Agammaglobulinemia and absent B lineage cells in a patient lacking the p85α subunit of PI3K

Abstract

Whole exome sequencing was used to determine the causative gene in patients with B cell defects of unknown etiology. A homozygous premature stop codon in exon 6 of PIK3R1 was identified in a young woman with colitis and absent B cells. The mutation results in the absence of p85α but normal expression of the p50α and p55α regulatory subunits of PI3K. Bone marrow aspirates from the patient showed <0.1% CD19(+) B cells with normal percentages of TdT(+)VpreB(+)CD19(-) B cell precursors. This developmental block is earlier than that seen in patients with defects in the B cell receptor signaling pathway or in a strain of engineered mice with a similar defect in p85α. The number and function of the patient's T cells were normal. However, Western blot showed markedly decreased p110δ, as well as absent p85α, in patient T cells, neutrophils, and dendritic cells. The patient had normal growth and development and normal fasting glucose and insulin. Mice with p85α deficiency have insulin hypersensitivity, defective platelet function, and abnormal mast cell development. In contrast, the absence of p85α in the patient results in an early and severe defect in B cell development but minimal findings in other organ systems.

Figure 1.

Mutation in PIK3R1 resulting in the absence of B lineage cells. (A) The intron/exon organization of PIK3R1 is shown, with the exons encoding p85α, p50α, and p55α color-coded to match the schematic drawing of the proteins. The specific mutation in exon 6 of PIK3R1 is shown above the gene. (B) The pedigree of the patient’s family is shown. Males who died of infection at <2 yr of age are indicated by a diagonal line through a square. The numbers within the square and circle indicate the number of males and females, respectively. The patient is indicated by an arrow. (C) PCR-based assay for mutation screening is shown. The lanes labeled HC contain digested PCR products from healthy controls. The lanes labeled Pt, M, and F contain digested PCR products from the patient, her mother, and her father, respectively.

Figure 2.

Analysis of B cell development in the blood and bone marrow by flow cytometry. (A) Ficoll-separated peripheral blood mononuclear cells were stained with PE-labeled anti-CD19 and FITC-labeled anti-CD20. The percentage of cells positive for CD19 and CD20 is indicated. The number of events shown is 20,000 for the healthy control and 250,000 for both the disease control (a patient with a mutation within a transcriptional regulatory element in intron 1 of BTK) and the patient. (B) Bone marrow cells were stained with PE-labeled anti-CD19 and FITC-labeled anti-CD34 or APC-labeled anti-CD19, FITC-labeled anti-TdT, and PE-labeled anti-VpreB. The percentage of cells within the lymphoid gate that fall into each of the sectors is shown. The number of events shown is 20,000 for the healthy control, 125,000 for the disease control (a patient with a mutation at the +1 donor splice site of intron 2 in BTK), and 250,000 for the patient. The peripheral blood analysis was performed on samples obtained from the p85α-deficient patient on three separate occasions. Bone marrow from the patient was obtained once and stained twice.

Figure 3.

Expression of p85α, p55α, and p50α and the p110 isoforms in hematopoietic cells and cell lines from healthy controls, disease controls, and the patient. (A) Expression of the indicated proteins in peripheral blood T cells isolated by negative selection and in activated T cells that had been stimulated with phytohemagglutinin and then supplemented with 10% IL-2 every 2–3 d for 2 wk. (B) Expression of the same proteins in primary B cells, EBV-transformed B cell lines, and NK cells. Because both the patient (Pt) and the disease controls (DC) lack B cells, expression of PI3K isoforms from two healthy controls (HC) is shown to document that the absence of p50α and p55α was a reproducible finding. The patient had too few NK cells to permit analysis. (C) Expression of PI3K isoforms in neutrophils and DCs. The immunoblots in A–C were sequentially probed with antibodies to p110δ, p85α N terminus, p85α C terminus, and actin. (D) The expression of p110 isoforms in activated T cells and neutrophils is shown. The disease control for NK cell analysis had mutations in λ5. All other disease controls had mutations in BTK.

Figure 4.

Cytokine production by cultured DCs from healthy controls, disease controls, and the patient. (A) The cell surface expression of monocyte-specific (CD14) and DC-specific (DC-SIGN) surface molecules was analyzed on cells immediately after selection on CD14 beads (post-CD14⁺ purification) and after a 7–8-d culture in GM-CSF + IL-4 with the addition of TNF for the final 24 h of culture (post-DC differentiation). The isotype control is shown in gray, and the specific staining for the healthy control, disease control, and patient cells are shown in black, blue, and red, respectively. (B) Representative qRT-PCR analysis of inflammatory messenger RNA (mRNA) expression by stimulated DCs. DCs isolated from control (black), disease controls (blue), or the patient (red) were stimulated with 100 ng/ml LPS (left) or 100 µg/ml curdlan (right), and messenger RNAs were analyzed by qRT-PCR with normalization to GAPDH. All the PCRs were performed twice with independent cDNA preps. The disease controls had mutations in BTK. Note the difference in scale in the response to LPS versus curdlan.