Epistatic interactions between mutations of TACI (TNFRSF13B) and TCF3 result in a severe primary immunodeficiency disorder and systemic lupus erythematosus

Abstract

Common variable immunodeficiency disorders (CVID) are a group of primary immunodeficiencies where monogenetic causes account for only a fraction of cases. On this evidence, CVID is potentially polygenic and epistatic although there are, as yet, no examples to support this hypothesis. We have identified a non-consanguineous family, who carry the C104R (c.310T>C) mutation of the Transmembrane Activator Calcium-modulator and cyclophilin ligand Interactor (TACI, TNFRSF13B) gene. Variants in TNFRSF13B/TACI are identified in up to 10% of CVID patients, and are associated with, but not solely causative of CVID. The proband is heterozygous for the TNFRSF13B/TACI C104R mutation and meets the Ameratunga et al. diagnostic criteria for CVID and the American College of Rheumatology criteria for systemic lupus erythematosus (SLE). Her son has type 1 diabetes, arthritis, reduced IgG levels and IgA deficiency, but has not inherited the TNFRSF13B/TACI mutation. Her brother, homozygous for the TNFRSF13B/TACI mutation, is in good health despite profound hypogammaglobulinemia and mild cytopenias. We hypothesised that a second unidentified mutation contributed to the symptomatic phenotype of the proband and her son. Whole-exome sequencing of the family revealed a de novo nonsense mutation (T168fsX191) in the Transcription Factor 3 (TCF3) gene encoding the E2A transcription factors, present only in the proband and her son. We demonstrate mutations of TNFRSF13B/TACI impair immunoglobulin isotype switching and antibody production predominantly via T-cell-independent signalling, while mutations of TCF3 impair both T-cell-dependent and -independent pathways of B-cell activation and differentiation. We conclude that epistatic interactions between mutations of the TNFRSF13B/TACI and TCF3 signalling networks lead to the severe CVID-like disorder and SLE in the proband.

Figure 1

Novel de novo TCF3 mutation discovered in a CVID family carrying C104R TACI variant. (a) Digenic inheritance of TNFRSF13B (c.310T>C, C104R TACI) and TCF3 (T168fx191) mutations in a three-generation New Zealand family. Whole-exome sequencing was performed on II.2, III.1 and III.2 (indicated by *). The proband (II.2) is indicated by an arrow. Circles, female; squares, male; gray, TNFRSF13B/TACI C104R mutation; blue TCF3 T168fsX191 mutation (as indicated). The proband (arrow, II.2) is heterozygous for both the TCF3 T168fsX191 and TNFRSF13B/TACI C104R mutations. Other family members who have inherited TCF3 T168fsX191 and TNFRSF13B/TACI C104R mutations are shown. CVID, common variable immunodeficiency disorder; SLE, systemic lupus erythematosus; sIgAD, selective IgA deficiency; T1D, Type 1 Diabetes, sHGUS, symptomatic hypogammglobulinaemia of uncertain significance; WT, wild-type. (b) Electropherograms showing the T168fsX191 mutation of TCF3 and C104R (c.310T>C) mutation of TACI gene in the proband II.2. The proband’s son (III.1) has inherited the TCF3 T168fsX191 mutation, but not the TNFRSF13B/TACI C104R mutation. The proband’s clinically unaffected daughter (III.2) has not inherited either mutation. The TCF3 T168fsX191 mutation was absent in the proband’s parents, indicating a de novo origin. (c) Schema of wild-type and truncated mutant TCF3 T168fsX191 gene. Exons coding E2A functional domains, activation domain 1 and 2 (AD1, AD2) and helix-loop-helix (HLH) domains are shown. (d) E2A (E47) protein expression was assessed by western blotting of lysates following 30 min PMA/ionomycin stimulation of PBMCS in the kindred, as indicated (U, unstimulated; S, stimulated). (e) PBMCs from the proband (II.2) and healthy control (HC) individuals (n=2) were unstimulated, or stimulated with PMA+ionomycin for 30 or 60 min as indicated, and cell lysates were analysed for p105 phosphorylation (P-p105, Ser933), expression of p105 and p50 by western blotting. Beta-actin was used as a protein loading control. Results are representative of two independent experiments.

Figure 2

Immunoglobulin isotype switching pathways showing nodes of intracellular signal integration between TACI and TCF3/E2A. T-cell-independent isotype switching occurs through TACI and TLRs while T-cell dependent switching occurs through CD40 and IL-4 or IL-21. Ligation of the B-cell receptor synergises with both pathways. TCF3/E2A contributes to the expression of AID, 14-3-3γ and Ig production and therefore influences both T-cell-dependent and -independent Ig switching pathways. 14-3-3γ is a scaffolding protein and targets AID to switch regions. Mutations are shown in red stars. BCR—B cell receptor. TLR, Toll-like receptor.

Figure 3

(a) Immunophenotyping, proliferation and isotype switching in TCF3/ TNFRSF13B/TACI mutant B cells. (a) Immunophenotyping results indicating proportions of naïve (CD20⁺CD27⁻) and memory (CD20⁺CD27⁺) B cells, and CD4⁺ and CD8⁺ T cells in PBMCs isolated from available family members as indicated, and representative healthy donor controls. (b) Relative proportions of IgM/G/A memory B cells from each family member and unrelated healthy donors (each as a proportion of total memory B cells). IgM-expressing cells are shown in black, IgG⁻ in gray and IgA⁻ isotype switched memory B cells in white, as indicated. (c) Total numbers of lymphocytes, B cells, naïve (CD20⁺CD27⁻) and memory (CD20⁺CD27⁺) B cells in peripheral blood from each family member and unrelated healthy donors (HD=12). Immunophenotyping and cell counts were performed in two separate experiments.

Figure 4

Severe defect in in vitro antibody production in proband demonstrates epistasis. (a) Immunoglobulin production from supernatants collected from in vitro cultures of naïve B cells isolated from PBMCs of each family member, stimulated as indicated with CD40L (100 ng ml^–1), IL-4 (50 ng ml^–1), IL-21 (50 ng ml^–1), CpG (1 μg ml^–1) and APRIL (500 ng ml^–1). Supernatants were assessed for secretion of IgG, IgA and IgM as indicated. (b) Representative Cell Trace Violet (CTV) plots and IgG isotype switched cells following in vitro stimulation of naïve B cells with CD40L+IL-4+IL-21 for 6 days (representative from two independent experiments). Cells were isolated, labelled with CTV, stimulated and collected after 6 days of culture and the division profiles and proportions of IgG expressing cells determined. FMO, fluorochrome minus one.

Figure 5

Severe defect in generation of antibody secreting cells in E2A/TACI-deficient cells. (a–c) Summary graphs from in vitro proliferation of naïve B cells stimulated as indicated with CD40L (100 ng ml⁻¹), IL-4 (50 ng ml⁻¹), IL-21 (50 ng ml⁻¹), CpG (1 μg ml⁻¹) and APRIL (500 ng ml⁻¹) as indicated. Isolated cells were collected after 5 days of culture, cell surface stained and analysed by flow cytometry for the (a) proportion of antibody secreting cells (ASC, CD27^hiCD38⁺) and (b) total number of ASC and (c) total lymphocyte number. Cell counts and proportion of ASC are shown for the proband, with both TNFRSF13B/TACI and TCF3 mutations in white; her son (III.1), expressing TCF3 T168fsX191 mutant B cells only (blue); TACI-deficient individuals (II.3, II.4, gray); and wild-type (III.2 and HD, black). Summary graphs of the proportions and total number of differentiated cells for all family members and healthy donors (HD, n=4) was performed in two independent experiments.

Figure 6

Quantitation of epistatic interactions of TCF3 and TACI mutations showing a greater net effect than the sum of each individual mutation. Total Serum Ig, clinical score and TNFRSF13B/TACI C104R and TCF3 T161fsX191 genotype for each family member, as indicated. The serum IgG for the proband II.2 was obtained in 2002. Normal serum Ig ranges (g l⁻¹) shaded in green. Lower graph: summary of total Ig levels detected in naïve B-cell cultures for each available family member, stimulated with APRIL, CpG, IL-4+IL-21 for 6 days as described. Indicated line is the total Ig level expected for the proband (II.2) calculated from the sum of deficits observed for each mutation alone (that is Ig level^III.2− (Ig^III.2−Ig^III.1)+(Ig^III.2−Ig^II.3)). Note that the clinical score for individual III.2 is 0.