Germline-activating mutations in PIK3CD compromise B cell development and function

Abstract

Gain-of-function (GOF) mutations in PIK3CD, encoding the p110δ subunit of phosphatidylinositide 3-kinase (PI3K), cause a primary immunodeficiency. Affected individuals display impaired humoral immune responses following infection or immunization. To establish mechanisms underlying these immune defects, we studied a large cohort of patients with PIK3CD GOF mutations and established a novel mouse model using CRISPR/Cas9-mediated gene editing to introduce a common pathogenic mutation in Pik3cd In both species, hyperactive PI3K severely affected B cell development and differentiation in the bone marrow and the periphery. Furthermore, PI3K GOF B cells exhibited intrinsic defects in class-switch recombination (CSR) due to impaired induction of activation-induced cytidine deaminase (AID) and failure to acquire a plasmablast gene signature and phenotype. Importantly, defects in CSR, AID expression, and Ig secretion were restored by leniolisib, a specific p110δ inhibitor. Our findings reveal key roles for balanced PI3K signaling in B cell development and long-lived humoral immunity and memory and establish the validity of treating affected individuals with p110δ inhibitors.

Figure 1.

GOF mutations in PIK3CD arrest peripheral B cell development and differentiation. PBMCs from healthy donors (n = 45–60) and patients with PIK3CD GOF mutations (n = 21–39) were labeled with mAbs against CD20, CD10, CD27, IgG, or IgA. The proportions of (A) B (CD20⁺) cells within the lymphocyte gate, (B) transitional, naive, and memory cells within the B cell population, and (C) IgG⁺ and IgA⁺ cells within the memory population were determined by flow cytometry. Histogram and contour plots are representative of healthy donors or PIK3CD GOF patients. Each symbol in the summary graphs corresponds to an individual donor or patient; horizontal bars represent the mean. Significant differences were determined by unpaired Student's t tests. ****, P < 0.0001. (D) Proportions of transitional (left panel) and memory (right panel) B cells in healthy donors (black) and PIK3CD GOF patients (red) were determined as a function of age.

Figure 2.

Peripheral B cells in PIK3CD GOF patients have an immature phenotype. Expression of (A) CD5, (B) CD38, (C) CD21, (D) CD23, (E) CD44, (F) CD19, (G) CCR7, (H) CXCR4, (I) CXCR5, or (J) BCL2 was determined on transitional, naive, and memory B cells in peripheral blood of healthy donors or patients with PIK3CD GOF mutations (n = 6–30). Histogram plots are representative of healthy donors (upper panels) or PIK3CD GOF patients (lower panels). The bar graphs depict the geometric mean fluorescence intensity (± SEM) of each indicated molecule. Significant differences were determined by unpaired Student's t tests. *, P < 0.05; **, P < 0.01; ****, P < 0.0001.

Figure 3.

PIK3CD GOF mutations block B cell development in the bone marrow at the pre-BII stage. BM aspirates from healthy donors (n = 6), patients with PIK3CD GOF mutations (n = 3), or one patient following hematopoietic stem cell transplant were labeled with mAbs against CD34, CD19, CD20, CD10, IgM, and CD27. Proportions of pro-B (CD19⁺CD34⁺CD10⁺CD20⁻IgM⁻), pre-BI (CD19⁺CD34⁻CD10⁺CD20⁻IgM⁻), pre-BII (CD19⁺CD34⁻CD10⁺CD20^dimIgM⁻), immature (CD19⁺CD34CD10⁺CD20⁺ IgM⁺), and recirculating mature (CD19⁺CD34⁻CD10⁻CD20⁺) B cells were determined. (A–C) Contour plots in show CD34 versus CD10 staining to identify pro-B cells, which were then further analyzed for pre-BI, pre-BII, immature, and recirculating mature according to differential expression of CD20 and CD10. (D) Mean ± SEM of the different subsets of B cells in the BM. (E) Representative contour plots showing CD10 versus CD27 staining on CD20⁺ B cells, and (F) mean ± SEM of CD10⁺CD27⁻, naive (CD20⁺CD10⁻CD27⁻) and memory (CD20⁺CD27⁺) B cells in the BM. Significant differences were determined by Student's t test. **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Figure 4.

Intact differentiation to plasmablasts but impaired class switching in PIK3CD GOF naive and transitional B cells. (A–G) Naive and/or transitional B cells from healthy donors or PIK3CD GOF patients were cultured for 5 d with CD40L alone or together with IL-21. (A and B) Proportions of CD38^hiCD27^hi plasmablasts generated from naive B cells were determined by flow cytometry (n = 7–8). (C and D) Expression of Blimp-1 was determined by (C) qPCR in naive and transitional B cells cultured with CD40L or CD40L/IL-21 (n = 5) and (D) flow cytometry of activated B cell blasts (CD27⁻CD38^-/lo) and plasmablasts generated from naive B cells cultured with CD40L or CD40L/IL-21 (n = 4–5). (E) Expression of AICDA was determined by qPCR in naive and transitional B cells cultured with CD40L or CD40L/IL-21 (n = 5). (F and G) Expression of intracellular IgM, IgG, and IgA by plasmablasts was determined by flow cytometry (n = 5–6). The FACS plots in A, D, and F are representative of cultured naive B cells from healthy donors or PIK3CD GOF patients; the graphs in B, C, E, and G depict the mean ± SEM of independent experiments using naive B cells from different donors and patients. (H and I) Sort-purified naive (H; n = 9–12) and transitional (I; n = 11–14) B cells from healthy donors or PIK3CD GOF patients were cultured for 5 or 7 d with CD40L and IL-21. Ig secretion was then determined. Values represent the mean ± SEM. Significant differences were determined by Student's t tests. *, P < 0.05; **, P < 0.01; ***, P < 0.001.

Figure 5.

PIK3CD GOF transitional B cells have impaired molecular responses to IL-21. (A and B) Transitional B cells from healthy donors and PIK3CD GOF patients were cultured with CD40L alone or together with IL-21 and gene expression profiles for each were obtained by microarray. (A) Heat map of log₂-transformed expression values for genes induced by IL-21 and involved in plasmablast generation. (B) Heat map of log₂-transformed expression values for key driver genes that distinguish CD40L/IL-21–treated transitional B cells from healthy donors and PIK3CD GOF patients identified by NMF. (C and D) qPCR analysis of human tonsil B cell subsets (N, naive; M, memory; GC, germinal center; PC, plasma cells) showing differential expression of genes identified in B that are down-regulated (C) or up-regulated (D) in human PCs relative to other B cell populations. Values represent the mean ± SEM of experiments using tonsils from 4 different donors. (E and F) Transitional B cells from healthy donors or PIK3CD GOF patients were cultured with CD40L alone or together with IL-21 for 5 d. (E) Proportions of CD27^hiCD38^hi plasmablasts generated in these cultures were determined by flow cytometry. Contour plots are representative of data obtained from two to three independent experiments using transitional B cells isolated from different healthy donors or PIK3CD GOF patients. (F) Relative expression levels of the indicated genes were determined in transitional B cells by qPCR. Data are expressed as mean fold change (±SEM; n = 3) of gene expression in transitional B cells stimulated with CD40L/IL-21 relative to transitional B cells stimulated with CD40L alone. The dashed line represents a fold-change of 1.0, indicating no change in the presence versus the absence of IL-21 by CD40L-stimulated transitional B cells.

Figure 6.

Mice with overactive PI3K show aberrant B cell development. (A) B cells were stained intracellularly for phosphorylated Akt (T308 and S473) and S6 (S235/236). Histograms show representative staining from WT (black), Pik3cd^E1020K heterozygous (Pik3cd^GOF, red), or Pik3cd^E1020K homozygous (Pik3cd^homGOF blue) mice and graphs give MFI relative to WT controls (mean ± SEM, n = 5). (B–D) BM and spleens from WT and Pik3cd^E1020K heterozygous GOF mice aged 8–12 wk were stained to identify different B cell populations. (B) B cell development in the BM. Flow cytometry plots showing representative staining of IgD versus IgM on B220⁺ cells. Numbers are percent IgM⁻IgD⁻, IgM^hi, and IgD^hi cells. IgM⁻IgD⁻ cells were further gated on CD24 and CD43 to identify pre-pro–, pro- and pre-B cells. Graphs give mean ± SEM (n = 9–12). (C) Transitional cells in the spleen. Flow plots show IgM versus CD23 on B220⁺CD93⁺ cells. Graphs show absolute numbers of T1 and T2/T3 cells as well as the percentage of each population that is IgM^hi (mean ± SEM, n = 9–12). (D) Percentages of follicular (CD23⁺CD21^lo) and MZ (CD21^hiCD23^lo) B cells were determined in the mature B cell population (CD93⁻) of the spleen. Flow plots show representative staining of CD21 versus CD23. Graphs show absolute numbers of follicular and MZ cells (mean ± SEM, n = 10–13). (E) Percentages of B1a (CD19⁺B220^loCD5⁺) or B1b (CD19⁺B220^loCD5⁻) cells were determined. Flow plots show representative staining of B220 versus CD19 gated on CD19⁺ cells (upper panel) and CD19 versus CD5 gated on CD19⁺B220^lo cells (lower panel). Graphs show absolute numbers of B1a and B1b cells in the spleen (mean ± SEM, n = 6–8). Significant differences were determined by unpaired Student's t tests. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Figure 7.

Pik3cd GOF B cells show a division-related decrease in switching to multiple IgG isotypes. (A–C) Follicular B cells (B220⁺CD93⁻CD23⁺CD21^lo) were sorted from the spleens of WT or Pik3cd^GOF mice. Cells were labeled with CFSE and stimulated with anti-CD40 + IL-4 (A), LPS + TGF-β (B) or LPS alone (C) for 4 d after which they were stained for IgG1, IgG2b and IgG3 as shown. Left panels show total percentage of cells of each isotype. Right panels show the percentage of switched cells in each division as determined by CFSE. Plots show mean ± SEM, n = 4–6. Significant differences were determined by paired t tests. **, P < 0.01; ***, P < 0.001.

Figure 8.

Pik3cd GOF B cells show defective switching but normal expansion and affinity maturation in vivo. WT or Pik3cd^GOF SWHEL cells were transferred to WT congenic hosts, which were then immunized with HEL-2x-SRBC. (A) The expansion of SW_HEL cells was tracked over time (mean ± SEM, n = 3–4 mice per group, representative experiment shown). (B) Percentage of cells with a plasmablast or GC phenotype was determined. (C) Percentage of cells that switched to IgG1 or were unswitched (IgM⁺) was determined in the plasmablast and GC populations. (D) Levels of HEL-specific serum Ig of various classes at day 5.5 were determined by ELISA. (B–D, each linked point shows mean ± SEM [n = 3–5] of single experiment). (E) Donor GC IgG1⁺ cells were sorted on day 10 and sequenced to identify mutations. Significant differences were determined by paired t tests. **, P < 0.01.

Figure 9.

A p110δ inhibitor partially alleviates the defects in B cell differentiation due to hyperactive PI3K signaling. (A–C) Sort-purified transitional B cells from healthy donors or PIK3CD GOF patients (n = 5) were cultured for 5 d with CD40L plus IL-21 in the absence or presence of the PI3K p110δ inhibitor leniolisib. Secretion of (A) IgM, (B) IgG, and (C) IgA was then determined. Data are expressed as mean percentage of Ig secreted (± SEM) by B cells stimulated with CD40L/IL-21. (D–G) Follicular B cells (B220⁺CD93⁻CD23⁺CD21^lo) were sorted from the spleens of WT or Pik3cd^GOF mice. Cells were labeled with CFSE and stimulated with anti-CD40 + IL-4 in the absence or presence of leniolisib for 4 d. (D and E) Cells were stained for expression of IgG1 and the total percentage of IgG1+ B cells detected (D) and the % of IgG1+ B cells in each division as determined by CFSE dilution (E) were determined. (F) Expression of Aicda in WT or Pik3cd^GOF B cells cultured with anti-CD40 + IL-4 was determined by qPCR. Plots show mean ± SEM, n = 3. (G) Plot of Aicda expression versus percentage IgG1+ cells generated following culture of WT or Pik3cd^GOF B cells with anti-CD40 + IL-4 in the absence or presence of leniolisib. Significant differences were determined by two-way ANOVA. **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.