Critical roles of chronic BCR signaling in the differentiation of anergic B cells into age-associated B cells in aging and autoimmunity

Abstract

Age-associated B cells (ABCs) with autoreactive properties accumulate with age and expand prematurely in autoimmune diseases. However, the mechanisms behind ABC generation and maintenance remain poorly understood. We show that continuous B cell receptor (BCR) signaling is essential for ABC development from anergic B cells in aged and autoimmune mice. ABCs exhibit constitutive BCR activation, with surface BCRs being internalized. Notably, anergic B cells, but not nonautoreactive B cells, contributed to ABC formation in these models. Anergic B cells also showed a greater propensity for in vitro differentiation into ABCs, which was inhibited by the expression of the transcription factor Nr4a1. Bruton's tyrosine kinase (Btk), a key BCR signaling component, was constitutively activated in ABCs from aged and autoimmune mice as well as patients with lupus. Inhibiting Btk reduced ABC numbers and ameliorated the pathogenicity of lupus mice. Our findings reveal critical mechanisms underlying ABC development and offer previously unrecognized therapeutic insights for autoimmune diseases.

Fig. 1.. Constitutive activation of BCR signaling in ABCs.

(A) Ca²⁺-mobilization profiles in FO B (FOB) cells (AA4.1⁻CD19⁺B220^hi CD21^intD23⁺) and ABCs (CD19⁺B220^hiAA4.1⁻CD21^loCD23⁻CD11c⁺) from aged mice (n = 3). All values are plotted as the fluorescence ratio. Bars show fluorescence ratio, and dots represent individual data in the unstimulated state. (B) Representative flow cytometry plots (top) and mean fluorescence intensity (MFI) (bottom) of intracellular staining for isotype control, pBtk, pSyk, pPLCγ2, pAkt, and pErk in FO B cells and ABCs from aged mice (n = 3). (C) Confocal images of p65 nuclear translocation (colocalized with DAPI) in unstimulated FO B cells and ABCs from pooled aged mice (n = 4). Scale bars, 10 μm; proportion of cells with p65 nuclear translocation. (D) FACS-sorted FO B cells and ABCs from aged mice (n = 3) were immediately subjected to detergent-based nuclear isolation. The nuclear translocation of p65 or NFAT1 at an unstimulated state was assessed by flow cytometry. Data are the means ± SEM. P values are from two-tailed unpaired Student’s t test in [(A), (C), and (D)] and paired Student’s t test in (B). Data are representative of three independent experiments in [(A) to (D)]. In the experiment (C), each dot shows the results of three independent experiments.

Fig. 2.. ABCs show cellular internalization of BCR at a steady state.

(A) Representative flow cytometry plots of surface (Surface) or combination of surface and intracellular staining (Surface + Intra) of IgM and IgD in FO B cells and ABCs from aged mice (n = 5) and summary of MFIs of IgM and IgD from the left plot. (B) Confocal images and the percentage of IgM internalized (defined as the proportion of cells with IgM colocalized with LAMP-1) in FO B cells and ABCs from pooled spleens of aged mice (n = 4). Scale bars, 5 μm. Data are the means ± SEM. P values are from two-tailed unpaired Student’s t test in (B) and two-way ANOVA followed by Tukey’s multiple-comparisons test in (A) (ns, not significant). Data are representative of three independent experiments in [(A) and (B)]. In the experiment (B), each dot shows the results of three independent experiments.

Fig. 3.. Anergic B cells can preferentially differentiate into ABCs.

(A) Representative flow cytometry plots of splenic ABCs (CD21^loCD11c⁺) from young WT (n = 6), MD4 (n = 6), and MD4 × ML5 mice (n = 6) and aged (46 to 65 weeks) WT (n = 6), MD4 (n = 9), and MD4 × ML5 mice (n = 6) (left) and the percentage of ABCs in mature B cells (CD19⁺B220^hiAA4.1⁻) (WT) or mature IgMa⁺IgMb⁻ B cells (MD4 and MD4 × ML5) (right). (B) Representative flow cytometry plots of HEL-binding B cells in young WT mice or in ABCs and non-ABCs of aged MD4 × ML5 mice and the percentage of HEL-binding B cells in aged MD4 × ML5 ABCs and non-ABCs (n = 3). (C) Representative flow cytometry plots (left) showing ABC differentiation phenotypes from B cells of young MD4 (n = 3) and MD4 × ML5 mice (n = 3) after 3 days of the indicated stimulations and the percentage of CD11c⁺T-bet⁺ ABCs gated on IgMa⁺IgMb⁻ B cells (right). (D) GFP and surface IgM expression on FO B cells from Nr4a1-GFP mice (n = 5). (E) Representative flow cytometry plots of GFP and pSyk expression in FO B cells after anti-IgM stimulation and MFI ratio (stimulated/unstimulated cells) of pSyk in GFP⁻ and GFP^hi FO B cells. (F) Representative histogram or MFI of GFP (Nr4a1) in FO B cells and ABCs obtained from aged Nr4a1-GFP mice (n = 5). (G) Gating for GFP^hi and GFP⁻ FO B cells. (H) Representative flow cytometry plot of GFP^hi (red) and GFP⁻ (blue) FO B cells from Nr4a1-GFP mice (n = 5) after 3 days of IL-21, R848, and anti-IgM stimulation and the percentage of CD11c⁺T-bet⁺ ABCs among B cells. Data are the means ± SEM. P values are from two-tailed unpaired Student’s t test in [(F) and (H)], paired Student’s t test in (E), and two-way ANOVA followed by Tukey’s multiple-comparisons test in [(A) and (B)]. Data are pooled from three experiments in [(A), (F), and (H)] and are representative of three experiments in [(C) to (E)].

Fig. 4.. Nr4a1 inhibits ABC generation.

(A) Schematic for retroviral transduction of Nr4a1 in anergic B cells. GFP^hi FO B cells from Nr4a1-GFP mice (n = 3) were FACS sorted, prestimulated with R848 (24 hours), then retrovirally transduced with empty (mock) or Nr4a1 vector, and cultured with R848 (24 hours). On day 2, IL-21, anti-IgM, and BAFF were added, and cells were cultured 2 more days before analysis. (B) Representative histogram of Nur77 protein expression in mock-transduced (blue) or Nr4a1-transduced (red) B cells on day 4. Dashed line, isotype control. (C) Representative flow cytometry plots of CD11b/CD11c staining on mock- or Nr4a1-expressing mCherry⁺ B cells and percentage of CD11b⁺CD11c⁺ among mCherry⁺ B cells on day 4. (D) Representative CD11c/T-bet staining on mock- or Nr4a1-expressing mCherry⁺ B cells and percentage of CD11c⁺T-bet⁺ ABCs among mCherry⁺ B cells on day 4. (E) Heatmap (row z-scores, n = 3) of DEGs between naive, mock-transduced, and Nr4a1-transduced B cells (log₂FC > 2.0, FDR < 0.05). (F) Venn diagram of up-regulated genes in mock versus naive and repressed genes by Nr4a1, shared/unique among naive, mock-transduced, or Nr4a1-transduced B cells, and ABC–up-regulated genes from aged (GSE237540) or MRL/lpr mice (fig. S3C, GSE260576). (G) Volcano plot of DEGs between mock- and Nr4a1-transduced B cells among genes highly expressed in mock versus naive (log₂FC > 2.0, FDR < 0.05). Red, up-regulated; blue, down-regulated in Nr4a1-expressing B cells. (H) GSEA of ABC–up-regulated genes from aged (GSE237540) and MRL/lpr mice (GSE260576) in DEGs between mock- and Nr4a1-transduced B cells. NES, normalized enrichment score. (I) GO biological pathway of genes down-regulated by Nr4a1. TNF, tumor necrosis factor. (J) GSEA of enriched BCR signaling pathway and Toll-like receptor cascades from DEGs between mock- and Nr4a1-transduced B cells. Data are the means ± SEM. P values are from two-tailed unpaired Student’s t test in (C). Data are representative of three independent experiments in [(A) to (D)].

Fig. 5.. Btk is essential for ABC formation and maintenance in aged mice.

(A) Representative flow cytometry plots of ABCs (CD21^loCD11c⁺) from aged mice (n = 3). Mature B cells (CD19⁺B220^hiAA4.1⁻) were divided into pBtk^hi and pBtk^lo populations, and the percentage of ABCs in each is shown. Data are the representative of two independent experiments. (B) Splenic FO B cells and ABCs from aged mice (n = 4) were analyzed for Btk/pBtk expression via Phosflow (left). MFI ratio of pBtk to Btk (right). (C) Western blotting analysis of pBtk and Btk in whole-cell lysates of FACS-sorted FO B cells and ABCs from aged mice (n = 4) stimulated with anti-IgM (10 μg/ml) for 3 min. (D) Representative flow cytometry plots of splenic ABCs (CD21^loCD11c⁺) and staining of CD11b and T-bet in aged (45 to 60 weeks) WT (n = 6) and aged (45 to 60 weeks) Btk-deficient (Btk KO) mice (n = 5) and the percentage of ABCs (CD21^loCD11c⁺) among B cells in aged WT and Btk KO mice. Data are pooled from two experiments. (E) Aged mice were given vehicle or acalabrutinib for 21 days (n = 10 per group, pooled from two experiments). Flow cytometry plots; right, FO B cell and ABCs (CD21^loCD11c⁺) percentages among mature B cells. (F) Aged mice were given vehicle or acalabrutinib for 7 days. Each group was tested in two experiments (n = 4 or 3 per experiment). The percentage of Ki67⁺ FO B cells and ABCs was compared. Representative histogram; right, pooled data (n = 7). Data are the means ± SEM. P values are from Mann-Whitney U test in (D), two-tailed unpaired Student’s t test in [(C) and (E)], and two-way ANOVA followed by Tukey’s multiple-comparisons test in (F).

Fig. 6.. ABCs in lupus mice exhibit similar characteristics to ABCs in aged mice.

(A) Representative flow cytometry plots of splenic FO B cells and ABCs (CD21^loCD11c⁺) obtained from mice 14 days after the transfer of bm12 lymphocytes. (B) MFIs of pBtk, pSyk, pPLCγ2, and pAkt in FO B cells and ABCs from bm12-induced lupus mice (n = 5). (C) MFIs of IgM by surface staining only or surface and intracellular staining of FO B cells and ABCs from bm12-induced lupus mice (n = 5). (D) Expression of the indicated proteins in FO B cells from aged mice (n = 3) (yellow), ABCs from aged mice (green), FO B cells from bm12-induced lupus mice (n = 3) (blue), and ABCs from bm12-induced lupus mice (red). (E) Representative histogram and MFIs of GFP expression in FO B cells and ABCs obtained from Nr4a1-GFP mice (n = 3) with induced lupus by bm12 lymphocyte transfer. (F) Lymphocytes from WT mice or bm12 mice were transferred into young MD4 (n = 4) and MD4 × ML5 mice (n = 4), and mice were analyzed on day 14. (G) Representative flow cytometry plots of splenic ABCs 14 days after the transfer of WT or bm12 lymphocytes and the percentage of ABCs (CD21^loCD11c⁺) among IgMa⁺IgMb⁻ mature (CD19⁺B220^hiAA4.1⁻) B cells in each group. (H) Representative flow cytometry plots of HEL-binding B cells in young WT mice or ABCs and non-ABCs in bm12-transferred MD4 × ML5 mice and the percentage of HEL-binding B cells in ABCs and non-ABCs in bm12-transferred MD4 × ML5 mice (n = 3). Data are the means ± SEM. P values are from two-tailed paired Student’s t test in [(B) and (H)] and one-way ANOVA followed by Tukey’s multiple-comparisons test in [(C) and (G)]. Data are the pool [(F) and (G)] or representative [(A) to (E)] of two independent experiments.

Fig. 7.. Btk inhibition prevents ABC formation and ameliorates autoimmunity in lupus mice.

(A) Mice started receiving either acalabrutinib (n = 5, bm12 acalabrutinib) or vehicle (n = 5, bm12 vehicle) 7 days after the bm12 lymphocyte transfer and were analyzed on day 28. WT mice receiving WT lymphocytes and vehicle (n = 5, WT vehicle) were set as the negative control. (B) Percentage of ABCs (AA4.1⁻CD21^loCD11c⁺) among B cells (CD19⁺B220^hi) in PBMCs from each group in (A). (C) Representative flow cytometry plots and the percentage and absolute number of ABCs and FO B cells. The results show the pool (total n = 10 per group) of two independent experiments. (D) Representative picture and weight of spleens from the mice, as shown in (A). (E) Percentage of ABCs (CD21^loCD23⁻CD11b⁺CD11c⁺) among B cells collected from the liver or kidney of each group. (F) Autoantibodies against histone, dsDNA, and Sm/RNP in vehicle- or acalabrutinib-treated mice serum were tested by enzyme-linked immunosorbent assay (ELISA). OD, optical density. (G) Representative renal immunostaining of immunoglobulin deposition in kidneys, stained with anti-IgG, and bright-field imaging from vehicle-treated WT and vehicle- or acalabrutinib-treated bm12-induced lupus mice. Scale bars, 50 μm. (H) Representative H&E-stained histological liver images. Black arrows show aggregated leukocytes. Scale bars, 30 μm. Data are the means ± SEM. P values are from one-way [(C) and (E)] or two-way [(B) and (F)] ANOVA followed by Tukey’s multiple-comparisons test. Data are the representative [(A), (B), and (D) to (H)] of two independent experiments.

Fig. 8.. Btk is crucial for the generation and maintenance of human ABCs.

(A) Flow cytometry plots of CD21, CD11c, and T-bet expression on PBMCs of healthy donors (n = 3) or patients with SLE (n = 11). (B) Percentage of ABCs (CD21^loCD11c⁺) among CD19⁺ B cells. (C) pBtk expression of ABCs (red) and CD21⁺CD11c⁻ B cells (blue) from PBMCs of patients with SLE (n = 11). pBtk MFI in each population was compared. (D) Schematic for ABC generation. CD19⁺ B cells were purified from PBMCs of healthy donors (n = 3) and cultured with R848, IL-21, anti-Ig, and BAFF (37°C, 3 days). B cells were subdivided into pBtk^hi and pBtk⁻ population and the percentage of ABCs in each. Dashed line, pBtk isotype control. (E) Schematic for acalabrutinib treatment from day 0. CD19⁺ B cells were cultured with R848, IL-21, anti-Ig, and BAFF supplemented with acalabrutinib or DMSO for 3 days. Representative flow cytometry plots of CD11c and T-bet expression in each group on day 3 and the frequency of CD11c⁺T-bet⁺ ABCs among live B cells. (F) Schematic for delayed acalabrutinib treatment from day 3. CD19⁺ B cells were cultured with R848, IL-21, anti-Ig, and BAFF for 3 days. On day 3, acalabrutinib or DMSO was added into the medium and cultured for an additional 2 days. Representative flow cytometry plots of CD11c and T-bet expression on day 5. Bottom right, frequency of CD11c⁺T-bet⁺ ABCs among live B cells. Data are the means ± SEM. P values are from two-tailed paired (C) or unpaired [(E) and (F)] Student’s t test and Mann-Whitney U test (B). Data are pooled from three experiments [(B) and (C)] or are representative of three experiments [(D) to (F)]. Data are n = 3 technical replicates [(E) and (F)].