Qualitative regulation of B cell antigen receptor signaling by CD19: selective requirement for PI3-kinase activation, inositol-1,4,5-trisphosphate production and Ca2+ mobilization

Abstract

Genetic ablation of the B cell surface glycoprotein CD19 severely impairs the humoral immune response. This requirement is thought to reflect a critical role of CD19 in signal transduction that occurs upon antigen C3dg coligation of antigen receptors with CD19 containing type 2 complement receptors (CR2). Here we show that CD19 plays a key accessory role in B cell antigen receptor signaling independent of CR2 coligation and define molecular circuitry by which this function is mediated. While CD19 is not required for antigen-mediated activation of receptor proximal tyrosines kinases, it is critical for activation of phosphatidylinositol 3-kinase (PI3-kinase). PI3-Kinase activation is dependent on phosphorylation of CD19 Y484 and Y515. Antigen-induced CD19-dependent PI3-kinase activation is required for normal phosphoinositide hydrolysis and Ca2+ mobilization responses. Thus, CD19 functions as a B cell antigen receptor accessory molecule that modifies antigen receptor signaling in a qualitative manner.

Figure 1

Expression levels of surface markers on the plasmacytoma cell lines. Flow cytometric analysis was performed on the J558Lμm3CD45⁺ (dotted lines), J558Lμm3CD45⁺CD19⁺ (thick lines), and J558L μm3CD45⁺ CD19⁺ (Y⁴⁸⁴F,Y⁵¹⁵F) (thin lines) cell lines. Shown is staining with secondary alone (A), biotinylated b-7-6 anti-IgM (B), biotinylated I32.5 anti-CD45 (C), or biotinylated HD37 anti-hCD19 (D) followed by PE-conjugated streptavidin.

Figure 2

The CD19-negative and -positive J558Lμm3CD45⁺ plasmacytoma variants mobilize [Ca²⁺]_i to different extent after antigen stimulation. J558Lμm3CD45⁺CD19⁻ and J558Lμm3CD45⁺CD19⁺ cells were loaded with indo-1 AM, and analysis of [Ca²⁺]_i initiated before antigen stimulation (250 ng NP₉BSA/ 10⁶cells/ml). Mean [Ca²⁺]_i (top) and cells responding (bottom) after antigen stimulation of J558Lμm3CD45⁺CD19⁻ and J558Lμm3CD45⁺CD19⁺. The analysis was conducted under conditions of 60 nM (dotted line) or 1.3 mM (solid line) extracellular free calcium concentration buffered as calculated by the CalCalc program (74). [Ca²⁺]_i was calculated according to Grynkiewicz et al. (75). Approximately 600 cells were analyzed per second. [Ca²⁺]_i in resting cells is 100 nM.

Figure 3

Antigen-induced IP₃ and total inositol phosphate release is greater in the CD19-positive than in CD19-negative J558Lμm3CD45⁺ cells. (A) Cells (10 × 10⁶/ml) were stimulated with 2.5 μg NP₉BSA for various times and the responses terminated by addition of 100% TCA (20% TCA final). After organic extraction the aqueous phase was assayed for IP₃ content in a [³H]IP₃ receptor binding inhibition assay (NEN-DuPont). The results are expressed as fold increase of IP₃ release over basal (3 pmol IP₃/2.5 × 10⁶ cell equivalents). Shown is mean fold increase from three independent experiments ± standard error. (B) Cells were labeled with Myo- [2-³H(N)]-inositol (21.0 Ci/mmol; 2 μCi/ml) in inositol-free medium containing 5% FCS at 10⁶ cells/ml for 18 h and the generation of inositol phosphates was measured after stimulation of cells for 25 min with 2.5 μg NP₉BSA/10⁷ cells/ml. The fold increase in inositol phosphate release was calculated; the two cell lines had comparable basal rates of inositol turnover as determined by the basal inositol phosphate release. Shown are the mean fold increases from three independent experiments ± standard error. Statistical significance was determined by JMP version 3.16 statistical software (SAS Institute, Cary, NC).

Figure 3

Antigen-induced IP₃ and total inositol phosphate release is greater in the CD19-positive than in CD19-negative J558Lμm3CD45⁺ cells. (A) Cells (10 × 10⁶/ml) were stimulated with 2.5 μg NP₉BSA for various times and the responses terminated by addition of 100% TCA (20% TCA final). After organic extraction the aqueous phase was assayed for IP₃ content in a [³H]IP₃ receptor binding inhibition assay (NEN-DuPont). The results are expressed as fold increase of IP₃ release over basal (3 pmol IP₃/2.5 × 10⁶ cell equivalents). Shown is mean fold increase from three independent experiments ± standard error. (B) Cells were labeled with Myo- [2-³H(N)]-inositol (21.0 Ci/mmol; 2 μCi/ml) in inositol-free medium containing 5% FCS at 10⁶ cells/ml for 18 h and the generation of inositol phosphates was measured after stimulation of cells for 25 min with 2.5 μg NP₉BSA/10⁷ cells/ml. The fold increase in inositol phosphate release was calculated; the two cell lines had comparable basal rates of inositol turnover as determined by the basal inositol phosphate release. Shown are the mean fold increases from three independent experiments ± standard error. Statistical significance was determined by JMP version 3.16 statistical software (SAS Institute, Cary, NC).

Figure 4

Tyrosine phosphorylation of signaling molecules upon antigen stimulation of the CD19-negative and CD19-positive cell lines. J558Lμm3 CD45⁺ variants were stimulated for 1 min with 2.5 μg NP₉BSA/10⁷cells/ml, lysed, and intermediary signaling molecules immunoprecipitated with various antibodies and fractionated on 10% SDS-PAGE. After electrophoretic transfer to PVDF membrane, both anti-phosphotyrosine and anti-effector immunoblotting were performed. Data are representative of at least three independent experiments.

Figure 5

PI3-kinase activity is significantly increased only upon antigen stimulation of the CD19-positive J558Lμm3CD45⁺ cells. CD19-negative and CD19-positive J558Lμm3CD45⁺ cells were stimulated for 2 min with 1 μgNP₉BSA/4 × 10⁶/ml, lysed and PI3-kinase immunoprecipitated with anti-p85 antibody. Immunoprecipitates were washed and assayed for P-I3 kinase activity. Fold increase in PI3-kinase activity after stimulation was determined using a PhosphorImager. The data shown is the mean fold increase over basal in eight independent experiments ± standard error of the mean. Statistical significance was determined by JMP version 3.16 statistical software (SAS Institute, Cary, NC).

Figure 6

Wortmannin inhibits the antigen-induced increase in PI3-kinase activity, [Ca²⁺]_i mobilization and IP₃ generation in the J558Lμm3 CD45⁺CD19⁺ plasmacytoma. (A) Cells were preincubated with DMSO (vehicle) or wortmannin at various concentrations (in 0.2% final DMSO concentration) for 25 min at 25°C, prewarmed for 4 min at 37°C, and then stimulated for 2 min with 1 μg NP₉BSA/4 × 10⁶ cells/ml before analysis of PI3-kinase activation as described in Fig, 4. Shown is the mean fold increase in activity over basal from four independent experiments ± standard error of the mean. Asterisk indicates a statistical significance of P <0.05 determined by JMP 3.16 statistical software. (B) To study the effects of wortmannin on the calcium response, J558Lμm3CD45⁺CD19⁺ cells were Indo-1 AM loaded and preincubated with DMSO (vehicle) or wortmannin at various concentrations (resulting in 0.2% final DMSO) for 25 min at 25°C, prewarmed for 4 min at 37°C, and stimulated at arrow with 250 ng NP₉BSA/10⁶ cells/ml. Shown is the mean [Ca²⁺]_i over time with analysis of ∼600 cells/s. [Ca²⁺]_i in resting cells is 100 nM. (C) J558Lμm3CD45⁺CD19⁺ cells were preincubated with DMSO or increasing concentrations of wortmannin, prewarmed, and stimulated with 2.5 μg NP₉BSA/10⁷/ml for 1 min before IP₃ production was measured by the IP₃ receptor binding inhibition assay. Shown is the mean fold increase over basal in three independent experiments ± standard error of the mean. Asterisk indicates a statistical significance of P <0.05 determined by JMP 3.16 statistical software.

Figure 7

In the CD19-positive J558Lμm3CD45⁺ cells, antigen-induced tyrosine phosphorylation of PLCγ1 and PLCγ2 is slightly inhibited by wortmannin, while Lyn, Ig-α and Syk inductive tyrosine phosphorylation is insensitive to wortmannin. J558Lμm3CD45⁺CD19⁺ cells were preincubated with DMSO or increasing concentrations of wortmannin, prewarmed and stimulated for 1 min with NP₉BSA (2.5 μg/10⁷ cells/ml) and lysed. Immunoprecipitation, SDS-PAGE fractionation and electrophoretic transfer were performed. For each immunoprecipitation, both anti-phosphotyrosine and anti-effector immunoblots are shown. Results are representative of at least three independent experiments.

Figure 8

Mutant CD19 (CD19Y⁴⁸⁴F,Y⁵¹⁵F) does not support antigen-induced [Ca²⁺]_i mobilization. J558Lμm3CD45⁺CD19⁻, J558Lμm3CD45⁺ CD19⁺ and J558Lμm3CD45⁺CD19⁺(Y⁴⁸⁴F,Y⁵¹⁵F) cells were loaded with Indo-1 AM, and analysis of [Ca²⁺]_i initiated before antigen stimulation (250 ng NP₉BSA/10⁶cells/ml). Shown is the mean [Ca²⁺]_i over time with analysis of ∼600 cells/s. [Ca²⁺]_i in resting cells is 100 nM.

Figure 9

Antigen induction of [Ca²⁺]_i mobilization in splenic B cells from normal mice is sensitive to wortmannin. Splenic B cells (ρ ⩾1.066) were isolated from spleens of normal mice, and loaded with Indo-1 AM. The cells were preincubated with DMSO or increasing concentrations of wortmannin for 25 min at 25°C. They were then warmed to 37°C for 4 min and stimulated with 13.2 μg F(ab′)₂RAMIG/10⁶/ml (at arrow). Shown is the mean [Ca²⁺]_i over time with analysis of ∼600 cells/s. [Ca²⁺]_i in resting cells is 70 nM.

Figure 10

BCR-mediated stimulation of PI3-kinase activation is diminished in splenic B cells from CD19^−/− mice compared with CD19^+/+ littermates. Splenic B cells (ρ ⩾1.066) were purified from spleens of CD19^−/− or CD19^+/+ mice, prewarmed, stimulated with F(ab′)₂ RAMIG (26.4 μg/5 × 10⁶/ml) for 2 min, lysed, and then immunoprecipitated with an anti-p85 antibody. Immunoprecipitates were washed and assayed for PI3-kinase activity. Fold increase in PI3-kinase activity after stimulation was determined using a PhosphorImager. (A) Primary autoradiographic data from a representative experiment. (B) Mean fold increase over basal in three independent experiments ± standard error of the mean. Statistical significance was determined by JMP version 3.16 statistical software (SAS Institute).

Figure 10

BCR-mediated stimulation of PI3-kinase activation is diminished in splenic B cells from CD19^−/− mice compared with CD19^+/+ littermates. Splenic B cells (ρ ⩾1.066) were purified from spleens of CD19^−/− or CD19^+/+ mice, prewarmed, stimulated with F(ab′)₂ RAMIG (26.4 μg/5 × 10⁶/ml) for 2 min, lysed, and then immunoprecipitated with an anti-p85 antibody. Immunoprecipitates were washed and assayed for PI3-kinase activity. Fold increase in PI3-kinase activity after stimulation was determined using a PhosphorImager. (A) Primary autoradiographic data from a representative experiment. (B) Mean fold increase over basal in three independent experiments ± standard error of the mean. Statistical significance was determined by JMP version 3.16 statistical software (SAS Institute).

Figure 11

BCR-mediated [Ca²⁺]_i mobilization is reduced in splenic B cells from CD19^−/− mice compared with CD19^+/+ littermates at both optimal and suboptimal doses of antibody. B cells (ρ ⩾1.066) were isolated from spleen of CD19^−/− and CD19^+/+ mice and loaded with Indo-1 AM. [Ca²⁺]_i mobilization was monitored before and after stimulation of the cells with F(ab′)₂ RAMIG. (A) Cells were stimulated with 13 μg/10⁶ cells/ml; shown is an overlay of the [Ca²⁺]_i mobilization pattern in an experiment where splenic B cells from two CD19^−/− and two CD19^+/+ mice were compared independently. This experiment was performed 10 times with similar results. (B) Cells were stimulated with a suboptimal dose of F(ab′)₂ RAMIG (1.3 μg/10⁶ cells/ml). Shown is the mean [Ca²⁺]_i over time with analysis of ∼600 cells/s. [Ca²⁺]_i in resting cells is 70 nM.