Dominant negative variants in ITPR3 impair T cell Ca2+ dynamics causing combined immunodeficiency

Abstract

The importance of calcium (Ca2+) as a second messenger in T cell signaling is exemplified by genetic deficiencies of STIM1 and ORAI1, which abolish store-operated Ca2+ entry (SOCE) resulting in combined immunodeficiency (CID). We report five unrelated patients with de novo missense variants in ITPR3, encoding a subunit of the inositol 1,4,5-trisphosphate receptor (IP3R), which forms a Ca2+ channel in the endoplasmic reticulum (ER) membrane responsible for the release of ER Ca2+ required to trigger SOCE, and for Ca2+ transfer to other organelles. The patients presented with CID, abnormal T cell Ca2+ homeostasis, incompletely penetrant ectodermal dysplasia, and multisystem disease. Their predominant T cell immunodeficiency is characterized by significant T cell lymphopenia, defects in late stages of thymic T cell development, and impaired function of peripheral T cells, including inadequate NF-κB- and NFAT-mediated, proliferative, and metabolic responses to activation. Pathogenicity is not due to haploinsufficiency, rather ITPR3 protein variants interfere with IP3R channel function leading to depletion of ER Ca2+ stores and blunted SOCE in T cells.

Figure 1.

Identification of de novo ITPR3 variants in five unrelated patients presenting with CID and ED. (A) Pedigrees of the five unrelated patients included in this study. Females are represented by circles, males by squares, and affected individuals (P1–P5) are denoted by filled, black symbols. P4 (IV) is part of a dizygotic twinship. P5 (V) was conceived using donor sperm. Where known, genotypes are displayed below each individual. (B) Sanger sequencing of genomic DNA confirming the heterozygous ITPR3 variants found in P1–P5; REF = reference sequence, ALT = alternative sequence; representative of two independent experiments. (C) Clinical photographs of P2, P3, P4, and P5 demonstrate some of the associated syndromic features, including dental abnormalities (microdontia, hypodontia, conical teeth, and enamel hypoplasia), nail dystrophy, and deformity of the feet and hands. (D) Flowchart detailing the strategy used to identify patients with heterozygous ITPR3 variants within the rare diseases cohort of the GEL 100,000 Genomes Project (GEL100KGP); Trios = genomes of probands and parents.

Figure 2.

Expression of ITPR members in HD and patient cells. (A) RT-qPCR for ITPR1, ITPR2, and ITPR3 transcripts in peripheral blood T cells isolated from HD and patients (P2, P4, and P5). Results shown are relative to beta-actin (ACTB) expression, are representative of two independent experiments, and are not statistically significant via multiple unpaired t tests with Holm-Sidak correction for multiple comparisons. (B) Sanger sequencing of cDNA showing expression of WT and variant ITPR3 transcripts for P1 (fibroblasts), P2 (T cells), P3 (LCLs), P4 (T cells), and P5 (T cells); REF = reference sequence, ALT = alternative sequence; representative of two independent experiments. (C) Protein expression of ITPR3 in LCLs from HDs, P3 and P4. (I) Western blot for ITPR3 and GAPDH using LCL lysates; representative of three independent experiments. (II) Densitometric quantification of ITPR3 protein abundance in LCLs relative to GAPDH. (D) Western blot for ITPR3 and GAPDH in ITPR3 KO HEK293T cells (ITPR3 KO) and in transfected ITPR3 KO cells expressing WT ITPR3 protein or the A196T, I2506N, and R2524C protein variants; representative of two independent transfection experiments. Source data are available for this figure: SourceData F2.

Figure S1.

Ca²⁺ flux in patient primary skin fibroblasts, primary T cells, and gene-edited cell lines expressing patient protein variants. (A) RT-qPCR for ITPR1, ITPR2, and ITPR3 transcripts in primary skin fibroblasts obtained from three HDs and three patients (P1, P2, and P3). Results shown are relative to beta-actin (ACTB) expression are representative of two independent experiments and are not statistically significant via multiple unpaired t tests with Holm–Sidak correction for multiple comparisons. (B) Western blot for ITPR3 and GAPDH in primary skin fibroblasts from three HDs and three patients (P1, P2, and P3); representative of two independent experiments. (C–J) The ratiometric Ca²⁺ indicator Indo-1 was used to measure cytoplasmic Ca²⁺ concentration. (C) Ca²⁺ flux after stimulation with anti-CD3 plus F(ab′)2 in thawed pre-HCT primary T cells from P1 (blue) and from a HD (grey) in Ca²⁺-free media and upon addition of CaCl₂. This single, independent experiment was analyzed separately from the data shown in Fig. 4 A due to poor cell viability. (D) Ca²⁺ flux after stimulation with anti-CD3/F(ab′)2 in primary T cells from P4 (blue) and from a HD (grey) in Ca²⁺-containing media and upon addition of ionomycin at 240 s; representative of three independent experiments; not significant, Mann–Whitney U test. (E) Ca²⁺ flux after stimulation with ionomycin in Ca²⁺-free media and upon addition of CaCl₂ in CTLs from P2 and P4 (light blue) versus one HD; representative of two independent experiments. (F) Ca²⁺ flux in fibroblasts from four HDs (grey) versus P1, P2, and P3 (blue) stimulated with ATP; traces are shown in I and graph of the AUC of the peaks in II. Graphs represent the mean of three separate experiments and error bars indicate SEM. (G) Ca²⁺ flux in WT (red) and ITPR3 KO Jurkat T cells (light grey), and ITPR3 KO cells stably transduced with WT ITPR3 (dark grey) after stimulation with (I) anti-CD3/F(ab′)2 or (II) thapsigargin in Ca²⁺-free media and upon addition of CaCl₂; representative of two independent experiments. (H) Ca²⁺ flux upon addition of CaCl₂ in Ca²⁺-free media in (H) Jurkat ITPR3KO T cells expressing WT or variant ITPR3 proteins; (I) shows representative plots of ITPR3 WT (left) and variant expressing (right) cells either pre-stimulated with anti-CD3 and addition of F(ab′)2 (at timepoint A) followed by addition of CaCl₂ (at timepoint B; darker line), or CaCl₂ alone (at timepoint A; paler line); (II) shows a comparison of Ca²⁺ influx upon addition of CaCl₂ alone (i.e., without agonist stimulation) in cells expressing WT ITPR3 or the various ITPR3 variant proteins; representative of two independent experiments. (I) Ca²⁺ flux upon addition of CaCl₂ in Ca²⁺-free media in CTLs from P2, P4, and one HD; representative of two independent experiments. (J) Ca²⁺ flux in ex vivo differentiated T cells from P5 and from one HD stimulated with anti-CD3/F(ab′)2 in (I) single positive CD4⁺ and (II) single positive CD8⁺ cells; representative of two independent experiments. Source data are available for this figure: SourceData FS1.

Figure 3.

In silico modeling of ITPR3 variants. (A) ITPR3 secondary structure with domains annotated. The location of the AAs affected by dominant variants are marked by pink circles (this study i.e., A196 in P1, I2506 in P2, and R2524 in P3–P5); and yellow circles (other studies [Neumann et al., 2023; Rönkkö et al., 2020; Terry et al., 2022]: V615, T1424, F1628, and R2524); the R1850Q polymorphism is marked by a grey circle. Missense and nonsense variants in gnomAD database are indicated by blue and red vertical lines, respectively. (B) ITPR3 tertiary structure with domains colored as indicated in panel A. The center shows the monomer of the cryo-EM structure 6DRC colored by domains as predicted by Pfam, with minor corrections in the ranges, namely coupling BTF (AA 3–230), IP₃ binding BTF (AA 233–433), first IP₃ binding helix bundle (437–707), second IP₃ binding helix bundle (1175–1334), coupling helix bundle (1335–1546), central helix bundle (1864–1974), and ion channel transmembrane domain (2191–2537). The latter contains, within its S6 helix, a cationic selectivity filter (formed by N2472 and D2478), which favors the passage of Ca²⁺ over other cations, and a gate (formed by I2517 and F2513) to prevent ions from passing unless the channel is activated (Paknejad and Hite, 2018). The three insets (right) show the overlay of the WT (turquoise) and the variant (salmon) chains with other chains shown in white. (C) Multiple sequence alignment of AA sequences in orthologous ITPR proteins across different species, with the inferred phylogenetic trees displayed on the left-hand side of the panel. AA residues altered in P1–P5 are annotated at the top of the figure and marked by red boxes. (D) Boxplot of transcriptomic distance to closest gnomAD variant for the de novo variants in P1–5 (blue) and gnomAD control variants (grey); *P = 0.02, Wilcoxon rank sum test. (E) Heatmap showing ΔΔG, a calculated parameter that predicts the effect of a protein change on stability, for selected mildly destabilizing ITPR3 variants across five different conformational models of ITPR3. A positive score (red) predicts that a variant is destabilizing, whilst a negative score (blue) predicts a variant to be stabilizing. Variants shown include those identified in P1–P5 of this study; V615M, a pathogenic dominant variant identified in familial CMT disease (Rönkkö et al., 2020); W168A, an in vitro generated variant that abolishes channel activity (Chan et al., 2010; Yamazaki et al., 2010); F1628L, a variant reported in a patient with later onset, milder immune deficiency (Neumann et al., 2023); P187S and I184S, two gnomAD control variants close to the variant detected in P1; and R2471H and I2511V, two further gnomAD control variants within the TMD close to the variants detected in P2–P5.

Figure 4.

Variant ITPR3 proteins impair Ca²⁺ flux in T cells. (A–C) In primary T cells: Ca²⁺ flux after stimulation with (A) anti-CD3 plus F(ab′)2 and (B) thapsigargin (Thap) in Ca²⁺-free media and after addition of CaCl₂. The left panels show representative examples of Ca²⁺ flux in primary T cells from P4 (blue) versus one HD (grey). The right panels show the ratio between the peak (100–150 s) versus baseline (0–30 s) in P2 (pre-HCT), P4 (untransplanted), and P5 (pre-HCT) versus 10 HDs. Independent replicates are included for two patients (n = 2 for P2 and P4) and three HDs. The lines show the median and the interquartile range. Triangles and circles distinguish between frozen and fresh samples, respectively (the HD indicated with an open triangle is the mother of P2, who carries the R1850Q polymorphism). *P < 0.05, Mann–Whitney U test. (C) Baseline cytosolic Ca²⁺ levels in primary patient T cells, expressed as fold change over HD. Data represent 12 HD and 12 patient samples (P2, P4, and P5), pooled across four independent experiments; **P <0.01, paired T test. (D–E) In CTLs: Ca²⁺ flux after stimulation with (D) anti-CD3/F(ab′)2 (left panel) and thapsigargin (right panel) in Ca²⁺-free media and after addition of CaCl₂ in CTLs from P2 and P4 (blue) versus one HD (grey). This is representative of four independent experiments. (E) Ca²⁺ flux after stimulation with ionomycin in Ca²⁺-free media in CTLs from P2 and P4 (light blue) versus one HD; representative of two independent experiments. **P < 0.01, Welch’s T test. (F–H) In Jurkat T cell lines: (F) western blots for ITPR1, ITPR2, ITPR3, and GAPDH from protein extracts from WT Jurkat T cells, ITPR3 KO Jurkat T cells, and stably transduced ITPR3 KO Jurkat T cells expressing WT ITPR3 or the A196T, I2506N, and R2524C proteins; representative of two independent experiments. (G–H) The left panels show representative Ca²⁺ flux after stimulation with (G) anti-CD3 and F(ab’)2 or (H) thapsigargin in one ITPR3 KO Jurkat T cell clone (light grey) and in stably transduced ITPR3 KO Jurkat T cells (derived from the same ITPR3 KO clone) expressing WT ITPR3 (dark grey) or the A196T, I2506N, and R2524C protein variants (shades of blue). The right panel shows the ratio between the peak versus baseline in the ITPR3 KO clone and in transduced KO cells expressing WT (three clones) or variant ITPR3 proteins (three clones for A196T, two clones for I2506N, and three clones for R2524C) in five independent experiments. ns: non-significant, *P < 0.05, **P = 0.01, and ***P < 0.01, Mann–Whitney U test. Source data are available for this figure: SourceData F4.

Figure 5.

T cell phenotype in non-transplanted patients with heterozygous ITPR3 variants. (A) Distribution of maturation-associated T cell subsets in patients (P2, P4, and P5) versus three pediatric HDs displayed as (I) UMAP projections in concatenated HD (left) and patient (right) samples and (II) stacked bar charts for CD4⁺ (left) and CD8⁺ T cells (right); HD data has been combined and mean and standard deviation (error bars) are plotted. Statistical significance is displayed in the chart in the center of the figure; *P < 0.05, Mann–Whitney U test, across three independent experiments. Maturation-associated subsets are colored according to phenotype and include naïve (TN: CD45RA⁺CD27⁺CCR7⁺CD95⁻; blue), stem central memory (TSCM: CD45RA⁺CD27⁺CCR7⁺CD95⁺; red), central memory (TCM: CD45RA⁻CD27⁺CCR7⁺CD95⁺; grey), transitional memory (TM: CD45RA⁻CD27⁺CCR7⁻CD95⁺; dark green), effector memory (TEM: CD45RA⁻ CD27⁻CCR7⁻CD95⁺; lilac), and terminally differentiated (TD: CD45RA⁺CD27⁻CCR7⁻CD95⁺; orange) CD4⁺ and CD8⁺ T cells, as well as effector CD27^dim (CD45RA⁺CD27^dimCCR7⁻CD95⁺; light green) CD8⁺ T cells. (B) Bar chart showing the percentage of CD31⁺ RTEs within naïve CD4⁺ T cells in patients versus pediatric HDs. (C) Bar chart (right) showing the percentage of PD-1⁺ cells in CD4⁺ and CD8⁺ T cells in patients versus pediatric HDs; representative histograms (left) are shown for P5 and one HD. (D) Bar chart (right) showing the percentage of CD8⁺ T cells positive for (I) granzyme B and (II) perforin in patients versus pediatric HDs. Representative histograms (left) are shown for P2 and one HD. (E) Representative TCRVβ spectratyping shown for P4 on isolated CD3⁺ T cells. All Vβ families are represented; all but one (TRBV30) have a skewed and sparse profile with oligoclonal expansions present and a median number of seven peaks. (F) Bar chart showing absolute numbers of Tregs in patients versus pediatric HDs. (G) Bar chart (right) showing the percentage of iNKT cells within T cells in patients versus pediatric HDs. Representative flow cytometry plots (left) for P5 and one HD are displayed. (H) Representative flow cytometry plots for P2 and one HD showing the percentage of TCR γδ⁺ and ⍺β⁺ cells with total CD3⁺ T cells. For B–D and F–G, representative of three independent experiments, bars represent the median values and lines and the interquartile range. *P < 0.05, Mann–Whitney U test. All statistics were performed using bilateral t tests. n/a, not available.

Figure S2.

Gating strategies for flow cytometric analyses and sorting. (A) FACS gating employed for (A) ratiometric analysis of cytoplasmic Ca²⁺ flux in primary T cells. (B) Immunophenotyping of T cells. TN: naïve; TSCM: stem cell memory; TCM: central memory; TM: transitional memory; TEM: effector memory; TD: terminally differentiated; CD27^dim: effector CD27^dim. (C) Enumeration of Tregs.

Figure 6.

Heterozygous ITPR3 variants primarily cause a hematopoietic defect correctable via HCT. (A) Graphs showing percentage of PBMCs undergoing apoptosis (AnnexinV⁺7AAD⁻) in patients (P2: two to three independent replicates; P4: one replicate) versus HDs at (I) baseline, representing apoptosis determined on day 0 in unstimulated PBMCs; this is shown alongside (II) fold change in percentage apoptotic cells in stimulated versus unstimulated PBMCs after culture with anti-CD3 and IL-2 for 6–7 days, followed by stimulation with PHA or anti-FAS IgM. Bars represent the media and lines the interquartile range; results are not statistically significant via multiple unpaired t tests with Holm–Sidak correction for multiple comparisons. (B and C) Flow cytometry plots showing proportions (within live CD45⁺ cells) of CD4⁻CD8⁻ double negative (DN) T cells, CD4⁺CD8⁺ DP T cells, CD3⁺TCRαβ⁺, and CD3⁺TCRγδ⁺ T cells (B) after 6 wk of ex vivo T cell differentiation of CD34⁺ cells isolated from P2 (bone marrow) and from one HD (cord blood) in a two-dimensional (2D) co-culture with OP9/DL1 stromal cells, and (C) after 6 wk of differentiation of CD34⁺ cells isolated from P5 (bone marrow) and from one HD (cord blood) in a three-dimensional (3D) co-culture with MS5/DL1 stromal cells in ATOs. All co-cultures were set up with at least three replicates in each experiment. (D) Distribution of maturation-associated T cells subsets displayed in stacked bar charts for CD4⁺ (left) and CD8⁺ T -cells (right) in P1, P2, P3, and P5 post-HCT at last follow-up versus adult (Ad.) and pediatric (Ped.) HDs; HD data has been combined and mean and standard deviation (error bars) are plotted. (E–G) Representative flow cytometry plots showing recovery of (E) Treg, (F) NKT, and (G) γδ T cells in patients post-HCT; representative of two to three independent experiments.

Figure 7.

T cells with heterozygous ITPR3 variants are impaired in their ability to express NF-κB and NFAT-dependent genes, proliferate, and upregulate metabolic enzymes in response to stimulation. (A) Graphs showing the percentage of CD25⁺ cells in CD4⁺ (left) and CD8⁺ T cells (right) stimulated with anti-CD3/CD28 compared with unstimulated for P2, P4, and two HDs across two independent experiments. (B) Nuclear translocation of NFAT after TCR engagement in CTLs from P2, P4, and two HDs in three independent experiments. The left panel shows representative images of (I) P185 target cells (yellow) and stimulated CD8⁺ CTLs (purple), (II–IV) intracellular staining for NFAT (blue) and nuclear staining with DAPI (red), and (V) a composite image illustrating the mildly reduced nuclear NFAT translocation in P4 compared to HD; scale bar = 5 μm. The right panels show the ratios between nuclear NFAT and cytoplasmic NFAT measured in MFI in unstimulated and stimulated CTLs from P2, P4 and the HDs in all three experiments; ns, non-significant, **P <0.01 and ***P <0.001, one-way ANOVA with Šídàk’s multiple comparisons test. (C) Genes activated by NF-κB and NFAT are downregulated in patient T cells compared with HD following anti-CD3/CD28 stimulation for 4 or 12 h (I) Graph showing normalized enrichment scores (NES) for NF-κB and NFAT transcription factor (TF) binding motifs within 5 Kb of the transcriptional start sites of genes significantly upregulated in HD versus patient T cells after 4 h of stimulation; filled symbols denote NF-κB motif analysis: square=bergman_dif_Rel, circle=homer_GGGGGAATCCCC_NFkB-p50_p52, triangle=transfac_pro_M01223; empty symbols denote NFAT motif analysis: square=taipale_NFATC1_full_NATGGAAANWWWWTTTYCMN_repr, circle=taipale_NFATC1_full_TTTTCCATGGAAAA_repr, triangle=cisbp_M5658. (II and III) Heatmaps showing expression level of genes significantly upregulated in HD versus patient T cells following stimulation that contain TF binding motifs for (II) NF-κB and (III) NFAT. (D) Bar chart showing proliferation of PBMCs from P1 (pre-HCT), P2 (pre-HCT), P4, and P5 (pre-HCT) versus HDs after stimulation with anti-CD3, PHA, or PMA and ionomycin. Proliferation was assessed by measuring the incorporation of titrated (³H) thymidine and quantified as counts per minute (cpm) over background across six independent experiments. Bars represent the median values and lines the interquartile range; *P < 0.05, **P < 0.01, t test with Sidak-Bonferonni correction for multiple comparisons. (E) Graphs showing expression (gMFI) of metabolic enzymes in CD4⁺ (left) and CD8⁺ T cells (right) stimulated with anti-CD3/CD28 compared with unstimulated cells for P2, P4, and two HDs across two independent experiments. (F) Gene set enrichment analyses of differentially expressed genes identified by RNA-sequencing of patients (P2, P4, and P5) and HD T cells (n = 3) stimulated with anti-CD3/28. Enrichment scores (ES) and P values are displayed for the following gene sets: (I) fatty acid metabolism, (II) tricyclic acid (TCA) enzyme complex, (III) mitochondrial biogenesis, (IV) mitochondrial genes, (V) protein import into the mitochondrial matrix, and (VI) bound by FOXP3.

Figure S3.

Functional T cell assays in patients with heterozygous ITPR3 variants. (A and B) Graphs showing the percentage of (A) CD25⁺ cells within CD4⁺ T cells and (B) CD40L⁺ cells within activated CD25⁺ CD4⁺ T cells after stimulation with PHA in patients (P2: three replicates; P4: two replicates) versus HDs, across three independent experiments; **P < 0.01, two-tailed t test. (C and D) Graphs showing interferon γ (IFNγ), tumor necrosis factor α (TNFα), and interleukin (IL)-2 production in patients (P2, P4, and P5) versus HD T cells after stimulation with (C) anti-CD3/28 and (D) PMA/Ionomycin. Results are expressed as a percentage of cytokine⁺ cells within CD4⁺ and CD8⁺ T cells after excluding naïve cells. (E) Degranulation capacity of CD8⁺ T cells after stimulation with anti-CD3/CD28. Results are expressed as a percentage of CD107a⁺ cells within total CD8⁺ T cells, across three independent experiments, two-tailed t test.