow In peripheral B cells carrying this risk many genes in B cell activation, including those in CD40, and receptor signaling as well as cytokine receptors (e.g., receptor and Moreover, these B cells show increased expression of as well as enhanced responses to contributing to the active peripheral phenotype of these genes is thought to be a mechanism by which B cells overcome the in function caused by the risk a family tyrosine is at the pre–B cell stage and is important in the of pre-BCR signaling the Ig and of the BCR and is known to bind the a complex with the B cell protein (encoded by the gene also a susceptibility for during BCR signaling The for and both result in the of expression and potential failure of adequate pre-BCR signaling (Table 1). A third and of associated with SLE that results in an in its leads to and of the protein This also impaired binding to B cell protein with 1 which is important in mice develop autoimmune disease, with anti-dsDNA autoantibodies and immune complex–mediated together with an increase in the proportion of cells subtype of B cells in of this subtype of B cells in other murine models of lupus is well and is thought to be in nephritis In a mechanism to that of the this may contribute to autoimmunity by central tolerance However, abnormalities in peripheral BCR signaling have also been with low BCR at but enhanced responses on activation, ability to T cells, and increased numbers of memory B cells on chromosome a risk that expression of a family tyrosine kinase important in the BCR signaling B cells from those carrying the risk show increased levels of expression and increased phosphorylation of BCR responses, and activation of mature B cells member of the kinase has also been associated with SLE GWAS identified in the region = × and in the region = × in women of with both variants being studies identified a third in women of in the 5′-UTR of the gene = This also and associated with disease within association between any variants and disease has been identified in of or has there been any association with in protein or levels However, data from SLE patients have them to have lower levels of to healthy that lower levels may be associated with disease In studies in B cell show that is required for signals in response to BCR crosslinking mice are to an immune complex–mediated nephritis as well as to Patients deficient in kinase differentiation and B molecules in the also have high levels of autoreactive B cells, suggesting that impaired signaling via these pathways results in survival of autoreactive possibly by to the signaling for clonal deletion in the bone marrow 12 (step 3 in Figure 1). These patients also impaired receptor editing and had in peripheral tolerance SNP in the of = 2 × has been associated with SLE in several GWAS 14. This is associated with increased levels and resulting in an enhanced 14. A in patients with SLE that variants in associated with certain disease and specific For example, some variants associated with nephritis and anti-dsDNA antibodies Although abnormalities in both central and peripheral tolerance are to memory B cells and cells contribute to disease One of the B cell findings in SLE is the high of class-switched memory cells 3. The of having a large proportion of these cells is First, these cells are for and have a lower activation to naive B cells. Second, they to and signals 3. In SLE patients have a of memory B cells in the which lack expression of memory B cell the numbers of which correlate with increased disease activity, and higher levels of autoantibodies The of autoreactive cells and their to as B cell and the of this cell subtype in disease In mice, cells develop early and to their models indicate a role for these cells in for example, antibodies from cells in diseased of mice 8. In these cells are thought to be for the production of which the of these antibodies are not altered with B cell and survival of cells and memory cells is complex and depends on a GC the GC, there is stringent selection of cells with receptors to become cells, while memory B cells a selection process B cells do increased affinity over and some studies suggest that they may the and cells in within the bone marrow, memory B cells are in the and in secondary lymphoid Various GWAS in SLE have identified an association with the gene encoding the protein member (Table 1 and 5 in Figure 2). This gene for a protein on the of memory B cells, with its receptor on cells. The interaction results in signaling in both the B cell and T cell activation in both cells, and of the immune response Although the of the identified on expression has not been fully we know that both and are at higher levels in patients with SLE, those with more active disease and with a more disease phenotype by the of These findings suggest that these contribute to the increased and more active phenotype of memory B cells in SLE. GWAS data also show disease association with many of the genes associated with the development of memory including and IKZF3 13, 14, (Table 1 and steps in Figure 2). The exact of the risk in of these genes on function is not yet but these that the may contribute to the dysregulation of the developmental pathways in memory and cells. particular we know that mice Aiolos fail to develop memory although they have normal initial responses to antigen and are at increased risk of B cell mice also develop a lupus-like disease in with typical autoantibodies and immune complex deposition This autoimmune phenotype as a result of a B cell defect with of both normal T cells and immune Two polymorphisms in have been identified as associated with risk 13, 14, (Table 1). Both BCR and signaling and are associated with an of memory B cells in (step in Figure 2). a protein that during BCR signaling by promoting phosphorylation of tyrosine as through interactions between them and receptor One of the risk polymorphisms is associated with which results in a protein that and more within the The of the other risk on expression is of tolerance in SLE is due in large to the discussed but it is important to that B cells develop in an abnormal in lupus patients to healthy of bone marrow from patients with SLE some findings First, there more cells within lupus bone marrow, pre–B cells to more nuclear self antigens and the development of autoreactive Second, there increased numbers of T cells, and cells, a These cells produced significant of high of the cytokine in the bone marrow associated with increased disease which has on B cell development of in bone marrow leads to development at the early stages and at the B cell stage the lymphoid in SLE is also altered The are with a high proportion of cells proliferation and normal cells are but in SLE there are in this process review, leading to of including nuclear components and the B cell response Although of risk variants the B cell as the cell of for the disease, risk variants affecting many of the immune from to have been described 14, Many of the risk loci associated with SLE discussed in this review also affect genes with For example, the IKZF1 has also been to affect expression of components and several response while is required for T cell receptor signaling 14, of this is the affecting which for B maturation protein 1 a transcription crucial for cell The of the risk however, is in cells which show a in expression as a result of the a not replicated in B cells in is associated with an increase in IL-6 secretion as well as increased numbers of T cells and GC B cells, with mice disease while the of a specific may be B cell dysfunction, the molecular may not be B cell Although the abnormalities in SLE are complex, and not fully B cells have a central role in the development of the disease, from immune complex production to secretion of SLE is an complex disease, the genetic association data that with many affecting several of the immune 13, 14. The of GWAS has identified many potential abnormal pathways to the of self tolerance in SLE in Table 1). The data the central role of B cells, with a significant proportion of identified genes affecting B cell function and this 13, 14, GWAS have also to explain between is to how is to have a different combination of risk variants resulting in a of immune dysfunction to that We can how variants in genes can help disease and or how variants in the gene for can B has been on of numbers (e.g., through targeting through but with our are being For example, Aiolos with the results in of genes B cell and antibody secretion in B cells from SLE patients as antibody to has been to block B cell differentiation and activation through of (encoded by how data from GWAS can Although the results from GWAS the of SLE pathogenesis and show that different B cell abnormalities are to disease in the the of central tolerance BCR signaling) and peripheral tolerance T cell–B cell are us to molecules can be through the exact pathways in an and the it for important and the final to be