Cytokine-producing B lymphocytes  key regulators of immunity

https://doi.org/10.1016/j.coi.2008.03.003Get rights and content

The successful use of B cell depletion therapy for the treatment of autoimmune disease has led to a resurgent appreciation of B cells as powerful regulators of immunity. However, to the surprise of many, B cells appear to regulate autoimmune conditions independently of their ability to produce autoantibodies. Indeed, disturbances in the ability of B cell subsets to present antigen, produce cytokines, and regulate the activities of T cells is emerging as a key feature in many inflammatory diseases. Here we review the recent literature describing cytokine-producing regulatory and effector B cell subsets in health and disease and discuss how future B cell-directed therapies might target the pathologic cytokine-producing effector B cell subsets without impacting the protective regulatory subsets.

Introduction

Rituximab, a humanized anti-CD20 antibody, is now being used clinically to deplete B cells in the treatment of multiple autoimmune diseases, including systemic lupus erythematosis (SLE) and rheumatoid arthritis (RA) [1••]. Although it would seem reasonable that B cell depletion therapy would ameliorate symptoms of disease by eliminating autoantibody-secreting B cells, many Rituximab-treated patients enter extended periods of clinical remission without reductions in serum autoantibody titers [1••]. These data suggest that B cells contribute to disease pathogenesis in an antibody-independent fashion, perhaps by modulating T cell responses.

Although it is not well appreciated, there is substantial evidence that B cells both amplify and suppress immune responses by mechanisms that do not involve antibody [1••]. For example, B cells respond to Toll-like receptor (TLR) ligands and present antigen. B cells also organize the structure of lymphoid tissues and regulate lymphangiogenesis. In addition, B cells produce cytokines and can be subdivided into discrete cytokine-producing ‘regulatory’ and ‘effector’ B subsets [2, 3••]. Regulatory B cells (Bregs) are distinguished by their ability to secrete IL-10 or TGFβ-1, while effector B cell populations produce cytokines such as IL-2, IL-4, TNFα, IL-6 (Be-2 cells) or IFNγ, IL-12 and TNFα (Be-1 cells). Here, we examine new data regarding the origin and functional properties of the effector and Breg subsets. We also review data showing that these cytokine-producing B cells can be either protective or pathologic and discuss the evidence that alterations in the types and amounts of cytokines made by B cells can affect autoimmune responses in both mice and humans.

Although B cells produce a variety of cytokines [2, 4], the first hint that cytokine-producing B cells regulate in vivo immune responses comes from studies examining the role of lymphotoxin (LT) and TNFα in lymphoid tissue organogenesis. These studies show that B-cell-derived LTα and TNFα control the development of follicular dendritic cells [5, 6], the formation of B cell follicles [7], and the development of specialized stromal cell subsets in the spleen [8]. Furthermore, LTα-expressing B cells regulate the formation of ectopic lymphoid tissue in autoimmune target organs [9], suggesting that cytokine-producing B cells contribute to pathology in autoimmune disease. More recent studies show that B-cell-derived TNFα also regulates T-cell-dependent antibody responses (Wojciechowski, Lund, unpublished data) as well as T cell responses to various pathogens ([10] and Lund et al., unpublished). Thus, the ability of B cells to produce cytokines in the TNF family is clearly important for multiple aspects of immunity.

Like T cells, B cells are not homogenous with respect to cytokine production. B cells primed by Th1 cells and antigen (Be-1 cells) make cytokines associated with type 1 immune responses, such as IFNγ and IL-12, while B cells primed by Th2 cells and antigen (Be-2 cells) make IL-2, IL-13, and IL-4; cytokines often associated with allergic responses [2, 11]. Interestingly, IFNγ-producing and IL-12-producing B cells can be identified in mice infected with pathogens that induce Th1 immunity [11, 12], while IL-4-producing B cells are found in mice infected with parasites that induce Th2 immunity [11, 13]. IFNγ-producing and IL-12-producing Be-1 like cells are also found in human peripheral blood and tonsil [14, 15, 16, 17] as well as ectopic lymphoid tissues [18], while IL-13-producing [19] and IL-4-producing [20] Be-2 cells are found in nasal polyps and germinal centers. Most excitingly, recent experiments clearly demonstrate that both mouse and human cytokine-producing effector B cells amplify effector T cell responses in a cytokine-dependent manner [11, 14, 15, 16].

Although some cytokines made by B cells amplify immune responses, IL-10-producing and TGFβ-producing B cells actively suppress immune responses [3••]. For example, IL-10-producing B cells are protective in models of colitis [21, 22, 23], promote remission in an EAE model [24, 25], and prevent induction of collagen-induced arthritis [26, 27•]. The activity of IL-10-producing Bregs is now documented in experimental models of tolerance [28, 29••, 30], tumor rejection [31], infectious disease [32, 33], and autoimmunity [21, 22, 23, 24, 26, 27•, 34, 35]. Likewise, there are also several papers describing the regulatory function of TGFβ-1-producing B cells [36, 37, 38]. Together, these publications demonstrate that there are distinct cytokine-producing effector and Breg subsets that can independently alter T-cell-mediated immune responses (Figure 1, Figure 2).

B cells are subdivided into two major lineages: B1 cells, which arise from fetal liver precursors and are enriched in mucosal tissues and the pleural and peritoneal cavities, and B2 cells, which arise from bone-marrow-derived precursors and are enriched in secondary lymphoid organs [39]. B1 cells can be further subdivided into B1a cells (CD11b+CD5+) and B1b cells (CD11b+CD5), while B2 cells can be subdivided into immature transitional cells (T1, T2, and T3) and mature follicular B cells (FO) or marginal zone (MZ) B cells. MZ B cells and B1 lineage B cells have the ability to respond very rapidly to inflammatory stimuli and antigen and can terminally differentiate into plasma cells in just one to two days. Thus, these B cells can play key roles in the early innate immune response to pathogens. By contrast, FO B cells represent an important component of the adaptive immune response. FO B cells can differentiate into short-lived plasma cells in three to five days after encountering with antigen or can enter a T-cell-dependent germinal center reaction where the B cells may undergo class switch recombination and affinity maturation. FO B cells that exit the germinal center seed the long-lived plasma cell or memory B cell pool.

Although little is known about the origins of effector B cells, the current data suggest that effector B cells are derived from FO B cells (Figure 1, Figure 2). First, both Be-1 and Be-2 cells can be generated from splenic B cells isolated from BCR transgenic mice that have few MZ B cells [11]. Second, experiments examining cytokine production by purified MZ and FO B cells indicate that only FO B cells produce IFNγ after TLR stimulation [40]. Finally, a recently identified long-lived pool of recirculating follicular B cells (FO II B cells) appears to have a greater propensity to make B-effector-associated cytokines like IFNγ, IL-12, IL-4, and IL-2 [41], suggesting that there may be specific subsets of FO B cells that are poised to develop into cytokine-producing Be-1 or Be-2 cells.

Many B cell subpopulations, including B1a, transitional, FO, and MZ B cells, produce IL-10 and have the potential to function as Bregs [3••]. Using adoptive transfer models, functional Bregs have been identified in the B1a B and MZ B lineages [21, 22, 27•, 29••, 30, 42]. As these B cells express TLRs and have the ability to respond rapidly to pathogen-derived products, they have been referred to as innate Bregs and are postulated to down-modulate inflammation associated with infection or disease [3••]. Consistent with this hypothesis, TLR ligands elicit poor inflammatory responses in neonatal mice because of a high frequency of IL-10-producing B1a cells [29••, 42]. The IL-10 made by TLR-activated neonatal B1a cells blocks IL-12 production by TLR9-activated DCs resulting in depressed Th1 priming and a predominant Th2 response [42] (Figure 3). Likewise, neonatal IL-10-producing B1a cells also suppress Th1 responses to alloantigens [30], perhaps explaining the long-standing observation that neonates are more tolerant to allografts than adult mice.

In neonatal mice, the IL-10-producing Bregs block TLR-induced inflammation and death [29••]. However, B1a cells are not able to suppress lethal inflammatory responses in adults [29••], suggesting that they are not a major source of adult Bregs. Instead, B2 cells from secondary lymphoid tissues play key roles in immune suppression in adults. For example, the transfer of normal B cells from mesenteric LNs (mLNs) into hosts with ulcerative colitis greatly diminishes local inflammation [21, 22]. Importantly, IL-10 production by the transferred B cells is crucial for blocking the inflammatory response [21, 22]. Although it is not yet known whether these IL-10-producing Bregs are derived from the FO or MZ lineage, the protective mLN B cells express CD1d [21] and high levels of CD19 [22] and thus phenotypically resemble splenic MZ B cells (Figure 2).

In addition to MZ B cells, MZ B cell precursors also suppress autoimmune disease in an IL-10-dependent manner. Immature B cells emerging from the bone marrow enter the spleen and move through additional rounds of selection and maturation (referred to as transitional T1, T2, and T3 cells) [43]. T2 B cells are phenotypically similar to MZ B cells and are reported to be enriched in MZ precursors [44, 45]. Transfer of T2 cells from arthritis-recovered mice to susceptible hosts prevented the recipients from developing Th1-mediated arthritis [27] (Figure 1). The T2 cells isolated from convalescent animals produce minimal amounts of IL-12 and IFNγ in response to antigen (collagen), but make larger quantities of IL-10 than either mature MZ or FO B cells [27], suggesting that the cytokine repertoire of these T2 B cells is biased toward anti-inflammatory cytokines. Importantly, IL-10 production by T2 B cells is crucial for their suppressive function, since IL-10-deficient T2 cells are unable to transfer protection to arthritis susceptible hosts [27]. Taken altogether, these data support the model that the Bregs and B effectors are separate subpopulations and that the best-studied IL-10-producing Bregs originate from the B lineages that contribute to innate immune responses, while the effector B cells originate from the FO B cell pool.

Although there are still many questions that remain regarding the developmental origins of effector and Bregs, progress has been made identifying the signals that regulate the differentiation of these subsets. Not surprisingly, cytokines play a key role in the ‘commitment’ of naïve B cells to the Be-1 or Be-2 lineages. Be-2 differentiation is dependent on the engagement of IL-4Rα on B cells [46] (Figure 2), while Be-1 cell development is dependent on the activation of the transcription factor T-bet and the IFNγR on B cells [47] (Figure 1). Interestingly, both Be-1 and Be-2 cell differentiation can proceed in vitro without BCR ligation, as long as peptide is provided to elicit cognate interactions between B cells and either Th1 or Th2 effectors [46, 47]. Indeed, Be-2 development is dependent on T cells that produce IL-4 and engage CD40 as well as CD80/CD86 [46]. By contrast, blocking the CD40/CD154 interactions between B cells and Th1 cells has little impact on Be-1 expansion or IFNγ production [2]. Despite the differences in the signals required to induce Be-1 or Be-2 development, both B cell subsets are competent to secrete antibody. However, the frequency of antibody secreting cells is increased in the in vitro generated Be-1 cultures [46], suggesting that the Be-1 culture conditions facilitate plasma cell differentiation.

Interestingly, unlike Be-2 differentiation, which is absolutely dependent on Th2 cells, Be-1 differentiation can proceed via a T-cell-independent mechanism [47] (Figure 1). For example, stimulation of naïve FO B cells with a single TLR ligand does not induce IFNγ production or Be-1 development [40]. However, stimulation of naïve B cells with multiple TLR ligands [40] or with a single TLR ligand and IL-12 and IL-18 [47, 48] promotes Be-1 differentiation and IFNγ production. As described earlier, IL-10-producing Bregs can also be activated by T-cell-independent stimuli, particularly TLR ligands [3••]. However, T-cell-dependent stimuli can also drive IL-10 production. In fact, both Be-1 and Be-2 cells make IL-10 and human peripheral blood B cells make IL-10 in response to many stimuli, including CD40 ligation [49, 50]. Furthermore, since T2 B cells from convalescent arthritic mice transfer protection much more efficiently than T2 cells from naïve mice [27], antigen priming may be required to activate these Bregs. Alternatively, the inflammatory environment may promote the differentiation of Bregs in the absence of antigen. In fact, type 1 interferons produced by TLR-activated DCs enhance IL-10 production by neonatal B1a cells activated with TLR2, TLR4, TLR7, and TLR9 ligands [29••] (Figure 3). However, type 1 interferons do not potentiate IL-10 production by adult B cells [29••] and instead appear to facilitate effector cell cytokine production (Lund, A Marshak-Rothstein, unpublished observation). Thus, the local cytokine milieu plays a crucial role in regulating the types and quantities of cytokines produced by B cells.

Both proinflammatory (TNFα, LTα, IL-12, and IL-6) and suppressive (IL-10) cytokines are produced by cultures of total peripheral blood B cells that contain a mixture of naïve and memory B cells [14, 15, 17, 50]. This balance of proinflammatory and anti-inflammatory cytokines is altered in B cells from patients with multiple sclerosis (MS), as B cells from MS patients make significantly less IL-10 than B cells from healthy individuals [51••]. Interestingly, this alteration in the proinflammatory to anti-inflammatory cytokine ratio is reversed in MS patients treated with either Mitoxantrone or Rituximab [51••]. This change in B cell cytokine production correlates with a significant reduction in the number of memory B cells in the treated patients [51••]. Consistent with this result, CD27+ memory B cells produce inflammatory cytokines like IL-12, LTα, and TNFα [17, 51••], while CD27neg naïve B cells make IL-10 [51••]. Thus, one of the major impacts of B cell depletion therapy is that the ratio of memory ‘effector’ B cells to naïve ‘regulatory’ B cells is reversed in favor of the naïve B cells, leading to a more suppressive or tolerogenic cytokine profile.

Likewise, changes in other B cell subsets also lead to alterations in the B cell cytokine repertoire. For example, IL-10 production by B cells is greatly increased in several mouse models of SLE [34, 52]  probably because of the high frequency of MZ B cells in these mice. In another example, stimulation of autoreactive, rheumatoid factor (RF)-specific murine B cells with immune complexes that simultaneously engage the BCR and TLR9 induces much higher levels of IL-2 than stimulation with either anti-IgM or CpG ODNs alone [53]. Furthermore, RF-specific B cells consistently produce more IL-2 in response to anti-IgM plus CpG ODNs than comparably stimulated normal B cells [53]. These data collectively suggest that the pattern of cytokines expressed by particular B cells is not fixed and that the types and amounts of cytokines produced by B cells can be influenced by the integration of signals from multiple receptors. Moreover, cytokine production by the overall B cell population in an individual can be altered by intrinsic changes in the activity of individual B cells or by changing the composition of the various B cell compartments.

Section snippets

Conclusions

New and accumulating evidence clearly demonstrates that B cells are not merely antibody-producing factories, but actively regulate immune responses by producing cytokines. IL-10-producing Bregs promote tolerance and suppress inflammatory responses, while effector B cells amplify humoral and cellular immune responses. The current evidence suggests that Bregs and effector B cells probably originate from different precursors  with the IL-10-producing B cells arising in large part from B1a cells and

References and recommended reading

Papers of particular interest, published within the annual period of review, have been highlighted as:

  • • of special interest

  • •• of outstanding interest

Acknowledgements

I would like to thank Dr Troy Randall for critically reviewing this manuscript. This work was supported by National Institutes of Health R01-AI0688056 and Trudeau Institute.

References (53)

  • V. Pistoia

    Production of cytokines by human B cells in health and disease

    Immunol Today

    (1997)
  • Y.-X. Fu et al.

    B lymphocytes induce the formation of follicular dendritic cell clusters in a lymphotoxin α-dependent fashion

    J Exp Med

    (1998)
  • M. Gonzalez et al.

    The sequential role of lymphotoxin and B cells in the development of splenic follicles

    J Exp Med

    (1998)
  • V.N. Ngo et al.

    Splenic T zone development is B cell dependent

    J Exp Med

    (2001)
  • C.M. Weyand et al.

    B cells in rheumatoid synovitis

    Arthritis Res Ther

    (2005)
  • L.C. Menard et al.

    B cells amplify IFN-gamma production by T cells via a TNF-alpha-mediated mechanism

    J Immunol

    (2007)
  • D.P. Harris et al.

    Reciprocal regulation of polarized cytokine production by effector B and T cells

    Nat Immunol

    (2000)
  • F. Ganapamo et al.

    CD19+ cells produce IFN-γ in mice infected with Borrelia burgdorferi

    Eur J Immunol

    (2001)
  • Q. Liu et al.

    The role of B cells in the development of CD4 effector T cells during a polarized Th2 immune response

    J Immunol

    (2007)
  • J.L. Schultze et al.

    Human non-germinal center B cell interleukin (IL)-12 production is primarily regulated by T cell signals CD40 ligand, interferon γ, and IL-10: role of B cells in the maintenance of T cell responses

    J Exp Med

    (1999)
  • M. Wagner et al.

    IL-12p70-dependent Th1 induction by human B cells requires combined activation with CD40 ligand and CpG DNA

    J Immunol

    (2004)
  • H. Shirota et al.

    B cells capturing antigen conjugated with CpG oligodeoxynucleotides induce Th1 cells by elaborating IL-12

    J Immunol

    (2002)
  • A. Gagro et al.

    Type I cytokine profiles of human naive and memory B lymphocytes: a potential for memory cells to impact polarization

    Immunology

    (2006)
  • C. Daridon et al.

    Polarization of B effector cells in Sjogren's syndrome

    Autoimmun Rev

    (2007)
  • O. Hajoui et al.

    Synthesis of IL-13 by human B lymphocytes: regulation and role in IgE production

    J Allergy Clin Immunol

    (2004)
  • A. Mizoguchi et al.

    Chronic intestinal inflammatory condition generates IL-10-producing regulatory B cell subset characterized by CD1d upregulation

    Immunity

    (2002)
  • Cited by (460)

    • Systems biology of B cells in COVID-19

      2024, Seminars in Immunology
    • B cells: The many facets of B cells in allergic diseases

      2023, Journal of Allergy and Clinical Immunology
    • Clinical efficacy of anti-CD20 antibodies in neurological and neuromuscular diseases

      2023, Resistance to Anti-CD20 Antibodies and approaches for their Reversal: Volume 2
    View all citing articles on Scopus
    View full text