Altered BCR and TLR signals promote enhanced positive selection of autoreactive transitional B cells in Wiskott-Aldrich syndrome

Abstract

Wiskott-Aldrich syndrome (WAS) is an X-linked immunodeficiency disorder frequently associated with systemic autoimmunity, including autoantibody-mediated cytopenias. WAS protein (WASp)-deficient B cells have increased B cell receptor (BCR) and Toll-like receptor (TLR) signaling, suggesting that these pathways might impact establishment of the mature, naive BCR repertoire. To directly investigate this possibility, we evaluated naive B cell specificity and composition in WASp-deficient mice and WAS subjects (n = 12). High-throughput sequencing and single-cell cloning analysis of the BCR repertoire revealed altered heavy chain usage and enrichment for low-affinity self-reactive specificities in murine marginal zone and human naive B cells. Although negative selection mechanisms including deletion, anergy, and receptor editing were relatively unperturbed, WASp-deficient transitional B cells showed enhanced proliferation in vivo mediated by antigen- and Myd88-dependent signals. Finally, using both BCR sequencing and cell surface analysis with a monoclonal antibody recognizing an intrinsically autoreactive heavy chain, we show enrichment in self-reactive cells specifically at the transitional to naive mature B cell stage in WAS subjects. Our combined data support a model wherein modest alterations in B cell-intrinsic, BCR, and TLR signals in WAS, and likely other autoimmune disorders, are sufficient to alter B cell tolerance via positive selection of self-reactive transitional B cells.

Figure 1.

Altered specificity of the naive B cell repertoire in WASp-deficient mice. (A) λ-LC usage in splenic B cell populations in 8–10-wk-old B6 (n = 6), Was^fl/fl × Mb-1^cre (n = 5), and Was^−/− (n = 7) mice assessed by flow cytometry. (B and C) Cloned WT and Was^−/− MZ B cell mAb reactivities toward self-antigens (dsDNA, high [PC-4]- and low [PC-14]-affinity phosphorylcholine, MDA-LDL, and Sm-RNP) via ELISA depicted using a pie chart (blue = reactive clones identified based on threshold of 0.5 OD value; gray = nonreactive clones), with percentages of reactive clones and total number of clones tested noted. MZ B cells were FACS sorted and gated based on B220⁺CD23^loCD1d^hiCD24^hiCD21^hi surface marker expression from splenocytes pooled from five to six WT or Was^−/− mice. (C) ELISA OD values of serial dilution curves of WT and Was^−/− MZ mAbs (100 ng/µl). (D) Proportion of low affinity (OD of 0.5–1.5) and high affinity (OD of 1.5–3) in reactive antibody clones to individual self-antigens. (E) Relative binding affinity displayed as AUC of reactive antibodies. (F) Selection of Id (M167)⁺ B cells in peripheral B cell subsets in 10–12-wk-old WT M167 Tg (n = 9) and Was^−/− M167 Tg (n = 9) mice. Error bars show SEM. Statistical analysis was performed using the Student’s t test: *, P < 0.05; **, P < 0.01; ***, P < 0.001. Data are representative of at least two experiments.

Figure 2.

High-throughput BCR heavy chain sequencing of splenic B cell subsets from WT, Was^−/−, and Was^fl/fl × Mb-1^cre mice. B cell populations were sorted (total of ∼0.5–10 × 10⁶ pooled cells/subset) from B6 WT, Was^−/−, and Was^fl/fl × Mb-1^cre mice using a minimum of five to six mice/genotype per experiment (three experiments total). RNA was isolated, sequenced, and analyzed using a 5′-RACE 454 platform (see Materials and methods). Data represent a mean of three experiments (15 mice per genotype). (A) Heavy chain variable (VH) gene family usage in bulk WT and Was^−/− B cells (∼10 × 10⁶ cells/sample). (B) VH gene family usage in sorted WT and Was^−/− B220⁺CD21^hiCD24^hiCD23^loCD1d^hi MZ B cells (∼0.5–10⁶ cells/sample). Additional sequence information is available in Table S1. Error bars show SEM. Statistical analysis was performed using the Student’s t test: *, P < 0.05; ***, P < 0.001. Data are representative of at least three experiments.

Figure 3.

Was^−/− B cells exhibit intact negative selection. (A–C) The HEL-Ig/mHel double Tg (DTg) chimeric transplant model was used to determine whether self-reactive Was^−/− B cells undergo deletion in the BM. Recipients were analyzed at 6 wk after transplant (n = 6–7 per group; data are representative of two independent experiments). (A and B) BM B cells were analyzed by FACS. (C) HEL-specific serum antibody levels. (D–F) The HEL-Ig/sHel DTg system was used to determine whether self-reactive Was^−/− B cells undergo efficient induction of anergy (n = 6–7 per group; data are representative of two independent experiments). (D and E) BM B cells were analyzed by FACS. (F) HEL-specific serum antibody levels. (G) Quantitative assay for RS rearrangement was used to estimate κ-LC rearrangement in BM and spleen B cell subsets of 6–8-wk-old WT or Was^−/− mice. Results are normalized to β-actin and presented as fold difference relative to WT κ⁺ FM B cells (n = 7 in each group except for Was^−/− fractions E [n = 6] and F [n = 5]). Error bars show SEM. Statistical analysis was performed using the Student’s t test: ***, P < 0.001. Data are representative of at least two independent experiments.

Figure 4.

Was^−/− B cells exhibit high levels of antigen-dependent clonal expansion. (A) 8–10-wk-old WT (n = 7), Was^−/− (n = 9), and Was^fl/fl × Mb-1^Cre (n = 6) mice were treated with BrdU in vivo for 24 h. BM and splenic B cell subsets were analyzed for BrdU incorporation via FACS. Data are representative of one of three experiments (WT, n = 3; Was^−/−, n = 3; Was^fl/fl × Mb-1^Cre, n = 3). (B) Cell cycle analysis of splenic B cell subsets via DAPI labeling in 8–10-wk-old WT (n = 7), Was^−/− (n = 9) and Was^fl/fl × Mb-1^Cre (n = 6) mice. (C) Serum BAFF levels in WT (n = 11), Was^−/− (n = 9), and Was^fl/fl × Mb-1^cre (n = 6) mice. (D) Representative data showing GFP staining of splenic T1 (left) and T2 (right) B cells in WT and Was^−/− Nur77 Tg mice. (E and F) Percentage of GFP^hi and GFP^lo T2 B cells (E) and MFI of GFP in B cell subsets (F) in WT and Was^−/− Nur77 Tg mice (n = 8/each). (G) Percentage of BrdU⁺ T2 GFP^hi and GFP^lo cells in WT (n = 7) and Was^−/− Nur77 Tg (n = 6) mice. (G and H) 8–10-wk-old WT M167 Tg (n = 5) and Was^−/− M167 Tg mice (n = 5) were treated with BrdU in vivo for 24 h. (H) Representative contour FACS plot showing BrdU incorporation in WT (top) versus Was^−/− (bottom) M167 (Id)⁺ T2 B cells. (I) Percentage of M167 Id⁺Ki-67⁺ and Id⁻Ki-67⁺ splenic T2 B cells in WT (n = 5) versus Was^−/− M167 (n = 6) Tg mice. Error bars show SEM. Statistical analysis was performed using the Student’s t test: *, P < 0.05; **, P < 0.01; ***, P < 0.001. Data are representative of at least two experiments.

Figure 5.

Antigen-specific selection of Was^−/− transitional B cells requires Myd88 signals. (A) Representative FACS analysis of Ki67 staining of splenic T1 (B220⁺CD24^hiCD21^lo) and T2 (B220⁺CD24^hiCD21^mid) B cells in WT, Was^−/−, and Was^−/−Myd88^−/− mice. (B) Percentage of Ki67⁺ B cells in WT (n = 12), Was^−/− (n = 13), and Was^−/−Myd88^−/− (n = 12) mice. (C) VH gene family usage in sorted WT, Was^−/−, and Was^−/−Myd88^−/− MZ B cells. Error bars show SEM. Statistical analysis was performed using the Student’s t test: **, P < 0.01. Data are representative of at least two experiments.

Figure 6.

WAS/XLT subjects exhibit an altered naive B cell repertoire enriched for self-reactive specificities. (A and B) BCRs were cloned from peripheral blood naive B cells from a 10-mo-old WAS subject (n = 54 mAbs) and two pediatric and one adult HC subjects (n = 87 antibodies). Percentages of mAb clones reactive to self-antigens are shown. Data are representative of one of two experiments. (B, top) Pie charts display frequency of ANA-IFA (antinuclear antibodies detected by IFA)–reactive clones. (bottom) Staining pattern of each IFA-reactive clone, defined as nuclear, cytoplasmic, or polyreactive (both nuclear and cytoplasmic), displayed according to overall percentages. (C) Human B cell subset gating and sorting strategy. (D–G) Sorted subsets from five HC and three WAS pediatric subjects were analyzed by Illumina high-throughput sequencing of the BCR heavy chain and combined to show mean VH family gene usage. (D) VH family usage of sorted transitional B cells (CD19⁺CD27⁺CD24^hiCD38^hi) displayed according to percentage of total unique clonotypes within the most abundant VH families; see Table S4 for full VH family usage. (E) VH family usage of naive B cells (CD19⁺CD27⁻). (F) VH family usage of IgM memory B cells (CD19⁺CD27⁺IgG⁻). (G) Percentage of VH4-34 within total unique, clonotypic sequences in naive B cell compartment. (H) Percentage of VH4-34 sequences in naive B cells in four HC (ages 23–29 yr) and four WAS adult subjects (ages 18–28 yr). Error bars show SEM. Statistical analysis was performed using the Student’s t test: *, P < 0.05; ***, P < 0.001. Data are representative of at least two experiments.

Figure 7.

WAS/XLT subjects exhibit increased selection for 9G4⁺ B cells specifically within the mature naive compartment. (A) Representative FACS histograms quantifying the percentage of 9G4⁺ peripheral blood transitional (left) and naive B cells (right) in age-matched HC and WAS subjects. (B) 9G4⁺ percentages of B cell subsets in HC (n = 5) and WAS subjects (n = 6). (C) Relative fold change in percentage of 9G4⁺ transitional versus naive B cells in WAS (n = 6) and HC subjects (n = 5). (D) Percentage of 9G4⁺ transitional versus naive B cells in HC (n = 4) and WAS adult subjects (n = 4). (E) Cumulative data showing the MFI of 9G4 staining in B cell subsets in HC (n = 6) and WAS pediatric subjects (n = 8). (F) Serum 9G4–specific IgG levels were determined using plasma collected from pediatric WAS (n = 8) and HC (n = 9) subjects. Titers are displayed as relative absorbance values, OD (450 nm). Error bars show SEM. Statistical analysis was performed using the Student’s t test: *, P < 0.05; **, P < 0.01. Data are representative of at least two experiments.