Mice overexpressing B cell activating factor of the TNF family (BAFF) develop systemic autoimmunity characterized by class-switched anti-nuclear Abs. Transmembrane activator and CAML interactor (TACI) signals are critical for BAFF-mediated autoimmunity, but the B cell developmental subsets undergoing TACI-dependent activation in settings of excess BAFF remain unclear. We report that, although surface TACI expression is usually limited to mature B cells, excess BAFF promotes the expansion of TACI-expressing transitional B cells. TACI+ transitional cells from BAFF-transgenic mice are characterized by an activated, cycling phenotype, and the TACI+ cell subset is specifically enriched for autoreactivity, expresses activation-induced cytidine deaminase and T-bet, and exhibits evidence of somatic hypermutation. Consistent with a potential contribution to BAFF-mediated humoral autoimmunity, TACI+ transitional B cells from BAFF-transgenic mice spontaneously produce class-switched autoantibodies ex vivo. These combined findings highlight a novel mechanism through which BAFF promotes humoral autoimmunity via direct, TACI-dependent activation of transitional B cells.
Although overexpression of the B cell survival cytokine, B cell activating factor of the TNF family (BAFF) has been implicated in the pathogenesis of systemic lupus erythematosus (SLE) in murine models and human studies (1), the underlying mechanisms through which BAFF promotes breaks in B cell tolerance and autoantibody production are unclear. BAFF and the related cytokine APRIL (a proliferation-inducing ligand) are TNF family cytokines with important roles in promoting peripheral B cell survival, development, and activation. BAFF exerts its impact on B cells by binding to both the BAFF receptor (BAFF-R) and transmembrane activator and CAML interactor (TACI), whereas APRIL binds TACI and the B cell maturation Ag (1). Because BAFF-R deletion causes the loss of peripheral B cells beyond the transitional stage (2), this receptor was anticipated to be the major contributor to the autoimmune phenotype of BAFF-transgenic (BAFF-Tg) mice. However, a recent study demonstrated that TACI deletion prevents humoral autoimmunity in BAFF-Tg mice (3), an observation that we confirm in this article.
Importantly, the specific B cell subset that is the target of TACI-dependent activation during BAFF-mediated autoimmunity is not known. Developing B cells progress through transitional stages in the spleen prior to entering follicular mature (FM) and marginal zone (MZ) B cell compartments, a process that is impacted by developmental cues, including BCR and BAFF-mediated signals. Although mature FM and MZ B cells are most frequently implicated in the generation of Ab responses, direct activation of transitional B cells was described (4, 5), suggesting a potential physiologic explanation for immature B cell transit in the periphery.
In this context, we report the surprising observation that transitional B cells are an important source of serum autoantibodies in BAFF-Tg mice. Mechanistically, self-ligand engagement promotes TACI upregulation on a subset of autoreactive transitional B cells, resulting in TACI-dependent activation and class-switched autoantibody production in high-BAFF settings. By identifying transitional B cells as the dominant splenic B cell subset promoting BAFF-mediated autoimmunity, our findings have important implications for the understanding of SLE pathogenesis and B cell autoimmunity in other high-BAFF settings, including the events following therapeutic B cell depletion or stem cell transplantation. In addition, although the proportion of TACI+ transitional B cells was markedly expanded in BAFF-Tg mice, we detected a distinct TACI+ transitional subset in wild-type mice at physiologic serum BAFF levels. Because infectious challenge can promote local BAFF generation (6), we predict that TACI-dependent activation of transitional B cells also contributes to rapid, T-independent Ab responses to blood-borne pathogens.
Materials and Methods
C57BL/6, BAFF-Tg (7), Taci−/− (8), Baffr−/− (9), Rag2.GFP-Tg (10), and Nur77.GFP-Tg (11) mice and relevant murine crosses were bred and maintained in the specific pathogen–free animal facility of Seattle Children’s Research Institute. All animal studies were conducted in accordance with Seattle Children’s Research Institute Institutional Animal Care and Use Committee–approved protocols. Mice were sacrificed at 3–6 mo of age.
Anti-murine Abs used included B220 (RA3-6B2), CD138 (281-2), CD21 (7G6), CD24 (M1/69), CD80 (16-10A1), CD86 (GL1), and CD5 (53-7.3) (all from BD Biosciences); CD19 (ID3), CD44 (IM7), CD21 (7E9), CD23 (B3B4), CD24 (M1/69), and TACI (1A1) (all from BioLegend); CD11c (N418), TACI (8F10-3), AA4.1 (AA4.1), activation-induced cytidine deaminase (AID; mAID-2), T-bet (4B10), and CD11b (M1/70) (all from eBioscience); B220 (RA3-6B2) (Life Technologies); TACI (166010) (R&D Systems); and goat anti-mouse IgM-, IgG-, IgA-, IgG1-, IgG2b-, IgG2c-, and IgG3-HRP conjugated, unlabeled, or isotype IgG2c (1079-02) (Southern Biotechnology).
Measurement of autoantibodies
HEp-2 anti-nuclear Ab (ANA) immunofluorescence and specific autoantibody ELISAs were performed, as described (12).
Spleen immunofluorescence staining
Acetone-fixed splenic sections were stained with B220-PE, CD3-allophycocyanin, and IgG2c-FITC, as described (12).
Flow cytometry, cell sorting, and in vitro B cell culture
Single-cell splenocytes or peritoneal cells were stained with fluorescence-labeled Abs for flow cytometry analysis, intracellular staining was performed using a fixation/permeabilization kit (BD Biosciences), and intranuclear staining was performed using the FOXP3 Fix/Perm Buffer Set (BioLegend). Cell sorting was performed on CD43-depleted splenocytes, using a FACSAria II sorter (BD Biosciences), with the following sort gates: FM, CD24intCD21int; MZ, CD21hiCD23lo; and transitional 1/2 (T1/T2), CD24hiCD21lo-int, with BAFF-Tg T1/T2 further subdivided as TACI− and TACI+. Sorted B cell subsets were cultured in RPMI 1640 at 2 × 105 cells/well in a 96-well plate, with or without R848 (5 ng/ml), at 37°C for 72 h prior to the collection of supernatant for Ab ELISA.
RT-PCR and κ-deleting recombination excision circle analysis
RT-PCR was performed with murine β2-microglobulin as control using the following primers: β2-microglobulin, 5′-CTTCAGTCGTCAGCATGGCTCG-3′ (forward) and 5′-GCAGTTCAGTATGTTCGGCTTCCC-3′ (reverse); Taci, 5′-ACCCCCAGTGTGCAGTAGAG-3′ (forward); RP, 5′-GGAGGTGGAAGTCAGGTCAG-3′ (reverse); Aicd, 5′-CCTCCTGCTCACTGGACTTC-3′ (forward) and 5′-GGCTGAGGTTAGGGTTCCAT-3′ (reverse); Tbx21, 5′-GGTGTCTGGGAAGCTGAGAG-3′ (forward) and 5′-CCACATCCACAAACATCCTG-3′ (reverse); and Igg2c, 5′-GGGAATTCGAGGTGCAGCTGCAGGAGTCTGG-3′ (forward) and 5′-GCTCAGGGAAATAACCCTTGAC-3′ (reverse). Replication history of sorted B cell subsets was determined by κ-deleting recombination excision circle (KREC) analysis (13).
Single-cell BCR cloning
Single-cell BCR cloning was performed as described (14). Briefly, Ig heavy and light (κ and λ) gene transcripts from sorted single GFP+ and GFP− T2 (CD21intCD24hi) cells from Rag2-GFP.BAFF-Tg mice were cloned into human IGG1, IGK, or IGL expression vectors, transfected into HEK293T cells, and mAbs purified from culture supernatants using protein A–agarose beads.
The p values were calculated using a two-tailed Student t test, a Mann–Whitney U test, or one-way ANOVA followed by the Tukey multiple comparison test with GraphPad software.
Results and Discussion
Humoral autoimmunity in BAFF-Tg mice requires TACI
BAFF-Tg autoimmunity is T cell independent, but it requires the signaling adaptor MyD88 (15). Because TLR signals are critical for humoral autoimmunity, lack of disease in Myd88−/−.BAFF-Tg mice has been presumed secondary to absent MyD88-dependent TLR activation (16). However, because TACI also signals via MyD88 (17), loss of TACI signals might also explain this phenotype. Although TACI was proposed to act as a negative regulator of BAFF signaling in B cells (18, 19), Taci−/−.BAFF-Tg mice exhibited a striking loss of IgG ANAs and RNA-associated Sm/RNP autoantibodies across the spectrum of Ab isotypes and Ig subclasses (Fig. 1A, 1B), without impacting BAFF-Tg B cell expansion (Supplemental Fig. 1A). In keeping with this idea, Figgett et al. (3) reported decreased autoimmunity in irradiated BAFF-Tg mice reconstituted with Taci−/− bone marrow. Together, these observations demonstrate that TACI is required for the development of humoral autoimmunity in BAFF-Tg mice.
Excess BAFF promotes TACI expression by a subset of transitional B cells
To begin to understand how TACI signals might promote BAFF-Tg autoimmunity, we first assessed surface TACI (sTACI) expression on developing B cell subsets in wild-type (WT) and BAFF-Tg mice. Consistent with prior reports, sTACI in WT mice was low on transitional (T1, CD21loCD24hi; T2, CD21intCD24hi) B cells but increased on mature (FM and MZ) B cells. Although sTACI on FM and MZ B cells did not differ significantly between WT and BAFF-Tg mice, a prominent subpopulation of T1 and T2 B cells in BAFF-Tg mice expressed sTACI at levels exceeding those of previously reported TACI+ MZ B cells (Fig. 1C, 1D, Supplemental Fig. 1B). As predicted, BAFF-R (BR3) levels were increased in mature (FM and MZ) relative to transitional T1 B cells in WT mice (data not shown). However, we were not able to compare BAFF-R expression between the TACI+ and TACI− transitional subsets in BAFF-Tg mice because BAFF-R is markedly decreased on their splenic B cells, consistent with physiologic receptor downregulation in the setting of high serum BAFF levels (20).
Increased sTACI correlated with a greater abundance of Taci transcripts, consistent with transcriptional regulation of TACI in a subset of BAFF-Tg transitional B cells (Fig. 1E). Although the proportion of TACI+ transitional cells was significantly increased in BAFF-Tg mice, a distinct subset of WT transitional cells also expressed higher levels of sTACI (Fig. 1C, 1D). This prominent expansion of TACI+ T1/T2 B cells in BAFF-Tg mice suggested that transitional cells might contribute directly to TACI-dependent humoral autoimmunity.
Notably, relative to WT transitional cells, BAFF-Tg T1 and T2 transitional cells exhibit reduced CD93 (AA4.1), a marker commonly used to define transitional B cells (21). In addition, the AA4.1− transitional subset was predominantly TACI+ (Fig. 1F), which suggested that TACI+ cells within the CD21lo or CD21intCD24hi gate might be derived from mature, rather than transitional, B cells (21). However, several lines of evidence strongly supported the conclusion that TACI+ B cells within the CD21loCD24hi and CD21intCD24hi gates are derived from immature, transitional B cells. First, using Rag2-GFP reporter mice (10) to label recent bone marrow emigrants, we consistently observed a lack of AA4.1 expression on a subset of transitional B cells at physiologic BAFF levels (Fig. 1G), indicating that the lack of AA4.1 expression per se does not imply a nontransitional origin for TACI+ T1 B cells. Moreover, Rag2-GFP+ T1 B cells with lower GFP expression (i.e., the subset of T1 B cells that has transited through the cell cycle leading to ∼50% dilution in the GFP signal) exhibit lower AA4.1 and increased sTACI, consistent with reciprocal regulation of these surface markers on early transitional B cells (Supplemental Fig. 1C). Second, TACI+ T1 B cells lacking AA4.1 expression develop in BAFF-R–null (Baffr−/−) mice at equivalent proportions to WT mice (Fig. 1H), despite the marked reduction in mature B cells in this strain (9). Third, BAFF-Tg TACI+ transitional B cells lacked the plasma cell marker CD138+, as well as the age-associated B cell markers CD11b and CD11c (22), indicating that this population is distinct from previously described activated B cell subsets in autoimmunity (Fig. 1I). Fourth, splenic TACI+ transitional B cells in BAFF-Tg mice were distinct from peritoneal B1b cells recently hypothesized to promote BAFF-mediated autoimmunity (23) (Supplemental Fig. 1D). Finally, the expansion of TACI+ transitional B cells in BAFF-Tg mice precedes the development of autoimmunity, arguing against the accumulation of TACI+ autoreactive B cells derived from mature B cell subsets (Supplemental Fig. 1E). Together, these data highlight the surprising observation that excess BAFF promotes the expansion of sTACI-expressing transitional B cells.
TACI expression marks an activated, proliferative subset of transitional B cells
Based on these findings, we next sought to better define the phenotype of the TACI+ transitional cell population. BAFF-Tg TACI+ transitional cells were significantly larger than BAFF-Tg TACI− or WT transitional cells, implying increased activation (Fig. 2A). Consistent with this idea, BAFF-Tg transitional cells exhibited increased CD44, CD80, and CD86 expression, and this increase in activation markers was specific to the TACI+ subset (Fig. 2B, 2C). Notably, the TACI+ T1 subset from WT and Baffr−/− mice exhibited a similar activated phenotype (Supplemental Fig. 1F).
To address whether this activated subset is cycling in vivo, we crossed BAFF-Tg with Rag2-GFP reporter mice to allow quantification of transitional cell proliferation by GFP dilution (10). Relative to WT Rag2-GFP reporter cells, an increased proportion of BAFF-Tg transitional cells were GFP−. The BAFF-Tg GFP− (i.e., divided) subset was TACI+, whereas GFP+ cells were TACI− (Fig. 2D). Consistent with these findings, BAFF-Tg transitional cells had proliferated based on KREC analysis, with the average number of cell divisions increased in TACI+ versus TACI− cells (Fig. 2E) (13). Together, these data demonstrate that excess BAFF promotes transitional B cell activation and proliferation and that sTACI marks this activated, cycling subset.
TACI expression correlates with BCR engagement in transitional B cells
Because BCR ligation upregulates TACI on mature B cells (24), we next assessed whether BCR signal strength correlates with sTACI expression using the Nur77-GFP reporter strain in which BCR signals activate GFP expression under control of the Nur77 regulatory region. In this model, mature FM and MZ B cells expressing a diverse (nontransgenic) BCR repertoire are predominantly GFPhi, whereas transitional cells exhibit a range of GFP expression dependent upon the relative strength of BCR engagement by self-ligand (11). Notably, although GFP expression in FM and MZ B cells did not differ between Nur77 and Nur77.BAFF-Tg mice, excess BAFF increased the proportion of GFPhi cells in the transitional compartment (Fig. 3A, data not shown). Subdividing transitional cells according to relative GFP expression demonstrated a correlation between BCR signal strength and sTACI on transitional cells in Nur77 and Nur77.BAFF-Tg mice, with relatively greater increases in sTACI in Nur77.BAFF-Tg mice (Fig. 3B).
BAFF-Tg TACI+ transitional cells express AID and T-bet
Dysregulated TLR7 signals can directly activate transitional B cells in TLR7-transgenic mice (5), resulting in transitional cell expression of AID and T-bet, a T-box transcription factor required for class-switch recombination to IgG2a/c (25). Strikingly, transitional B cells expressed AID and T-bet in BAFF-Tg mice, and this change was specific to TACI+ cells (Fig. 3C–E). In addition, quantitative PCR analysis demonstrated increased Aicd (encoding AID) and Tbx21 (encoding T-bet) mRNA transcripts in sorted BAFF-Tg transitional cells (Fig. 3F). Consistent with the role for AID and T-bet in class-switch recombination, we observed prominent extrafollicular IgG2c+ cells within the splenic red pulp in BAFF-Tg mice. In contrast, IgG2c+ cells were much less abundant and restricted to the follicles in Taci−/−.BAFF-Tg mice (Fig. 3G). Although the primary B cell subset(s) contributing to this extrafollicular IgG2c+ population remains to be definitely identified, transitional B cells in BAFF-Tg mice expressed abundant IgG2c mRNA transcripts (Fig. 3H), implying that the TACI+ transitional subset contributes directly to this activated population.
TACI+ cells are enriched for self-reactivity, exhibit somatic hypermutation, and spontaneously produce class-switched autoantibodies
Despite central tolerance mechanisms, a significant proportion of early transitional cells exhibit autoreactivity or polyreactivity (26). To test the hypothesis that the novel TACI+ transitional subset is enriched for autoreactivity, we cloned BCRs from single GFP− versus GFP+ T2 B cells sorted from Rag2-GFP.BAFF-Tg mice, an approach that is designed to directly assess the relative self-reactivity in cycling versus noncycling B cells, respectively. Consistent with our hypothesis, BCRs cloned from GFP− (i.e., divided, TACI+) T2 B cells exhibited reactivity to a panel of autoantigens, including significantly increased reactivity with phosphorylcholine (PC) and a trend toward greater dsDNA and Sm/RNP reactivity (Fig. 4A).
Strikingly, BCRs from the cycling GFP− transitional subset demonstrated increased Ig H and L chain mutation frequency, whereas mutations were rare in GFP+ BAFF-Tg T2 B cells and BAFF-Tg MZ B cells, as well as in BCRs from naive control B cells (Table I). Observed mutations were characteristic of AID-mediated somatic hypermutation, in that nucleotide substitutions were biased for G to A and C to T transitions and were predominantly localized to CDRs (Supplemental Fig. 2). Interestingly, one T2 GFP−-derived mAb exhibiting polyreactivity to PC, Sm/RNP, and dsDNA (Fig. 4A, highly reactive clone highlighted with dotted line) contained the greatest number of replacement mutations (four total; H chain S64I [CDR2], M91I [FWR3]; κ L chain S14F [FWR1], N56D [CDR2]). Together, these data indicate that engagement of self-antigen by autoreactive transitional B cells is sufficient to promote AID-mediated somatic hypermutation in the setting of BAFF excess.
|.||Sequences (n) .||≥1 Mutation (%) .||≥2 Mutations (%) .|
|Rag2-GFP.BAFF-Tg T1/T2 GFP+||118||1.69||0.0|
|Rag2-GFP.BAFF-Tg T1/T2 GFP−||192||18.75*,**||11.98*|
|.||Sequences (n) .||≥1 Mutation (%) .||≥2 Mutations (%) .|
|Rag2-GFP.BAFF-Tg T1/T2 GFP+||118||1.69||0.0|
|Rag2-GFP.BAFF-Tg T1/T2 GFP−||192||18.75*,**||11.98*|
Total number of sequences analyzed and H and L chain (κ and λ) mutation frequency in single-cell sorted B cells from the indicated subsets.
p < 0.0001, mutated sequences in Rag2-GFP.BAFF-Tg GFP− T1/T2 BCRs versus WT FM, WT MZ, and BAFF-Tg MZ cells.
p < 0.0001, mutated sequences in Rag2-GFP.BAFF-Tg GFP− T1/T2 BCRs versus Rag2-GFP.BAFF-Tg GFP+ BCRs.
Finally, to address whether this autoreactive transitional subset contributes to BAFF-Tg autoantibodies, we quantified ex vivo Ig production by sorted WT and BAFF-Tg B cell subsets. Although BAFF-Tg TACI+ transitional B cells and, to a lesser extent, MZ B cells produced IgM, only the BAFF-Tg TACI+ transitional cell subset spontaneously produced class-switched IgG. Further, addition of the TLR7 ligand, R848, markedly increased IgM and IgG by BAFF-Tg TACI+ transitional cells. Although TLR7 activation increased MZ IgM secretion, R848 exerted minimal impact on IgG formation by FM, MZ, or TACI− transitional cells (Fig. 4B, upper panels). Importantly, secreted Ab exhibited reactivity to the RNA-associated Ag Sm/RNP, with TACI+ transitional and MZ subsets from BAFF-Tg mice producing anti-Sm/RNP IgM. Remarkably, only BAFF-Tg transitional cells produced class-switched Sm/RNP IgG Ab ex vivo, and the TACI+ transitional subset was the dominant source for anti-Sm/RNP IgG, spontaneously and after R848 stimulation (Fig. 4B, lower panels). Thus, although the MZ is enriched for autoreactive specificities and expanded in BAFF-Tg mice (27, 28), TACI-expressing transitional cells from BAFF-Tg mice produce markedly greater levels of class-switched IgG autoantibodies ex vivo.
In summary, our results demonstrate a critical role for TACI in the pathogenesis of BAFF-mediated humoral autoimmunity and identify a distinct subpopulation of activated, cycling transitional B cells enriched for autoreactive BCR specificities and characterized by prominent sTACI expression. Excess BAFF was shown to rescue low-affinity, autoreactive B cell clones from deletion at the transitional stage, suggesting that BAFF promotes autoimmunity by skewing the mature B cell compartment toward autoreactivity (20, 29). Our combined in vivo and in vitro observations strongly support a complementary model wherein direct activation of transitional B cells generates a pool of IgG Ab-secreting cells that produce pathogenic autoantibodies in BAFF-Tg mice, highlighting the significant and surprising contribution of TACI+ transitional cells to BAFF-mediated autoimmunity. Although serum BAFF levels in BAFF-Tg mice likely exceed those of the majority of patients with SLE and other autoimmune disorders, we predict that our findings may inform disease pathogenesis in the subset of subjects characterized by the highest BAFF levels. In addition, dysregulated transitional B cell activation is likely to be relevant in other clinical scenarios, including autoimmune disease relapse after treatment with B cell–depletion therapies and de novo humoral autoimmunity following stem cell transplantation. Importantly, although our data focused on events leading to loss of B cell tolerance, the developmental program identified in this study is more likely to have evolved to resist infection. Because developing transitional cells exhibit a broad range of BCR specificities, including self-reactivity, polyreactivity, and foreign reactivity, we predict that direct activation of this unique subset may contribute to rapid, TACI-dependent, T cell-independent Ab responses against blood-borne or other systemic infections.
This work was supported by National Institutes of Health Grants R01HL075453, R01AI084457, R01AI071163, and DP3DK097672 (all to D.J.R.) and K08AI112993 (to S.W.J.). Additional support was provided by the Benaroya Family Gift Fund (to D.J.R.), the American College of Rheumatology/Rheumatology Research Foundation Rheumatology Scientist Development Award (to S.W.J.), and the Arnold Lee Smith Endowed Professorship for Research Faculty Development (to S.W.J.).
The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.
The online version of this article contains supplemental material.
Abbreviations used in this article:
activation-induced cytidine deaminase
B cell activating factor of the TNF family; BAFF-R, BAFF receptor
κ-deleting recombination excision circle
systemic lupus erythematosus
transmembrane activator and CAML interactor
The authors have no financial conflicts of interest.