CD24 Regulates the Formation of Ectosomes in B Lymphocytes

Abstract

CD24 is a glycophosphatidylinositol-linked protein that regulates B cell development. We previously reported that stimulation of CD24 on donor B cells promotes the transfer of functional receptors to recipient B cells via extracellular vesicles (EVs). However, the mechanisms regulating CD24-mediated formation of bioactive EVs are unknown. Using bioinformatics, we found a connection between CD24, and PI3K/AKT, tran and mTOR. To determine if these pathways regulate EV release, we used flow cytometry to follow the transfer of EVs carrying lipid-associated GFP and surface IgM from donor to recipient B cells. Using chemical and genetic inhibition, we found that a PI3K/mTORC2/ROCK/actin pathway regulates bioactive EV formation via activation of acid sphingomyelinase (aSMase) upstream of PI3K. Using single EV analysis, we found that CD24 regulates the formation of the subset of bioactive EVs that are taken up by recipient cells and not total EVs. Interestingly, we also found that ROCK and aSMase modulate ectosome but not exosome formation, when CD24 is stimulated. Lastly, through live cell imaging, we found that PI3K and ROCK are required for inducing membrane dynamics associated with EV formation. These data suggest that this pathway regulates bioactive EV release that, in turn, could regulate B cell development.

Keywords: B lymphocyte; CD24; acid sphingomyelinase (aSMase); ceramide; ectosome; extracellular vesicle.

FIGURE 1

CD24 expression is associated with the PI3K‐Akt signalling and mTOR pathways. (A) Hierarchical cluster analysis showing the clade of 44 genes (41 annotated) with similar differential gene expression patterns as CD24a. (B) Pathway network analysis generated from the differentially expressed 41‐gene list identified the PI3K/Akt and mTOR signalling pathway as being associated with CD24 expression in developing B cells. The red boxes indicate nodes with terms related to PI3K‐Akt or mTOR signalling. (C) WEHI‐231‐GFP cells were pre‐treated with LY294002 or DMSO (vehicle control) for 15 min then cells were stimulated with isotype antibody (isotype) or anti‐CD24 stimulating (CD24) antibody for the times indicated. PIP₃ levels were analysed by ELISA, n = 4, significance was determined by a one‐way ANOVA followed by the Sidak's multiple comparison test *p < 0.05.

FIGURE 2

CD24‐mediated EV transfer is regulated by PI3K. (A and B) WEHI‐231‐GFP cells were pre‐treated with LY294002 or DMSO for 15 min, then stimulated with anti‐CD24 (CD24) or isotype control (isotype, iso) for 15 min, followed by washout and then a 24 h co‐culture with WEHI‐303‐tdTomato cells. (A) Percent GFP and tdTomato double‐positive cells and (B) Percent IgM and tdTomato double‐positive cells after 24 h incubation. n = 6, statistical significance determined by a two‐way ANOVA (interaction significant at p = 0.0003 for A and p = 0.0002 for B) followed by the Sidak's multiple comparison test *p < 0.05, ****p < 0.001. (C) Total cell lysates from WEHI‐231‐GFP cells pre‐treated with DMSO or LY2940002 for 15 min, then stimulated with the above antibodies for different times indicated. Phosphorylated Akt and total Akt expression levels were determined by immunoblotting. (D) WEHI‐231‐GFP cells were transfected with scrambled control siRNA or PI3Kδ siRNA, and PI3Kδ and GAPDH levels determined by immunoblotting. Shown are representative western blots from three replicates with molecular weight marker positions indicated on the left of each blot. (E and F) WEHI‐231‐GFP cells with or without PI3Kδ siRNA knock‐down were stimulated with anti‐CD24 or isotype for 15 min, followed by washout and then a 24 h co‐culture with WEHI‐303‐tdTomato cells. (E) Percent GFP and tdTomato double‐positive cells. (F) Percent IgM and tdTomato double‐positive cells after 24 h incubation. n = 3, statistical significance determined by a two‐way ANOVA (interaction significant at p = 0.0015 for E and p = 0.0088 for F) followed by the Sidak's multiple comparison test **p < 0.01, ***p < 0.005.

FIGURE 3

CD24‐mediated EV transfer is dependent on Akt and mTOR signalling pathways. (A and B) WEHI‐231‐GFP cells were pre‐treated with MK‐2206 or DMSO for 15 min, then stimulated with anti‐CD24 (CD24) or isotype control (Isotype) for 15 min, followed by washout and then a 24 h co‐culture with WEHI‐303‐tdTomato cells. (A) Percent GFP and tdTomato double‐positive cells and (B) Percent IgM and tdTomato double‐positive cells after 24 h incubation. n = 4, statistical significance determined by a two‐way ANOVA (interaction significant at p = 0.0074 for A and p = 0.0002 for (B) followed by the Sidak's multiple comparison test *p < 0.05, **p < 0.01, ****p < 0.001. (C and D) WEHI‐231‐GFP cells were pre‐treated with Torin 1, Rapamycin, JR‐AB2‐011 or DMSO for 15 min, then stimulated with anti‐CD24 or isotype control for 15 min, followed by washout and then a 24 h co‐culture with WEHI‐303‐tdTomato cells. (C) Percent GFP and tdTomato double‐positive cells and (D) Percent IgM and tdTomato double‐positive cells after 24 h incubation. n = 4, statistical significance determined by a two‐way ANOVA (interaction significant at p = 0.0143 for C and p = 0.0066 for D) followed by the Sidak's multiple comparison test *p < 0.05, **p < 0.01, ***p < 0.005, and ****p < 0.001.

FIGURE 4

CD24‐mediated EV transfer is dependent on ROCK. (A and B) WEHI‐231‐GFP cells were pre‐treated with Y27632 or DMSO for 15 min, then stimulated with anti‐CD24 (CD24) or isotype control (isotype, iso) for 15 min, followed by washout and then a 24 h co‐culture with WEHI‐303‐tdTomato cells. (A) Percent GFP and tdTomato double‐positive cells and (B) Percent IgM and tdTomato double‐positive cells after 24 h incubation. n = 6, statistical significance determined by a two‐way ANOVA (interaction significant at p = 0.03 for A and p = 0.0069 for B) followed by the Sidak's multiple comparison test *p < 0.05, **p < 0.01, ***p < 0.005, ****p < 0.001. (C) Total cell lysates from WEHI‐231‐GFP cells pre‐treated with DMSO or Y27632 for 15 min, then stimulated with the above antibodies for different times indicated. Phosphorylated Cofilin and total Cofilin expression levels were determined by immunoblotting. Shown are representative western blots from three replicates with molecular weight marker positions indicated on the left of each blot. (D) WEHI‐231‐GFP cells were transfected with scrambled control siRNA or ROCK siRNA, and ROCK and GAPDH levels determined by immunoblotting. (E and F) WEHI‐231‐GFP cells with or without ROCK siRNA knock‐down were stimulated with anti‐CD24 or isotype for 15 min, followed by washout and then a 24 h co‐culture with WEHI‐303‐tdTomato cells. (E) Percent GFP and tdTomato double‐positive cells and (F) Percent IgM and tdTomato double‐positive cells after 24 h incubation. n = 3, statistical significance determined by a two‐way ANOVA (interaction significant at p = 0.0011 for E and p = 0.0013 for F) followed by the Sidak's multiple comparison test ***p < 0.005, ****p < 0.001.

FIGURE 5

CD24‐mediated EV transfer is controlled by actin cytoskeleton re‐organisation. (A and B) WEHI‐231‐GFP cells were pre‐treated with Cytochalasin D or DMSO for 15 min, then stimulated with anti‐CD24 (CD24) or isotype control (isotype) for 15 min, followed by washout and then a 24 h co‐culture with WEHI‐303‐tdTomato cells. (A) Percent GFP and tdTomato double‐positive cells and (B) Percent IgM and tdTomato double‐positive cells after 24 h incubation. n = 4, statistical significance determined by a two‐way ANOVA (interaction significant at p = 0.0017 for A and p = 0.0027 for B) followed by the Sidak's multiple comparison test **p < 0.01, ***p < 0.005, and ****p < 0.001.

FIGURE 6

CD24‐mediated EV transfer is controlled by aSMase activity. (A and B) WEHI‐231‐GFP cells were pre‐treated with imipramine or DMSO for 15 min, then stimulated with anti‐CD24 (CD24) or isotype control (isotype) for 15 min, followed by washout and then a 24 h co‐culture with WEHI‐303‐tdTomato cells. (A) Percent GFP and tdTomato double‐positive cells and (B) Percent IgM and tdTomato double‐positive cells after 24 h incubation. n = 4, statistical significance determined by a two‐way ANOVA (interaction significant at p = 0.0006 for A and p = 0.0006 for B) followed by the Sidak's multiple comparison test ***p < 0.005, ****p < 0.001. (C and D) WEHI‐231‐GFP cells were pre‐treated with ARC39 or DMSO for 15 min, then stimulated with anti‐CD24 (CD24) or isotype control (isotype) for 15 min, followed by washout and then a 24 h co‐culture with WEHI‐303‐tdTomato cells. (C) Percent GFP and tdTomato double‐positive cells and (D) Percent IgM and tdTomato double‐positive cells after 24 h incubation. n = 6, statistical significance determined by a two‐way ANOVA (interaction significant at p < 0.0001 for C and p < 0.0001 for B) followed by the Sidak's multiple comparison test ***p < 0.005, ****p < 0.001. (E) WEHI‐231‐GFP cells were pre‐treated with imipramine or DMSO (vehicle control) for 15 min then cells were stimulated with isotype antibody (isotype) or anti‐CD24 stimulating antibody (CD24) for the times indicated. PIP₃ levels were analysed by ELISA, n = 5, significance was determined by a one‐way ANOVA followed by the Sidak's multiple comparison test *p < 0.05, **p < 0.01.

FIGURE 7

CD24 does not change overall numbers of secreted EVs while PI3K, ROCK and aSMase increase ectosome production when CD24 is stimulated. WEHI‐231‐GFP cells were pre‐treated with LY294002, Y27632, imipramine or ARC39 for 15 min, then stimulated with anti‐CD24 (CD24) or isotype control (isotype) for 15 min, followed by a washout and 1 h incubation in EV‐depleted media. EVs were stained with (A) membrane stain vFRed, (B) vFRed and GFP, (C) vFRed and ectosomal markers (CD9 and ANXA2) or (D) vFRed and exosomal markers (CD63 and LAMP‐1). n = 4, statistical significance determined by a two‐way ANOVA, followed by Sidak's multiple comparison test *p < 0.05, **p < 0.01, ***p < 0.005, ****p < 0.001. (A) Stimulation significant at p < 0.0001, (B) Stimulation significant at p < 0.0001 and inhibitor treatment significant at p < 0.01, (C) Stimulation significant at p < 0.0001 and (D) Stimulation and inhibitor treatment were not significant.

FIGURE 8

CD24 induces dynamic membrane movement in B cells via PI3K and ROCK. WEHI‐231‐GFP cells were pre‐treated with LY294002 (CD24+LY) or Y27632 (CD24+Y) or DMSO for 15 min, then stimulated with anti‐CD24 (CD24) or isotype (Iso) control. (A) The cells were captured at 30 min after stimulation. White arrows indicate where EV release was seen in the videos. Scale bar = 2 µm. (B) Cell activation was determined using the coefficient of variance (cv) of the area over 10 images covering 5 min and visualised with violin plots (the filled black circle indicates the mean). Five to 12 cells were analysed per video for a total of 38–40 cells analysed per treatment group. Pairwise comparisons using the Kolmogorov–Smirnov test were used to determine significant differences: *p < 0.05, **p < 0.01.