Batf coordinates multiple aspects of B and T cell function required for normal antibody responses

Abstract

Batf belongs to the activator protein 1 superfamily of basic leucine zipper transcription factors that includes Fos, Jun, and Atf proteins. Batf is expressed in mouse T and B lymphocytes, although the importance of Batf to the function of these lineages has not been fully investigated. We generated mice (Batf(DeltaZ/DeltaZ)) in which Batf protein is not produced. Batf(DeltaZ/DeltaZ) mice contain normal numbers of B cells but show reduced numbers of peripheral CD4(+) T cells. Analysis of CD4(+) T helper (Th) cell subsets in Batf(DeltaZ/DeltaZ) mice demonstrated that Batf is required for the development of functional Th type 17 (Th17), Th2, and follicular Th (Tfh) cells. In response to antigen immunization, germinal centers were absent in Batf(DeltaZ/DeltaZ) mice and the maturation of Ig-secreting B cells was impaired. Although adoptive transfer experiments confirmed that this B cell phenotype can be driven by defects in the Batf(DeltaZ/DeltaZ) CD4(+) T cell compartment, stimulation of Batf(DeltaZ/DeltaZ) B cells in vitro, or by a T cell-independent antigen in vivo, resulted in proliferation but not class-switch recombination. We conclude that loss of Batf disrupts multiple components of the lymphocyte communication network that are required for a robust immune response.

Figure 1.

Profile of T and B cells in Batf^ΔZ/ΔZ mice. (A) Schematic of Batf KI and Batf exon 3–deleted (ΔZ) alleles. Batf exons 1–3 are numbered. Filled triangles indicate loxP sites. Arrows indicate genotyping primers. Numbered arrows indicate primers used to identify targeted ES clones. S, SpeI; X, XbaI; E, EcoRI; B, BamHI; Xh, XhoI; P, PstI. (B) A representative DNA blot from five independent experiments detects homozygous deletion of Batf exon 3 in Batf^ΔZ/ΔZ mice. (C) Splenocytes from Batf^KI/KI and Batf^ΔZ/ΔZ mice were stimulated with 2.5 ng/ml PMA and 125 ng/ml ionomycin (P + I) for 6 h. Cell extracts were immunoblotted using anti-HA mAb to detect Batf and anti–Hsp-90 mAb as a control for sample loading. A representative blot from three independent experiments is shown. (D) Cells from spleen and PP of Batf^+/+ and Batf^ΔZ/ΔZ mice were analyzed by flow cytometry using anti-CD4 and anti-CD8 mAb to detect T cells and anti-CD19 mAb to detect B cells. The mean percentage of total cells of each set is shown on the left, and the mean cell number per spleen or per mouse PP is shown on the right. Data are means of three experiments performed with three mice per group (n = 9). Error bars indicate SE. *, P < 0.05.

Figure 2.

Th17 and Th2 differentiation is impaired in Batf^ΔZ/ΔZ mice. (A) Cells from spleens and PP of Batf^ΔZ/ΔZ and Batf^+/+ mice were surface stained using anti-CD4 mAb. After stimulation with 50 ng/ml P + 1 µg/ml I with 1 µg/ml monensin for 4 h, intracellular mAb staining was used detect IFN-γ–, IL-4–, and IL-17–expressing cells by flow cytometry. Representative flow plots are shown, along with the mean percentage of CD4⁺ cells positive for each cytokine, calculated from two experiments with three mice per group (n = 6). Error bars indicate SE. *, P < 0.05. (B) CD4⁺ cells from spleens of Batf^ΔZ/ΔZ and Batf^+/+ mice (n = 3) were stimulated with anti-CD3ε mAb for 48 h. RNA was isolated, converted to cDNA, and assayed in duplicate for the indicated transcripts by qPCR. The mean relative mRNA expression is shown. Error bars indicate SE. *, P < 0.05. (C and D) Naive T cells from Batf^ΔZ/ΔZ and Batf^+/+ spleens were cultured for 5–6 d with 5 µg/ml anti-CD3ε, 2 µg/ml anti-CD28, and 20 U/ml rhIL-2 and the following skewing conditions: Th2, 20 ng/ml rmIL-4, 10 µg/ml anti–IFN-γ mAb, and 10 µg/ml anti-IL-12 mAb; Th17 was described in Wang et al. (2009). Cytokine gene expression was detected by qPCR as in B. Data shown were averaged from five (Th2) and three (Th17) independent RNA preps assayed in duplicate. Error bars indicate SE. *, P < 0.05.

Figure 3.

Dysgammaglobulinemia in Batf^ΔZ/ΔZ mice. (A) ELISA was performed to measure the indicated Ig in sera of Batf^+/+ and Batf^ΔZ/ΔZ mice. Shown are mean results from four mice per group (n = 4) assayed in duplicate. Error bars indicate SE. *, P < 0.05. (B) Sera from Batf^+/+ and Batf^ΔZ/ΔZ mice, immunized with sRBC or mock injected with PBS, were used in ELISA to detect sRBC-specific IgM or IgG. Mean results from one (n = 3) of two experiments are shown. Error bars indicate SE. *, P < 0.05. (C) Spleen sections from mice in B (n = 3 for each genotype) were incubated with primary anti-IgG1 and anti-IgG2c Abs. Complexes were detected using biotinylated secondary Abs and Vectastain ABC reagent. Shown are representative images, counterstained with H (no E) and photographed at 40×. f, follicle. Bars, 50 µm. (D) Spleen sections from mice in B (n = 3 for each genotype) were stained with H + E (left) and photographed at 4× (bars, 250 µm) or 20× (insets; bars, 50 µm). GCs (arrows) were detected on additional sections (right) using biotinylated PNA and Vectastain ABC reagent, counterstained with H (no E; 20×; bars, 50 µm) or using biotinylated PNA, anti–mouse B220, DAPI, and fluorescently labeled secondary mAbs (60×; bars, 50 µm).

Figure 4.

Dysfunctional Tfh cells in Batf^ΔZ/ΔZ mice. (A) Cells from spleen and PP of Batf^ΔZ/ΔZ and Batf^+/+ mice were stained with anti-CD4, anti-CD44, anti-CD62L, and anti-CXCR5 mAb (or isotype control) and Tfh cells detected by flow cytometry. Representative plots (n = 9) showing CXCR5 expression after gating on CD44⁺CD4⁺CD62L⁺ or CD44⁺CD4⁺CD62L⁻ cells are presented. (B) Datasets from A (n = 9) are plotted as the mean percentage of CD62L⁻CXCR5⁺ of total CD4⁺ (left) or number of CXCR5⁺ cells per organ or per total mouse PP (right) with SE. *, P < 0.05. (C) 5 × 10⁵ cells from PP were allowed to migrate to rmCXCL13 for 3 h. Migrated cells were stained with anti-CD4 and anti-CD62L mAbs and analyzed by flow cytometry. Migration is the number of migrating cells (+ ligand) minus the number migrating cells (no ligand) expressed as a percentage of CD4⁺CD62L⁻ or CD4⁺CD62L⁺ cells in the starting population. Mean results from three experiments are shown (n = 3). Error bars indicate SE. *, P < 0.05.

Figure 5.

Limited response of Batf^ΔZ/ΔZ CD4⁺ T cells to T cell–dependent antigen. (A) CD4⁺ T cells from Batf^+/+ or Batf^ΔZ/ΔZ mice were injected into T cell–deficient mice, which were immunized with 5 × 10⁸ sRBC. After 8 d, Tfh cells in spleen and lymph nodes were quantified as in Fig. 4 A. Representative plots (n = 3) are shown. (B) Mean number of CXCR5⁺ cells from A per organ or per mouse lymph node (LN) is shown (n = 3). Error bars indicate SE. *, P < 0.05. (C) ELISA using sera isolated from mice in A to detect sRBC-specific IgM and IgG1. Mean and SE are shown (n = 3). *, P < 0.05.

Figure 6.

Batf^ΔZ/ΔZ B cells do not undergo CSR. (A) B cells purified from Batf^+/+ and Batf^ΔZ/ΔZ mice were cultured in media, or media supplemented with 20 µg/ml LPS, with or without 20 ng/ml IL-4. After 24 h, DNA synthesis was quantified by BrdU labeling for 16 h. Shown is mean BrdU incorporation, relative to Batf^+/+ or Batf^ΔZ/ΔZ cells in media (set to 1.0), from three experiments (n = 3) performed in triplicate. Error bars indicate SE. (B) B cells cultured as in A were assayed for surface Ig expression by flow cytometry (top) and for secreted Ig by ELISA (bottom). The mean and SE were calculated from three experiments (n = 3). N, not detected. (C) RNA from cells in B was assayed for the indicated transcripts using qPCR. Data are averaged from three experiments (n = 3) performed in triplicate and expressed relative to the Batf^+/+ media control (set to 1.0). Error bars indicate SE. N, not detected. (D) Sera from Batf^+/+ and Batf^ΔZ/ΔZ mice immunized with TNP-LPS or PBS were analyzed by ELISA for TNP-specific IgG1. Mean results from three mice per group (n = 3) are plotted. Error bars indicate SE. (E) Spleen sections from mice in D (n = 3 for each genotype) were stained as in Fig. 3 C to detect IgG1- and IgG2c-producing cells. Representative images are shown (20×). f, follicle. Bars, 50 µm.