Deletion of IgG-switched autoreactive B cells and defects in Fas(lpr) lupus mice

Abstract

During a T cell-dependent Ab response, B cells undergo Ab class switching and V region hypermutation, with the latter process potentially rendering previously innocuous B cells autoreactive. Class switching and hypermutation are temporally and anatomically linked with both processes dependent on the enzyme, activation-induced deaminase, and occurring principally, but not exclusively, in germinal centers. To understand tolerance regulation at this stage, we generated a new transgenic mouse model expressing a membrane-tethered gamma2a-reactive superantigen (gamma2a-macroself Ag) and assessed the fate of emerging IgG2a-expressing B cells that have, following class switch, acquired self-reactivity of the Ag receptor to the macroself-Ag. In normal mice, self-reactive IgG2a-switched B cells were deleted, leading to the selective absence of IgG2a memory responses. These findings identify a novel negative selection mechanism for deleting mature B cells that acquire reactivity to self-Ag. This process was only partly dependent on the Bcl-2 pathway, but markedly inefficient in MRL-Fas(lpr) lupus mice, suggesting that defective apoptosis of isotype-switched autoreactive B cells is central to Fas mutation-associated systemic autoimmunity.

Figure 1

Design, in vitro testing and generation of γ2a-macroself Ag Tg mice (pURF-Tg). (A) Schematic representation of anti-mouse Igγ2a membrane-bound macroself Ag. A single chain Fv generated from the heavy and light chain variable genes of the 20.8.3 hybridoma is linked to the hinge and membrane proximal domains of rat IgG1 followed by transmembrane and cytoplasmic tail region (Tm/Cy) of H-2K^b gene. (B) Schematic representation of the DNA construct encoding the γ2a^a-macroself Ag used for transfection and microinjection to generate Tg mice. Introns are represented as thin lines. (G₄S)₃ refers to linker codons in one letter amino acid code: GGGGSGGGGSGGGGS. Li stands for the Ig light chain leader exon and first intron. The human ubiquitin C promoter controls the γ2a-macroself Ag transcription. (C) Flow cytometry analysis of transiently transfected HEK 293T cell line with pURF Tg and EGFP-containing vectors. GFP⁺ gated cells were analyzed. (Left), Staining with an anti-rat IgG1 monoclonal Ab compared with empty vector transfected cells. (Right) Binding specificity of the macroself Ag to mouse IgG2a^a (bold line) and IgG2a^b (thin line). γ2a-macroself Ag transfected cells were incubated first with IgG2a^a or IgG2a^b mouse monoclonal Abs followed by a biotin-conjugated anti-mouse IgG2a,b step and binding was revealed with streptavidin conjugate. (D) Flow cytometry analysis of γ2a-macroself Ag expression in bone marrow, spleen and lymph nodes of pURF Tg mice detected with an anti-rat IgG1 monoclonal Ab (thin line, non-Tg cells; bold line, pURF Tg cells).

Figure 2

Absence of IgG2a^a Ab response in IgH^a/b γ2a-macroself Ag Tg mice after priming and recall immunization with NP-KLH in RIBI adjuvant. (A) Mean±SD of total IgG2a^a, IgG2a^b, IgG1^a, IgG2a and IgG1 Ab titers measured by ELISA assays of serum derived from B6 (open lozenges), B6-Igh^a congenic (grey lozenges), IgH^a/b littermate control mice (non-Tg, open squares) and IgH^a/b pURF-Tg mice (black circles) after priming and recall immunization. (B) Mean±SD of NP₇-specific-IgG2a^a, -IgG2a^b, -IgG1^a, -IgG2a and -IgG1 Ab titers measured by ELISA assays on serum derived from IgH^a/b pURF-Tg (black circles) and littermate control mice (open squares) as indicated. The average concentrations of serum Ab and days after priming and recall immunization are indicated on the y- (logarithmic scale) and x-axis, respectively. Three to five mice were analyzed per group at each time point.

Figure 3

Absence of IgG2a^a memory B cell formation in IgH^a/b γ2a-macroself Ag Tg mice. (A) Flow cytometry analysis of spleen cells derived from pURF-Tg and non-Tg littermate control mice 7d post-priming and 5d after recall immunization with NP-KLH in RIBI adjuvant. NP-specific B220^high/CD138⁻ and B220^low/CD138⁺ cells were visualized by successive gating on a broad forward scatter versus side scatter window to include B cell blasts, then on PI⁻, Dump⁻ (CD4⁻, CD8⁻, IgD⁻) to exclude dead cells, T cells and naive B cells. NP-binding IgG1-positive cells were visualized by gating on B220^high cells. The histograms on the right represent the mean numbers of NP-specific and NP-specific B220^high/IgG1⁺ cells in non-Tg (grey bars) and pURF-Tg mice (black bars). (B) Flow cytometry analysis of NP-specific, B220^high/IgG2a^a+ and B220^high/IgG2a^b+ cells in pURF-Tg and non-Tg mice after priming and recall immunization. The histograms on the bottom represent the average numbers of NP-specific IgG2a^b-, IgG2a^a-positive cells and the IgG2a^b:IgG2a^a cell ratio in littermate control (grey bars) and pURF-Tg (black bars) mice. The percentage of positive cells (± SEM) is indicated in each gate. Three mice were analyzed per group at each time point except for the non-Tg mice after Ag recall (n=2).

Figure 4

Absence of cytoplasmic IgG2a^a-expressing B cells after Ag priming in γ2a-macroself Ag Tg mice. (A) Flow cytometry analysis of B220⁺, cytoplasmic IgG2a^a, IgG2a^b, IgG1 and IgG2b expressing spleen B cells identified by successive gating on side scatter versus forward scatter, Dump⁻ (CD8⁻, CD4⁻, F4/80⁻, Gr1⁻, IgM⁻, IgD⁻) and cytoplasmic Igκ⁺ (cIgk) cells, 14 days after priming with NP-KLH in RIBI adjuvant. The frequency of positive cells, rounded to the nearest 1%, is indicated next to each gate. (B) Histograms representing the mean±SE percentages and numbers of Dump⁻, cIgκ⁺, cIgG2a^a+; cIgG2a^b+; cIgG1⁺ and cIgG2b⁺ cells in littermate control (open bars) and pURF Tg (dark bars) mice. The numbers of mice analyzed are indicated in parenthesis.

Figure 5

Quantification of V_H-γ2a transcripts and molecular parameters of γ2a CSR in pURF Tg and littermate control mice (non-Tg) after Ag priming. (A) RNA samples purified from total spleen cells were subjected to reverse transcription and PCR reaction. Amplification products were quantified by Southern blot using an IgG2a specific probe. RNA samples without reverse transcription were used as negative control (“-“ labeled lines). Four-fold serial template dilutions were tested. (Top row) PCR detection of V_H-γ2a (V_H-CH1-hinge) transcripts. (Middle row) Detection of post switched germline transcripts driven by the Iμ promoter (Iμ-γ2a post-switch). (Third row) Detection of Iγ2a germline transcripts driven by the Iγ2a promoter (Iγ2a-γ2a germline). (Bottom row) Actin PCR is used as a DNA loading control. Representative blots are shown. Numbers above blots indicate days after Ag priming. Blots were cropped for clarity. No additional bands were detected in the full-length blots. (B) Relative quantification of V_H-γ2a transcripts as detected in A. (C) Quantification of γ2a (CH1-hinge) transcripts by quantitative real time PCR in RNA isolated from spleen cells of the indicated mouse types. A minimum of three mice was analyzed per group at each time point. (D) Relative quantification of Iμ-γ2a post-switched and (E) Iγ2a-γ2a germline transcripts as detected in A from three experiments. The results are representative of three to four independent experiments.

Figure 6

Flow cytometry and NP-specific ELISPOT analyses of tolerance induction in radiation chimeras using γ2a-macroself Ag Tg hosts. IgH^a/b/CD45.2, wild type or Bcl2 Tg spleen cells were used to reconstitute sub-lethally irradiated IgH^b/CD45.1 pURF-Tg or non-Tg control recipients. Spleen donor cells are indicated on the left side of each arrow. Mice were analyzed 14 day post reconstitution and priming with NP-KLH in RIBI adjuvant. (A) Analysis of Igκ⁺ spleen cells for surface expression of B220 and IgG2a^a (top plots) or IgG2a^b (bottom plots). Plasma cells were identified as falling in the lower right boxes, as indicated. The percentage in each gate, rounded to the nearest 1%, is indicated in the upper right corner of each plot. (B) Average numbers (± SE) of IgG2a^a- and IgG2a^b-expressing B cells in the spleen of the indicated radiation chimeras. (C) Mean±SE numbers of IgG2a^a and IgG2a^b NP-specific Ab-forming cells (AFC) per 10⁶ spleen cells counted by ELISPOT assay. The numbers of mice analyzed per group are indicated in parentheses.

Figure 7

MRL-Fas^lpr γ2a-Tg mice fail to delete IgG2a-expressing B cells after NP-KLH priming. Six week-old MRL-Fas^lpr and MRL-Fas^lpr pURF-Tg mice were analyzed by flow cytometry for cytoplasmic IgG2a and IgG2b expression two weeks after immunization with NP-KLH in RIBI adjuvant. Age-matched B6-Igh^a pURF-Tg and littermates (non-Tg) were used as controls. (A) Spleen cells from the indicated mice were analyzed by flow cytometry for B220⁺, cytoplasmic IgG2a, and IgG2b-expressing spleen B cells as described in Fig. 4 except that an anti-TcRβ was added in the Dump channel to counter select double-negative T-cells. Top panels are Dump⁻ cells. The percentage in each gate, rounded to the nearest 1%, is indicated. (B) Mean±SE of total, NP-specific and anti-chromatin IgG2a Abs in the serum of naïve and immunized littermate B6-Igh^a and MRL-Fas^lpr controls and pURF-Tg mice, 14 days after Ag priming. Three to five mice were analyzed in each group. (C) Allotype-specific serum IgG2a in (B6-Fas^lprxMRL-Fas^lpr)F1 pURF mice born to IgG2a^b allotype (B6-Fas^lpr) mothers. IgG2a levels below the detection limit are indicated as <1 μg/ml. *P<0.0001 between nTg and Tg mice. (D) Allotype-specific serum IgG2a and IgG1 in Fas-deficient B6-Igh^a pURF Tg mice. Littermates for this experiment were (B6-Igh^a Fas^lpr/+ × B6-Igh^a Fas^lpr/+ pURF Tg)F1 mice. IgG2a concentrations <1 μg/ml were below detection. n=5-9 mice, except 35 d Tg lpr/+ where n=2. *P<0.05.