B Cell Development and T-Dependent Antibody Response Are Regulated by p38γ and p38δ

Abstract

p38MAP kinase (MAPK) signal transduction pathways are important regulators of inflammation and the immune response; their involvement in immune cell development and function is still largely unknown. Here we analysed the role of the p38 MAPK isoforms p38γ and p38δ in B cell differentiation in bone marrow (BM) and spleen, using mice lacking p38γ and p38δ, or conditional knockout mice that lack both p38γ and p38δ specifically in the B cell compartment. We found that the B cell differentiation programme in the BM was not affected in p38γ/δ-deficient mice. Moreover, these mice had reduced numbers of peripheral B cells as well as altered marginal zone B cell differentiation in the spleen. Expression of co-stimulatory proteins and activation markers in p38γ/δ-deficient B cells are diminished in response to B cell receptor (BCR) and CD40 stimulation; p38γ and p38δ were necessary for B cell proliferation induced by BCR and CD40 but not by TLR4 signaling. Furthermore, p38γ/δ-null mice produced significantly lower antibody responses to T-dependent antigens. Our results identify unreported functions for p38γ and p38δ in B cells and in the T-dependent humoral response; and show that the combined activity of these kinases is needed for peripheral B cell differentiation and function.

Keywords: B cell; lymphocyte; p38MAPK; p38γ; p38δ; spleen.

FIGURE 1

Analysis of B cell populations in BM from WT, p38γ/δ^–/–, p38γ/δ^f/f and CD19-Cre^KI/+-p38γ/δ^f/f mice. Cells were isolated from femoral and tibiae. (A) Total cell number in BM of the indicated genotypes. (B) Cell suspensions were stained with the indicated antibodies and analysed by flow cytometry. Representative flow cytometry profiles and analysis are shown. (C) Frequency and (D) total cell number of Pro-B cells (CD19⁺ IgM^– B220⁺ CD43⁺), Pre-B cells (CD19⁺ IgM^– B220⁺ CD25⁺), immature (CD19⁺ B220⁺ IgM⁺IgD^–) and mature (CD19⁺ B220⁺ IgM⁺ IgD⁺) B cells in adult mouse BM. Percentage of each B-cell population was determined relative to CD19⁺ cell. Each dot in (A,C,D) represents a single mouse (n = 9–14). ns, not significant.

FIGURE 2

Analysis of B cell populations in the spleen. (A) Total cell number in spleens of adult mice of the indicated genotypes. Each dot represents a single mouse (n = 12–18). (B) B cell frequency and (C) total number in spleens of adult mice of the indicated genotypes; B cells were gated as CD19⁺ cells. Each dot represents a single mouse (n = 12–18). (D) Purified B cells were cultured in absence or presence of BAFF (0.1 μg/ml) and analysed for cell survival at the indicated time points by flow cytometry. B cell survival frequencies for each genotype are shown. Data are the mean ± SD of n = 3 mice per genotype. (E) Representative dot plots for CD21 and CD23 expression in splenocytes from mice of the specified genotypes. Frequencies of gated follicular (FO) B cells (CD21⁺CD23⁺) and marginal zone (MZ) B cells (CD21^hiCD23^–) (gates) are indicated. (F,H) Frequency and (G,I) total cell number of FO and MZ B cells in adult mouse spleens. Each dot represents a single mouse (n = 9-14). (J) Representative photomicrographs of hematoxylin/eosin-stained spleen sections from adult WT and p38γ/δ^–/– mice. (K) Fluorescence images of IgD (green), IgM (blue) and MOMA-1 (red) in representative tissue sections from spleens of WT and p38γ/δ^–/– mice. Right panels show a higher magnification of a B cell follicle and surrounding MZ. ns, not significant, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, p-value is also indicated in number.

FIGURE 3

Migration and immune synapse formation of p38γ/δ deficient B cells. (A) B cells were cultured in absence (none) or presence of CXCL13 or CCL21 gradients, in Boyden chambers, during 3 h. Migratory B cells were collected and counted by flow cytometry. Migration frequencies for each condition and genotype are shown; data are the mean ± SD of n = 3 mice per genotype. (B,C) WT and p38γ/δ deficient B cells were settled on ICAM-1-containing planar lipid bilayers uncoated or CXCL13-coated, and monitored for cell adhesion and motility by real-time microscopy. (B) Frequencies of B cell migration (left panel) and adhesion (right panel); data are from two independent experiments (n = 4 mice per genotype). Dashed line indicates mean value of basal migration/adhesion (in absence of CXCL13). (C) Values of mean speed (left panel) and straightness index (right panel) for motile B cells; data from representative experiment are shown (n = 4 mice per genotype). (D–F) WT and p38γ/δ deficient B cells were settled on ICAM-1-containing planar lipid bilayers in the presence of tethered su-Ag at the specified densities, and monitored for IS formation. (D) DIC, IRM and fluorescence su-Ag images at the contact plane of representative IS-forming B cells. Bar, 2 μm. Data acquisition was performed at medium quality (512 × 512 pixels). (E) Frequencies of B cell adhesion (upper panel) and su-Ag central cluster (cSMAC; bottom panel). (F) Values of contact area (IS area; in μm²; left) and total su-Ag fluorescence at the IS (in arbitrary units, AU; right) for B cells with established IS in each case. Data from a representative experiment are shown in (E,F) (n = 3 mice per genotype). Each dot in (B,E) represents a single image field, and in (C,F) a single cell.

FIGURE 4

B cell activation in the absence of p38γ and p38δ. (A–C) Splenocytes from WT (black) or p38γ/δ^–/– (white) mice were cultured alone (none) or with anti-BCR (1 and 10 μg/ml) for 20 h. CD19⁺-gated B cells were analysed for expression of CD69 and CD86 activation markers by flow cytometry. (A) Representative profiles of surface CD69 expression (filled histograms) in B cells stimulated with the indicated ligands; grey line, isotype control. (B) Frequency of activated B cells, gated as CD69⁺CD86⁺, in the total B cell population for each condition. (C) CD69 mean fluorescence intensity (MFI) values in CD19⁺-gated B cells. In (B,C) each dot represents a single mouse (n = 7–8). (D,E) Violet tracer-labelled splenocytes of the specified mouse genotypes were cultured alone (none) or with anti-BCR (1 and 10 μg/ml) plus IL-4 (20 ng/ml) for 96 h. CD19⁺-gated B cells were analysed for violet tracer signal dilution by flow cytometry. (D) Representative violet tracer profiles of WT (filled histograms) and p38 isoform-deficient (black line) B cells stimulated as above; grey line, violet tracer profile of unstimulated B cells after 96 h in culture. (E) Proliferation index of B cells in each condition; values obtained with FlowJo software. Each dot represents a single mouse (n = 4-7). (F,G) Splenocytes from WT or p38γ/δ^–/– (white) mice were cultured alone or in presence of anti-CD40 (1 μg/ml) for 20 h. (F) Representative profiles of surface CD69, CD86, and CD25 expression (filled histograms) in CD40-stimulated B cells. Grey line indicates unstimulated B cells at 20 h. (G) Frequency of CD69⁺CD86⁺ B cells (top, right) and of CD25⁺ B cells (bottom, right), and MFI values of CD69, CD86 and CD25 (left panels) for CD19⁺-gated B cells in each condition (none, unstimulated; CD40, anti-CD40). (H) Splenocytes were cultured alone (none) or in presence of anti-CD40 (1 μg/ml) plus IL-4 (20 ng/ml) for 96 h. Left, representative violet tracer profiles of WT and p38γ/δ deficient B cells (black line); filled histogram, unstimulated B cells at 96 h. Right, frequencies of proliferating B cells (with diluted violet tracer level) in each condition. Each dot in (G,H) represents a single mouse (n = 5-6). (I,J) Purified B cells from WT or p38γ/δ^–/– mice were unstimulated or stimulated with (I) anti-CD40 (1 μg/ml) or (J) anti-BCR (1 μg/ml) for the times indicated and cell lysates immunoblotted with the indicated antibodies using the Odyssey infrared imaging system. Figure shows representative immunoblots from three independent experiments. ns, not significant, *p < 0.05, **p ≤ 0.01, ***p ≤ 0.001.

FIGURE 5

Antibody response to T-dependent antigen in p38γ- and p38δ-deficient mice. (A) Adult WT and p38γ/δ^–/– mice were immunised i.p. with NP-KLH in alum (T-dependent). Serum samples were analysed by ELISA for TNP-specific antibodies of different isotypes (IgM, IgG1, IgG2a, IgG2b, IgG3, IgA) at the indicated times. O.D., optical density. (B) Serum samples from NP-KLH immunised mice were analysed for the presence of NP-specific high affinity antibodies, using NP₃ as substrate compared to NP₃₀. Values of the ratio between NP₃-ELISA and NP₃₀-ELISA for the indicated IgG isotypes in each time point and for each mouse are shown. Each dot in (A,B), represents a mouse (n = 4 mice/group); ns, not significant, *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001. (C–F) Adult WT and p38γ/δ^–/– mice were immunised i.p. with NP-KLH in alum. Spleens and BMs were analysed at day seven post-immunisation for the presence of GC B cells (B220⁺ IgD^neg/low GL7⁺ CD95⁺; CXCR4^high for centroblast and CXCR4^+/low for centrocytes), plasma cells (PC; CD138^high B220^low, IgM⁺ or IgG1⁺), and T follicular helper cells (Tfh; CD3⁺ CD4⁺ CXCR5⁺ PD1⁺) by flow cytometry (gating strategy shown in Supplementary Figures S6A,C). Representative dot plots of (C) GL7/CD95 for B220⁺ IgD^neg/low-gated B cells (top panels), and FSC-A/CXCR4 for GL7⁺ CD95⁺-gated B cells (bottom panels); (D) CD138/B220 for total splenocytes (top panels), and IgG1/IgM for CD138^high B220^low-gated PC (bottom panels); (E) as in (D) but for total BM; and (F) CXCR5/PD-1 for CD3⁺CD4⁺-gated T cells, are shown. Regions used for GC, PC and Tfh analysis are indicated in each case. Frequencies of the analysed populations are shown (right panels); each dot is a mouse. Data are from one experiment (n = 5 mice per group). (G,H) B cells isolated from spleens of WT and p38γ/δ^–/– mice were cultured with (G) anti-CD40 mAb plus IL-4 (T-dependent stimulation) or (H) LPS (T-independent stimulation) for 96 h. Production of different Ig isotypes (IgM, IgG1, IgG2a, IgG2b, IgG3) was measured by ELISA, using the appropriate supernatant dilutions; O.D. values obtained with supernatants from none-stimulated B cells were subtracted. Data show mean ± SD for four experiments (n = 8).