Dominant-negative mutations in human IL6ST underlie hyper-IgE syndrome

Abstract

Autosomal dominant hyper-IgE syndrome (AD-HIES) is typically caused by dominant-negative (DN) STAT3 mutations. Patients suffer from cold staphylococcal lesions and mucocutaneous candidiasis, severe allergy, and skeletal abnormalities. We report 12 patients from 8 unrelated kindreds with AD-HIES due to DN IL6ST mutations. We identified seven different truncating mutations, one of which was recurrent. The mutant alleles encode GP130 receptors bearing the transmembrane domain but lacking both the recycling motif and all four STAT3-recruiting tyrosine residues. Upon overexpression, the mutant proteins accumulate at the cell surface and are loss of function and DN for cellular responses to IL-6, IL-11, LIF, and OSM. Moreover, the patients' heterozygous leukocytes and fibroblasts respond poorly to IL-6 and IL-11. Consistently, patients with STAT3 and IL6ST mutations display infectious and allergic manifestations of IL-6R deficiency, and some of the skeletal abnormalities of IL-11R deficiency. DN STAT3 and IL6ST mutations thus appear to underlie clinical phenocopies through impairment of the IL-6 and IL-11 response pathways.

Graphical abstract

Figure 1.

AD IL6ST deficiency. (A) Pedigree of the eight unrelated families showing familial segregation of the c.2277_2281dup (p.T761fs) mutant IL6ST allele in kindred A, the c.2155dup (p.I719fs) mutant IL6ST allele in kindreds B and C, the c.2199C>A (p.C733*) mutant IL6ST allele in kindred D, the c.2121del (p.L708*) mutant IL6ST allele in kindred E, the c.2277T>G (p.Y759*) mutant IL6ST allele in kindred F, the c.2224dup (p.S742fs) mutant IL6ST allele in kindred G, and the c.2261C>A (p.S754*) mutant IL6ST allele in kindred H. M, mutant. Individuals of unknown genotype are labeled “E?”. The father of P12, who died at 26 yr of age, is shown in gray because he suffered from a phenotype partially compatible with HIES (asthma, multiple infections, rheumatoid arthritis, and diverticulitis). (B) Representative photographs and scanner images of patients, with scoliosis (P5), lung pneumatocele (P5, P6, P9, and P10), bronchomalacia (P3, red arrow), cold abscess (P8, black arrow), and sinusitis (P9, green arrow). (C) Schematic representation of GP130 and population genetics of IL6ST alleles. All reported predicted heterozygous “LOF” alleles and their positions in GP130 are indicated. The L708*, I719fs, C733*, S742fs, S754*, Y759*, and T761fs mutations found in the patients under study are colored in red. Mutants reported in gnomAD are colored in black, green, or blue. Two variants from gnomAD with a premature stop codon before the transmembrane domain were selected for subsequent functional analysis and are colored in green (E316* and T555fs). Three variants from gnomAD with a premature stop codon lacking one (G913fs), two (E899*), or three (S789*) STAT3-binding residues were selected for subsequent functional analysis and are colored in blue. EC, extracellular domain; IC, intracellular domain; SP, signal peptide; TM, transmembrane domain.

Figure 2.

The I719fs and T761fs IL6ST mutants accumulate at the cell surface and are LOF. (A) GP130-deficient HEK293T cells were transfected with an empty pCMV6 plasmid (EV) or with pCMV6 plasmids encoding the WT or the E316*, T555fs, I719fs, T761fs, S789*, E899*, or G913fs GP130 mutants. Total protein was extracted, left untreated (top), or treated (bottom) for 1 h with PNGase F to eliminate N-glycosylation, and subjected to immunoblotting with a mAb against GP130 (amino acids 365–619, clone E8). GAPDH was used as a loading control. Western blots representative of three independent experiments are shown. (B and C) GP130-deficient HEK293T cells were transfected as described in A. After 48 h of incubation, cells were harvested, stained for extracellular GP130 expression, and analyzed by flow cytometry. An image showing representative staining (B) and a recapitulative graph depicting GP130 mean fluorescence intensity (MFI; C) are shown. The bars and error bars represent the mean of three different experiments and the standard deviation, respectively. (D) GP130-deficient HEK293T cells were transfected as described in A. After 24 h of incubation, the cells were stimulated for 15 min with IL-6/IL-6Rα or IFN-α, or left unstimulated, and the phosphorylation of STAT1 (pY701) and of STAT3 (pY705) was then evaluated. Representative results from three independent experiments are shown. (E) Luciferase assay to assess STAT3 activity. GP130-deficient HEK293T cells were transfected with an empty pCMV6 vector (EV) or a plasmid encoding the WT or the indicated GP130 mutants, plus a pGL4.47 reporter plasmid carrying the luciferase cDNA downstream from five SIEs. After 24 h, cells were stimulated with the indicated cytokine (+) or were left unstimulated (−) for another 24 h before the measurement of luciferase activity. The horizontal dotted line indicates the luciferase activity after stimulation, with the indicated cytokine, of the cells transfected with the empty pCMV6 vector. The results shown are the mean and standard error of the mean for a technical duplicate. Luciferase assays representative of three independent experiments are shown.

Figure S1.

Expression of the GP130 mutants and phosphorylation assays of STAT1 and STAT3. (A) GP130-deficient HEK293T cells were transfected with an empty pCMV6 plasmid (EV) or with pCMV6 plasmids encoding the WT, L708*, C733*, S742fs, S754*, or Y759* GP130 mutants. Total protein was extracted and subjected to immunoblotting with a mAb against GP130 (amino acids 365–619, clone E8). GAPDH was used as a loading control. Western blots representative of three independent experiments are shown. (B and C) GP130-deficient HEK293T cells were transfected as described in A. After 48 h, the cells were harvested, stained for extracellular GP130, and analyzed by flow cytometry. Representative staining (B) and a recapitulative graph of GP130 mean fluorescence intensity (C) are shown. The bars and error bars represent the mean of three independent experiments and the standard deviation, respectively. (D and E) GP130-deficient HEK293T cells were transfected with an empty pCMV6 plasmid (EV) or with pCMV6 plasmids encoding the WT, or E316*, T555fs, L708*, I719fs, C733*, S742fs, S754*, Y759*, T761fs, S789*, E899*, G913fs GP130 mutants. After 24 h of incubation, the cells were stimulated for 15 min with the indicated GP130-dependent cytokines (red line) or left unstimulated (black line), and the phosphorylation of STAT3 (pY705; D) and STAT1 (pY701; E) was then evaluated by flow cytometry. The results obtained with the I719fs and T761fs mutants are representative of the results obtained with the other mutants found in HIES patients (L708*, C733*, S742fs, S754*, and Y759*). Representative results from three independent experiments are shown.

Figure S2.

Molecular characterization of the GP130 mutants. (A) STAT3 activity, as assessed in a luciferase assay. GP130-deficient HEK293T cells transfected with an empty pCMV6 vector (EV) or with a plasmid encoding the WT or the indicated GP130 mutant plus a pGL4.47 reporter plasmid carrying the luciferase cDNA downstream from five SIEs. Cells were stimulated with the indicated cytokine (+), 24 h after transfection, or were left unstimulated (−) for another 24 h before the measurement of luciferase activity. The horizontal dotted line indicates the luciferase activity after the stimulation, with the indicated cytokine, of cells transfected with the empty pCMV6 vector. The results shown are the mean and standard error of the mean for a technical duplicate. A luciferase assay representative of two independent experiments is shown. (B) STAT3 activity, as assessed in a luciferase assay, after IFN-α stimulation. GP130-deficient HEK293T cells were transfected as described in A. Cells were stimulated with IFN-α (+), 24 h after transfection, or were left unstimulated (−) for another 24 h before the measurement of luciferase activity. The horizontal dotted line indicates the luciferase activity after the stimulation, with the indicated cytokine, of cells transfected with the empty pCMV6 vector. The results shown are the mean and standard error of the mean for a technical duplicate. A luciferase assay representative of two or three independent experiments is shown. (C) GP130-deficient HEK293T cells were transfected with an empty pCMV6 plasmid (EV) or with pCMV6 plasmids encoding the WT or I719fs, S754*, and T761fs GP130 mutants. After 24 h of incubation, the cells were starved of serum for 16 h and then stimulated for 15 min with the indicated GP130-dependent cytokines or left unstimulated, and the phosphorylation of ERK1/2 (pT202/pY204; top panel) or STAT3 (Y705, positive control; lower panel) was evaluated by flow cytometry. Representative results from two independent experiments are shown.

Figure 3.

The I719fs and T761fs GP130 mutants are DN over the WT GP130. GP130-deficient HEK293T cells transfected with a pCMV6 EV or encoding the WT GP130 (25 ng) and various concentrations (25–200 ng) of pCMV6 vector encoding the indicated GP130 mutants (Mut) plus a pGL4.47 reporter plasmid carrying the luciferase cDNA downstream from five SIEs. GP130-deficient HEK293T cells transfected with only the WT (25 ng) or the mutant GP130 (200 ng) were used as controls. Cells were stimulated, 24 h after transfection, with the indicated cytokine (+) or were left unstimulated (−) for another 24 h before the measurement of luciferase activity. The results shown are the mean and standard error of the mean for a technical duplicate. Luciferase assays representative of two independent experiments are shown.

Figure S3.

Assay of the negative dominance of the GP130 mutants. (A–G) GP130-deficient HEK293T cells transfected with an empty pCMV6 vector or a vector encoding the WT GP130 (25 ng) and various amounts (25–200 ng) of pCMV6 vector encoding the indicated GP130 mutant (Mut) plus a pGL4.47 reporter plasmid carrying the luciferase cDNA downstream of five SIEs. GP130-deficient HEK293T cells transfected with only the WT (25 ng) or the mutant GP130 (200 ng) were used as controls. Cells were stimulated with the indicated cytokine 24 h after transfection or were left unstimulated for another 24 h before the measurement of luciferase activity. The results shown are the mean and standard error of the mean for a technical duplicate. Luciferase assays representative of two independent experiments are shown.

Figure 4.

DN mutants hijack IL-6R. GP130-deficient HEK293T cells transfected with an empty pCMV6 vector or a vector encoding the WT GP130 (25 ng) and a pCMV6 vector encoding the I719fs GP130 mutant (25 ng) plus a pGL4.47 reporter plasmid carrying the luciferase cDNA downstream from five SIEs. GP130-deficient HEK293T cells transfected with only the WT (25 ng) or the mutant GP130 (25 ng) were used as controls. A pCMV6 vector encoding IL-6R (25 ng) was also used for cotransfection, where indicated. Cells were stimulated with IL-6 or IL-6 plus IL-6Rα, as indicated, 24 h after transfection or were left unstimulated for another 24 h before the measurement of luciferase activity. The results shown are the mean and standard error of the mean for a technical duplicate. Luciferase assays representative of two independent experiments are shown.

Figure 5.

GP130 production and activity in the patients’ primary fibroblasts. (A and B) GP130 levels in primary fibroblasts, as evaluated by flow cytometry. (A) Representative traces of GP130 expression in primary fibroblasts from P5 and a healthy control are shown, together with traces for the isotypic control. (B) The graph shows the MFI of GP130 in five independent controls, and two patients (P2 and P5). The bars and error bars for the patients represent the mean of a duplicate and the standard deviation, respectively. (C) Total protein was extracted from primary fibroblasts and subjected to immunoblotting with a mAb against GP130 (amino acids 365–619, clone E8). GAPDH was used as a loading control. Western blots representative of three independent experiments are shown. (D) Primary fibroblasts from P2 and P5 and three healthy controls (Ctl, only one shown) were stimulated (Stim) for 15 min with IL-6, IL-6/IL-6Rα, IL-11, LIF, OSM, and IFN-α or left unstimulated (NS), and the phosphorylation of STAT3 (pY705, top panel) and STAT1 (pY701, bottom panel) was then evaluated by flow cytometry. FACS plots representative of two independent experiments are shown.

Figure S4.

Primary fibroblasts. (A) Kinetics of GP130 degradation. Primary fibroblasts from P6, P7, and one healthy control were incubated for the times indicated (30 min to 4 h) with cycloheximide (CHX) to inhibit protein synthesis. GP130 levels were then evaluated by flow cytometry. The expression levels shown are normalized relative to the nonstimulated time point (100%). The means and standard errors of the means from three independent experiments are shown. (B) p-STAT3 in P9. Primary fibroblasts from P9 and one healthy control (Ctl) were stimulated (Stim) for 15 min with IL-6, IL-11, LIF, OSM, and IFN-α or left unstimulated (NS), and the phosphorylation of STAT3 (pY705) was then evaluated by flow cytometry.

Figure 6.

Serum CRP levels suggest normal acute phase in patients. CRP kinetics of P5, P6, and P10, and CRP levels in P5, P6, and P10 during the follow-up period. Arrows indicate infections.

Figure 7.

Leukocyte immunophenotyping. (A) Frequencies of monocyte subsets, as assessed by measuring the expression of CD16 and CD14, for controls (n = 23; C) and patients (P) with AD IL6ST mutations (n = 6). ns, not significant. (B) Frequencies of cDC2 (Lin⁻HLA-DR⁺CD11c⁺CD1c⁺CD141⁻), cDC1 (Lin⁻HLA-DR⁺CD11c⁺CD1c⁻CD141⁺), and plasmacytoid DCs (Lin⁻HLA-DR⁺CD11c⁻CD123⁺) among the PBMCs of controls (n = 32) and patients with AD IL6ST mutations (n = 6). (C) Frequency of CD27⁺ memory cells within the B cell compartment of controls (n = 60) and patients with AD IL6ST mutations (n = 6). (D) Frequency of IgM⁺, IgA⁺, and IgG⁺ cells within the memory B cell compartment of controls (n = 39–60) and patients with AD IL6ST mutations (n = 6). (E–G) NK cell immunophenotyping for controls (n = 58–60) and patients with AD IL6ST mutations (n = 6), showing the frequency of CD56^bright cells within the NK-cell compartment (E), the terminal differentiation profile of the CD56^dim compartment (F), and the frequency of NKG2C⁺NKG2A⁻ cells within the CD56^dim compartment (G). (H and I) Frequency of naive (CD45RA⁺CCR7⁺), central memory (CD45RA⁻CCR7⁺), effector memory (CD45RA⁻CCR7⁻), and T_EMRA (CD45RA⁺CCR7⁻) cells among the CD4⁺ (H) and CD8⁺ (I) T cells of controls (n = 65) and patients with AD IL6ST mutations (n = 7). (J) Frequency of γδ T cells (CD3⁺TCR-γδ⁺), MAIT (CD3⁺CD161⁺TCR-vα7.2⁺), and iNKT (CD3⁺TCR-iNKT⁺) cells among the T cells of controls (n = 31–59) and patients with AD IL6ST mutations (n = 6). (K) Frequency of T reg (CD3⁺CD4⁺CD25^hiFoxP3⁺) cells in the CD4⁺ T cell compartment of controls (n = 59) and patients with AD IL6ST mutations (n = 7). (L) Frequency of Th subsets within the CD4⁺ memory compartments of controls (n = 60–64), and patients with AD IL6ST mutations (n = 7). Subsets were defined as follows: Tfh (CXCR5⁺), Th1 (CXCR5⁻CXCR3⁺CCR4⁻CCR6⁻), Th2 (CXCR5⁻CXCR3⁻CCR4⁺CCR6⁻), Th1* (CXCR5⁻CXCR3⁺CCR4⁻CCR6⁺), and Th17 (CXCR5⁻CXCR3⁻CCR4⁺CCR6⁺). (A–L) t tests were used for all comparisons.

Figure 8.

In vitro functional assays of CD4⁺ Th cells. (A–C) Secretion (ng/ml) of IL-2, Th1 (IFN-γ and TNF; A), Th2 (IL-4, IL-5, and IL-13; B), and Th17 (IL-17A, IL-17F, and IL-22; C) cytokines by memory CD4⁺ T cells after 5 d of stimulation with anti-CD2/CD3/CD28 mAb-coated beads. C, control; P, patients. (D) Frequency of Th17 (IL-17A, IL-17F, and IL-22) cytokine-positive memory CD4⁺ T cells after 4 d of culture under Th0 conditions (anti-CD3/CD2/CD28 antibody-coated beads). (E) Secretion (ng/ml) of Th17 (IL-17A and IL-17F) cytokines by naive CD4⁺ T cells after 5 d of culture under Th0 cell–polarizing conditions (anti-CD2/CD3/CD28 mAb-coated beads) or Th17 cell–polarizing conditions (anti-CD2/CD3/CD28 mAb-coated beads together with IL-1β, IL-6, IL-21, IL-23, and TGF-β). Mann–Whitney tests were used for all comparisons. ns, not significant. *, P < 0.05; **, P < 0.01.

Figure 9.

GP130 production and activity in primary lymphocytes and monocytes from patients. (A–E) GP130 levels, as evaluated by flow cytometry, in primary CD56^bright and CD56^dim NK cell subsets (A), naive and memory B cells (B), naive and memory CD8⁺ T cells (C), naive and memory CD4⁺ T cells (D), and monocytes (E). The graphs show the MFI of GP130, as measured by flow cytometry, in eight controls (C) and six patients (P2–P7). Representative data for GP130 expression in naive B cells (B) and monocytes (E) from P2 and a healthy control are shown, together with data for the isotypic control. (A–E) t tests were used for all comparisons. (F) PBMCs from P6 (red lines) and a control (black lines) were stimulated for 15 min with IL-6, IL-6/IL-6Rα, IL-27, or IFN-α or left unstimulated, and the phosphorylation of STAT3 (pY705) and STAT1 (pY701) was then evaluated in the indicated subsets. (G) Purified monocytes from P5 (red lines) and a control (black lines) were stimulated for 15 min with IL-6, IL-6/IL-6Rα, IL-27, or IFN-α or left unstimulated, and the phosphorylation of STAT3 (pY705) and STAT1 (pY701) was then evaluated. ns, not significant.

Figure S5.

GP130 expression and function in PBMCs. (A) GP130 levels in EBV-B cell lines from two controls (C1 and C2) and P8. Representative data are shown on the left and are summarized in a graph on the right. (B) PBMCs from P5 (red lines, left plots), P6 (red lines, right plots), and two controls (black lines) were stimulated for 15 min with IL-6, IL-6/IL-6Rα, IL-27, or IFN-α or left unstimulated, and the phosphorylation of STAT3 (pY705) was then evaluated in the indicated subsets. (C) PBMCs of P2-P4 (red lines) and controls (black lines) were stimulated for 15 min with IL-6, IL-6/IL-6Rα, IL-27, IFN-α, IL-10, or IL-21 or left unstimulated, and the phosphorylation of STAT3 (pY705) was then evaluated in B cells. (D) EBV-B cell lines from P8 and a control were stimulated for 15 min with IL-6, IL-27, or IL-21 or left unstimulated, and the phosphorylation of STAT3 (pY705) was then evaluated.

Figure 10.

IL6ST DN mutation mechanism of action. (A) At a 1:1 Mut:WT ratio, truncated GP130 is predicted to impair 75% of GP130 downstream signaling hexamers (IL-6/IL-6R/GP130 or IL-11/IL-11RA/GP130) and 50% of GP130 downstream signaling trimers (IL-27/IL-27RA/GP130, LIF/LIF-R/GP130, or OSM/OSM-R/GP130). (B) The accumulation of mutant GP130 is predicted to have a stronger impact on IL-6 family cytokines forming hexamers (IL-6 and IL-11) than on those forming trimers (IL-27, LIF, and OSM) with GP130. (C) Mechanism of GP130 accumulation at the cell surface. Due to the missing recycling motif, truncated GP130 is not recycled in the basal state, unlike WT GP130. As a result, in heterozygous individuals, truncated GP130 is predicted to accumulate at the cell surface, whereas WT GP130 is predicted to be stably expressed due to a balance between de novo production and proteosomal degradation.