Severe Allergic Dysregulation Due to A Gain of Function Mutation in the Transcription Factor STAT6

Abstract

Background.

Inborn errors of immunity (IEI) have been implicated in causing immune dysregulation, including allergic diseases. The signal transducer and activator of transcription 6 (STAT6) is a key regulator of allergic responses.

Objective.

We sought to characterize a novel gain-of-function (GOF) STAT6 mutation identified in a child with severe allergic manifestations.

Methods.

We performed whole-exome and targeted gene sequencing, lymphocyte characterization, and molecular and functional analyses of mutated STAT6.

Results.

We report a child with a missense mutation in the DNA binding domain of STAT6 (c.1114G>A, p.E372K) who presented with severe atopic dermatitis, eosinophilia and elevated IgE. Naive lymphocytes from the affected patient displayed increased T_H2 and suppressed T_H1 and T_H17 cell responses. The mutation augmented both basal and cytokine-induced STAT6 phosphorylation without affecting dephosphorylation kinetics. Treatment with the Janus kinase 1/2 inhibitor ruxolitinib reversed STAT6 hyperresponsiveness to IL-4, normalized T_H1, and T_H17, suppressed the eosinophilia and improved the patient’s atopic dermatitis.

Conclusions.

We identified a novel IEI due to STAT6 GOF mutation that gave rise to severe allergic dysregulation. Janus kinase inhibitor therapy could represent an effective targeted treatment for this disorder.

Graphical Abstract

Capsule summary.

Baris et al identify a novel inborn error of immunity disease due to a STAT6 gain of function mutation that results in severe allergic dysregulation and which was effectively managed by Janus kinase inhibitor therapy.

Introduction

A number of monogenic inborn errors of immunity (IEI) give rise to allergic dysregulation as a prominent disease manifestation^{1, 2}. Originally, heterozygous loss-of-function mutations in the signal transducer and activator of transcription 3 (STAT3) were first identified as the cause of autosomal dominant form of the hyper IgE syndrome (HIES)³. Afterward, Dedicator of Cytokinesis 8 (DOCK8) deficiency was discovered to cause the autosomal recessive form of the disease^{4, 5}. Subsequently, other genetic causes of HIES have been identified, encompassing mutations in PGM3, ZNF341, IL6ST, IL6R, ERBIN, TGFBR1, TGFBR2 and CARD11 genes^6–10. Recently, pertaining to IEI disorders with atypical allergic manifestations, the term of primary atopic disorders was emerged to unify these disorders under one umbrella^{9, 11}, encompassing more than 35 monogenetic disorders^{2, 8}. Gain-of-function (GOF) mutations in Janus kinase (JAK)-STAT signaling pathways have also been associated with atopic manifestations¹², including ones targeting STAT3, STAT5B and JAK1^13–16.

STAT6 is the central transcription factor that mediates the biological effects of interleukin 4 (IL-4) and IL-13^17–19. The binding of IL-4 and IL-13 to IL-4R complexes triggers the phosphorylation of conserved tyrosine residues in the cytoplasmic domain of the common IL-4 receptor alpha (IL-4Rα) subunit by the receptor-associated JAK kinases^{20, 21}. STAT6 is subsequently recruited through the binding of its tandem Src homology 2 (SH2) domains to the IL-4Rα phosphotyrosine docking sites²². Once bound to the receptor, STAT6 is phosphorylated by the JAK kinases, resulting in its dimerization via the SH2 domains and its translocation into the nucleus, where it activates the expression of target genes by binding to a specific DNA sequence motif^{23, 24}.

The IL-4R-STAT6 axis plays a critical role in type 2 immunity, directing protective responses to parasites and toxins, promoting B cell development, activation and class switching to IgE and IgG1, and directing overall tissue repair^{22, 25, 26}. By the same token, its dysregulation plays a pathogenic role in different allergic diseases²⁷. STAT6 signaling promotes the differentiation of T helper type 2 (T_H2) cells by upregulating the expression of the master transcriptional regulator GATA binding protein 3 (GATA3)^{28, 29}. Furthermore, STAT6 signaling renders T_H2 cells resistant to regulatory T (Treg) cell-mediated suppression³⁰, while excessive STAT6 signaling in Treg cells results in their acquisition of a T_H2 cell-like phenotype that contributes to the allergic response³¹. STAT6 single nucleotide polymorphisms are associated with increased serum IgE, eosinophil, food allergy, and asthma^{32, 33}. Exaggerated STAT6 activity due to somatic STAT6 mutations is often linked to malignancies, especially Hodgkin and non-Hodgkin lymphomas^{34, 35}. A transgenic mouse model (STAT6VT) expressing a lymphoid tissue-restricted constitutively active STAT6 GOF variant created by introducing mutations in the SH2 domains exhibited IL-4-independent augmented T_H2-cell response. These mice developed atopic dermatitis (AD)-like phenotype with blepharitis and allergic lung inflammation^{36, 37}. In line with the known critical function of IL-4R-STAT6 axis in the initiation and promotion of T_H2 responses, we herein report a patient with germline sporadic GOF mutation in STAT6 as the molecular etiology of a previously unknown syndromic HIES. We showcase the impact of perturbations in T-cell responses downstream of altered STAT6 signaling and deliver a proof-of-concept for improving immune cell activation and differentiation in patient cells by Janus Kinase inhibitors (Jakinibs).

Materials and Methods

Clinical assessments.

The local ethics committee from Marmara University and Boston Children’s Hospital approved the clinical and research studies protocol and written informed consent was obtained from the family. We documented the clinical and demographic features of the patient and provided data regarding long-term follow-up.

Antibodies and flow cytometry.

To determine detailed lymphocyte subsets, the following monoclonal antibodies (mAbs) were used: Fluorescein isothiocyanate (FITC)-conjugated CD3 (UCHT1, BC, FRA), Allophycocyanin (APC)-Alexa Fluor 700 (APC-A700) CD4 (13B8.2, BC), Krome Orange (KO) CD45 (J33, BC), Alexa Fluor 750 (APC-A750) CD45RA (2H4DH11LDB9, BC), Phycoerythrin-Cyanin 7 (PC7) CD8 (SFCI21Thy2D3, BC), Phycoerythrin-Texas Red-x (ECD) CD45RO (UCHL1, BC), PC5.5 CD25 (B1.49.9, BC), Alexa Fluor 647 Foxp3 (259D, BD Biosciences, Calif, USA), PE P-STAT6 (A15137E, Biolegend), FITC “Chemoattractant receptor homologous molecule expressed on TH2-type cells” (CRTH2) (BM16, Biolegend), APC-GATA binding protein 3 (GATA3) (16E10A23, Biolegend), BV605 CCR4 (L291H4, Biolegend), PE RORγt (Q31–378, BD Biosciences), CD27 (0323, Biolegend), CD19 (4G7, Biolegend), CD23 (EBVCS-5, Biolegend), IgE (MHE-18, Biolegend), IgD (IA6–2, Biolegend), CD38 (HB-7, Biolegend), CD20 (2H7, Biolegend). Peripheral blood lymphocyte subset analyses, upregulation, and proliferation assays were performed by flow cytometry as described previously^38–40. For lymphocyte subset analysis, 100μl of whole blood was incubated with mAbs against surface markers for 20 minutes in the dark at room temperature. Red cells were lysed and washed before acquisition. For Treg cells visualization, after isolation of peripheral blood mononuclear cells (PBMCs), the cells were fixated and permeabilized, followed by overnight incubation for intracellular Foxp3 staining. All stained cells were acquired with a Navios EX cytometer (Beckman Coulter) and analyzed with Kaluza Analysis Software (Version 2.1). CD25 upregulation and proliferation assay were performed by isolation of peripheral blood mononuclear cells and stimulation in anti-CD3/anti-CD28 (1μg/ml each) 96-well plates for 3 days, following labeling with CellTrace Violet (Thermo Fisher) according to manufacturer’s instructions. CD69 upregulation was evaluated after 24 hours of stimulation with anti-CD3/anti-CD28 (1μg/ml each). For intracellular IL-4, IL-17A, IL-10 and IFN-γ detection, cell suspensions (1×10⁶ cells) were incubated with protein transport inhibitor containing monensin (BD Bioscience, USA), PMA (50 ng/ml) and ionomycin (1 μg/ml) for 6 hours. After incubation, cells were fixed. For intracellular cytokine staining, fixed cell pellets were resuspended in permeabilization reagent containing saponin (Thermo Fisher Scientific, USA), mixed with fluorescent-labeled cytokine specific antibody. After 45 minutes of incubation, cells were washed twice and cell surface staining was performed with fluorescent-labeled antibodies against CD4, CD45RA and CD45RO. Stained cells were acquired by Navios EX cytometer (Beckman Coulter) and analyzed by FlowJo software (TreeStar, Ashland, Ore).

Analysis of phosho(p)-STATs by phosphoflow.

Total PBMCs from either Healthy controls or Patient pre and post-treatment were stimulated at 37°C in non-supplemented RPMI 1640 with IL-4 (20 ng/ml), IL-2 (10 ng/ml) or IL-6 (20 ng/ml). For the STAT6 dephosphorylation following cytokine withdrawal, cells were washed with RPMI after stimulation with IL-4 and incubated in RPMI at 37°C and examined for p-STAT6 levels at 0 to 30 min, as previously described for STAT1 GOF mutations^{41, 42}. Reactions were stopped and cells were permeabilized using a Foxp3/transcription factor staining buffer (eBiosciences) and Perm buffer III (BD Biosciences). Cells were stained using PE anti pY641-STAT6 (A15137E, Biolegend), APC anti-pY694-STAT5 (A17016B.Rec, Biolegend), Efluor450 Foxp3 (236A/E7, ThermoFischer), APC-Cy7 CD3 (OKT3, Biolegend), BV785 CD4 (OKT4, Biolegend), and/or BV605 CD8 (HIT8a, Biolegend) mAbs. Samples were acquired on a Fortessa cytometer (BD) and data were analyzed using the FlowJo software.

In vitro B cell IgE class switching assay.

CD19⁺IgD⁺CD27⁻CD23⁻IgE⁻ naïve B cells were purified by fluorescent-activated cell sorting (FACS) and cultured in vitro for 4 days at 30,000 cells/tissue culture plate well in the presence of either an IgG1 isotype control antibody (Biolegend, Cat# 400165, 0.5μg/ml) or with an anti-CD40 mAb (0.5μg/ml, Biolegend, Cat# 334350, 0.5μg/ml)+IL-4 (20ng/ml, Peprotech). At the end of the culture period the cells were stained with a viability dye and anti-CD19,-CD27, -CD23 and-IgE mAbs and analyzed by flow cytometry.

ELISA for human total IgE.

The levels of IgE in cells supernatants were determined by Sandwich ELISA. 96-well plates were coated with 2 μg/well of Purified Mouse Anti-Human IgE (BD Biosciences). For the standard curve Chimeric Human IgE anti NP (Bio-Rad) was used. The supernatants were incubated at 4°C overnight, followed by a detection antibody incubation step with Goat anti-Human IgE Fc Secondary Antibody, HRP (ThermoFisher Scientific). The reaction was developed using TMB as substrate (Thermo Scientific) and read at 450 nm.

Generation of STAT6 variant plasmids and expression of STAT6 variants.

A plasmid used for transfection studies contained full-length STAT6 in a pCMV6 entry vector with a C-terminal GFP tag (pCMV6-AC-STAT6-GFP) which was purchased from OriGene (RG210065). The directed STAT6 mutagenesis (c.G1114A; p.E372K) plasmid was generated with standard cloning techniques by using pCMV6-AC-STAT6-GFP as a template. Transient expression of STAT6 variants in HEK293T cells were accomplished using GeneJuice transfection reagent (Merck Millipore) according to manufacturer’s recommendations. Briefly, the day before transfection, HEK293T cells were seeded at a density of 5×10⁵ cells per well in 6 well plates in 3ml Dulbecco modified Eagle medium (DMEM) with 10% FBS (Gibco, Life Technologies; Rockville, MD, USA) and incubated for 24 hours at 37°C. The next day, HEK293T cells were transiently transfected with 2ug of either pCMV6-AC-STAT6-GFP wild type or STAT6 mutant plasmid using GeneJuice transfection reagent. Cells were harvested after 24h after transfection for phospho-flow cytometry analysis.

Luciferase reporter assays.

The p4xSTAT6-Luc2P luciferase reporter plasmid encoding 4xSTAT6 binding sites [TTCCCAAGAA] was obtained from Addgene (catalogue no. 35554) and used to assess STAT6 promoter activity. The pRL-SV40 Renilla luciferase control reporter vector was obtained from Promega (catalogue no. E2261) and was used in co-transfection studies with p4xSTAT6-Luc2P for internal normalization. HEK293 cells were seeded in 96-well plates overnight at a density of 30,000 cells per well and transfected with 20ng of plasmids encoding either GFP-tagged WT or p.E372K STAT6. The cells were further transfected with 100ng of p4xSTAT6-Luc2P and 4ng the pRL-SV40 vectors using Fugene HD transfection reagent (Promega, catalogue no. E2311), according to manufacturer’s recommendations. The next day, transfected cells were stimulated with 100ng/mL of IL-4 (PeproTech, catalogue no. 200–04), and transfected but otherwise untreated cells were used as a control group. Firefly and Renilla luciferase activities, as indicated by relative luminescence units (RLU), were determined using Dual-Glo luciferase assay kits (Promega, catalogue no. E2920), according to the manufacturer’s instructions. Luciferase activity was measured on BioTek Synergy HTX Multi-Mode Microplate Reader.

Immunofluorescence imaging.

Confocal microscopic analysis of STAT6 and p-STAT6 nuclear translocation was carried out following previously described methods⁴³. HEK293T cells were grown on the coverslips and transfected with indicated plasmid. After stimulation with 100ng/mL of IL-4 for 1 and 4h, the cells were fixed for 20 min in 4% paraformaldehyde and permeabilized with 0.3% Triton X-100. After blocked with 5% bovine serum albumin, the cells were stained with a mouse anti-p-STAT6 mAb [clone 4H253, Santa Cruz Biotechnology, catalogue number sc71793) and anti-STAT6 rabbit mAb (clone D3H4, Cell signaling, catalogue number 5397), incubated at 4 °C overnight. Alexa Fluor 555-goat anti-mouse IgG1 (Invitrogen, catalogue number A21127) and Alexa Fluor Plus-647 conjugated goat anti-rabbit (Invitrogen, catalogue number A32733) secondary antibodies (1:500 dilution) were used for visualizing. After, cells were washed with PBS and stained with 4,6-diamidino-2-phenylindole (DAPI) (Sigma; 1:10,000 dilution). Images were acquired with a Zeiss LSM880 confocal microscope and ZEN imaging software. Ten fields were selected randomly and total cells in the field were analyzed for the percentage of STAT6 nuclear localization using ImageJ software.

Whole exome sequencing.

The genetic diagnosis was made by whole-exome sequencing (WES) and the detected variant was confirmed by Sanger sequencing. Briefly, genomic DNA was extracted from peripheral blood samples and 1μg of DNA was used for exome capture using the IDT XGen exome target design or Agilent SureSelect Human All Exon. Generated libraries were sequenced using 75 bp paired-end sequencing on an Illumina NovaSeq-6000 and BGISeq-500 platform. Captured fragments were sequenced to achieve a minimum of 85% of the target bases covered at 20x or greater coverage. Analysis of WES data was performed using “Variant Explorer Pipeline” (VExP) to narrow down potential candidate variants⁴⁴. Raw data were processed, filtered and analyzed according VExP recommendations. Candidate genes were further evaluated by our research team (Table E1 in the Online Repository). Sanger sequencing was performed to confirm the mutation identified by the WES. Briefly, genomic DNA was amplified by a polymerase chain reaction and amplimers were sequenced using the Big Dye Terminator v1.1 Cycle Sequencing Kit (Applied Biosystems; Life Technologies, Darmstadt, Germany) on an Applied Biosystems 3130 Genetic Analyzer.

Statistical analysis.

The data expressed as mean and standard error of mean (SEM). Analysis of luciferase reporter activity, p-STAT6 phosphorylation and p-STAT6 nuclear translocation in HEK293 cells was carried out using 2-way ANOVA with post-test analysis. Differences in values were considered significant at a p value <0.05.

Results

Identification of a novel STAT6 mutation (STAT6^E372K) causing severe allergy dysregulation.

The index case is a ten-year-old boy born to consanguineous parents. He presented in the newborn period with severe atopic eczema, incessant itching, severe growth retardation, generalized lymphadenopathy, and pneumonia. He had generalized skin rash since the newborn period and itching was severely affecting his quality of life. Eczema and itching did not benefit from any local therapies and oral antihistamines. The eczema was associated with a robust IgE response to aero-and food allergens, including high levels of specific serum IgE to mite, grass, wheat, soy, cow milk, egg, tomato, kiwi, almond, and nuts. He experienced urticaria after ingestion of kiwi, and almond, and displayed dysphagia with vomiting following intake of wheat products, egg white, and tomato. A skin biopsy was performed at the age of seven years, which revealed chronic dermal inflammation characterized by perivascular and perifollicular eosinophils, follicular mucinosis, and fibrosis (Fig. 1A). During this period, the patient developed generalized lymphadenopathy involving the cervical, axillary, and inguinal lymph node, which persisted despite antibiotic therapy, prompting screening for malignancy. A bone marrow aspirate and biopsy were normal. An inguinal lymph node biopsy revealed findings consistent with reactive follicular hyperplasia, and paracortical enlargement. Meanwhile, he developed fever and painful swellings in his skin and axillary lymph nodes, and an ultrasound detected a local tissue abscess. Staphylococcus aureus was isolated and was treated with a systemic anti-staphylococcal antibiotic regimen.

Figure 1. Identification of a STAT6^E372K mutation in a child with severe allergy dysregulation.

(A) Scattered eosinophils in dermis, hematoxylin and eosin staining (H&E), 40X (a). Parakeratosis, spongiosis and acanthosis of epidermis (black arrows), H&E, 20X (b). Mucin accumulation dissociating epithelial cells of follicular epithelium, dilatation of follicle (black arrows), periodic acid schiff and Alcian Blue 2.5 staining, 10X (c). Gastrointestinal pathology shows esophagitis and scattered eosinophils with marked fibrosis in the subepithelial area of the esophagus (black arrows), H&E; 20X, 40X and trichrome staining; 40X, respectively (d, e, f). Inflammation with eosinophils in lamina propria, intraepithelial and intravascular areas of sigmoid colon, (H&E); 20X, 40X, 40X, respectively (g, h, i). (B) Pedigree of STAT6 GOF patient. Double lines indicate consanguinity; a filled black square depicts the patient. Males and females are distinguished by squares and circles, respectively. (C) Sanger sequencing analysis of the STAT6 c.1114G>A mutation in the index patient versus a healthy control. (D) Schematic diagram of STAT6 protein domains. The depicted domains are the α-Helix domain (α-H), DNA binding domain (DBD), linker domain (LD), SH2 Src homology 2 domain (SH2), and a transactivation domain (TAD). The heterozygous mutation localizes to the DBD. Multiple sequence alignment analysis of STAT6 proteins of different species demonstrates conservation of the STAT6E372 residue. (E, F) Structural modeling of the interaction of the STAT6^E372K and STAT3^N420K with DNA.

One year later, he was admitted to the hospital with a fever, and cough. Chest X-ray showed pneumonia and parapneumonic effusion. Methicillin resistance Staphylococcus aureus was detected in his sputum and blood cultures. Lung computed tomography revealed ground-glass densities and interlobular septal thickenings, predominantly in the peripheral and paramediastinal parts of both lungs. On physical examination, his weight (18 kg) and height (118 cm) were below the 3^rd percentile and he had mild developmental delay. There were xerotic excoriated lichenified eczematous lesions covering all the body, accompanied by hyperpigmentation. Coarse facial appearance with broad nasal bridge, increased alar width, and deep-set eyes, mild high palate arch, hyperextensibility, digital clubbing, and enlarged cervical, axillary and inguinal lymph nodes were also notable. He had normal dentition but with enamel hypoplasia (Fig. E1 in the Online Repository). The scoring of AD (SCORAD) and National Institute of Health-HIES scores were 96 and 44; respectively^{45, 46} (Table E2 in the Online Repository). Nevertheless, he had no evidence of AD-HIES-associated pneumotocele formation or vascular anomalies as evidenced by normal lung computed tomography, brain magnetic resonance imaging, and angiography. He was commenced on prophylactic antibiotic therapy and intravenous immunoglobulin replacement. Due to difficulty in swallowing, we performed upper and lower gastrointestinal endoscopy and showed findings of esophagitis and colitis with mild eosinophilic infiltrations (Fig. 1A). The endocrinologic evaluation revealed growth hormone (GH) deficiency (GH: 0.93 mIU/L (5.4–10.3), IGF-1: 72 ug/L (85.7–343)), and he was scheduled to start GH replacement therapy. His immunological workup was particularly notable for persistent eosinophilia (3,100/mm³). He had normal serum IgG, IgA, and IgM levels but high IgE (44,226 IU/ml), and protective protein antigen IgG responses, and IgM responses to isohemagglutinins. He had increased CD3⁺ and CD3⁺CD4⁺ T counts with a predominance of effector memory CD4⁺ T cells in the flow cytometric analysis (Table 1). T-cell proliferation with anti-CD3/anti-CD28 antibodies and phytohemagglutinin (PHA), and upregulation of CD25 and CD69 were normal in the patient when compared with the healthy controls (Fig. E2 in the Online Repository).

Table 1.

Immunological evaluation of the patient with STAT6^E372K gain-of-function mutation.

Parameters	Before ruxolitinib			On ruxolitinib	Normal range for age
Age (yrs)	8	9	10	10
Complete Blood Count
Lymphocyte (/mm 3 )	5,100	2,700	1,800	3,660	1,500–8,500
Neutrophil (/mm 3 )	9,600	4,900	4,800	2,600	1,500–6,500
Eosinophil (/mm 3 )	3,100	2,130	1,140	440	0–500
Hemoglobin (g/dl)	9.5	10.6	8.6	11.4	10–15.5
Thrombocyte (/mm 3 )	831,000	633,000	167,000	492,000	150,000–450,000
Immunoglobulins and antibody responses
IgG (mg/dl)	1610	1391	1803	1410	842–1953
IgA (mg/dl)	260	152	188	114	62–398
IgM (mg/dl)	116	178	173	109	54–392
IgE (IU/ml)	44,626	50,000	39,700	37,550	<50
Anti-Hbs (mIU/mL)	154	On IVIG	On IVIG	On IVIG	>10
Anti-Mumps (Index)	1.53	On IVIG	On IVIG	On IVIG	>1.1
Isohemagglutinin titer (Anti A-IgM)	1/256	On IVIG	On IVIG	On IVIG	≥1/8
Isohemagglutinin titer (Anti B-IgM)	1/64	On IVIG	On IVIG	On IVIG	≥1/8
Lymphocyte subsets (/mm 3 )
CD3+ T cells	4,539	2,268	1,585	3,100	1,214–4,130
CD3+4+ T cells	3,162	1,674	1,200	2,340	607–2,110
CD3+8+ T cells	1,224	540	325	640	380–2,084
CD19+ B cells	255	120	80	296	197–867
CD16+56+ NK cells	255	126	90	114	111–963
CD4+ T-cell subsets (%)
CD4 + CD45RA + CCR7 +	61	NA	51	72	40–78
CD4 + CD45 RA − C CR7 +	11	NA	12	9	17–53
CD4 + CD45 RA − C CR7 −	22	NA	30	14	2–14
CD4 + CD45RA + CCR7 −	5	NA	6	6	0–43
CD8+ T-cell subsets (%)
CD8 + CD45RA + CCR7 +	46	NA	53	77	22–67
CD8 + CD45RA − CCR7 +	3	NA	3	7	1–10
CD8 + CD45RA − CCR7 −	28	NA	25	9	6–41
CD8 + CD45RA + CCR7 −	22	NA	18	6	18–74
CD19 cell subsets (%)
CD19 + CD27 − IgD +	69	NA	50	71	55–90
CD19 + CD27 + IgD +	6	NA	7	7	7–31
CD19 + CD27 + IgD −	18	NA	25	12	2–52
CD21lowCD3low B cells	11	NA	3	12	1–13
IgRT	-	400 mg/kg/per 3 weeks	400 mg/kg/per 3 weeks	400 mg/kg/per 3 weeks
Antibacterial prophylaxis	-	TMP-SMX	TMP-SMX	TMP-SMX
Antifungal prophylaxis	-	-	Fluconazole	-
Antiviral prophylaxis	-	-	-	Aciclovir

Abbreviations: Anti-Hbs: Anti-hepatitis B surface antibody, IgRT: Immunoglobulin replecment therapy, NA: Not-available; PHA: Phytohemagglutinin; TMP-SMX: Trimethoprim-sulfamethoxazole. Abnormal values are shown in a bold manner. CD4⁺ naïve T cells (CD4⁺CD45RA⁺CCR7⁺), central memory CD4⁺ T cells (CD4⁺CD45RA⁻CCR7⁺), effector memory CD4⁺ T cells (CD4⁺CD45RA⁻CCR7⁻), terminally differentiated effector memory CD4⁺ T cells (CD4⁺CD45RA⁺CCR7⁻), CD8⁺ naive T cells (CD8⁺CD45RA⁺CCR7⁺), central memory CD8⁺ T cells (CD8⁺CD45RA⁻CCR7⁺), effector memory CD8⁺ T cells (CD8⁺CD45RA⁻CCR7⁻), terminally differentiated effector memory CD8⁺ T cells (CD8⁺CD45RA⁺CCR7⁻), naive mature B cells (CD19⁺CD27⁻IgD⁺), non-switched memory B cells (CD19⁺CD27⁺IgD⁺), switched memory B cells (CD19⁺CD27⁺IgD⁻), autoreactive B cells (CD21^lowCD3^low).

Identification of pathogenic STAT6 variant.

WES analysis revealed a novel heterozygous G>A substitution in exon 22 of STAT6 gene (c.1114G>A, p.E372K; based on STAT6 isoform 1, NM_003153.5) (Table E1 in the Online Repository). This mutation, which mapped to the DNA binding domain (DBD) of STAT6 and was confirmed by Sanger sequencing, has not been previously reported in the genomAD, ExAC, dbSNP, or 1000 Genomes Project databases, and it involved an amino acid residue that is conserved among species (Fig. 1B–D). Interestingly, this variant was previously identified in the COSMIC (Catalog of somatic mutations in cancer) database as a somatic mutation associated with lymphoid malignancies⁴⁷. It was not detected in DNA sequences of other available family members, including the mother and siblings, suggesting that it was sporadic. The deleteriousness of the variant was suggested by a high Combined Annotation Dependent Depletion (CADD) score of 27.9. The in-silico pathogenicity scores of SIFT, REVEL, Mutation tester, and PROVEAN were also damaging. The mutation was predicted to minimally affect protein stability as evaluated in silico by PremPS (Fig. E3A in the Online Repository)⁴⁸. Multiple sequence alignment analysis of STAT6 with other STAT proteins produced by ClusterW and ESpript (espript.ibcp.fr/ESPript/ESPript/) showed that the STAT6^E372K variant equivalent of STAT3^N420K, which has been reported as GOF in the literature (Fig. E3B in the Online Repository)^{49, 50}. Locating within the DNA-interacting loop, the STAT6^E372K variant is homologous to STAT3^N420K GOF mutants. Similar to STAT3^N420K, STAT6^E372K also features an elongated side chain comparing to wild type (Fig. 1E, F). Such extension increases the possibility of additional protein-DNA contact. Moreover, change of the side charge from neutral (STAT3^N420K) or negative (STAT6^E372K) to positive is likely to enable the interactions to the negatively charged DNA phosphate groups. Overall, these analyses suggested that the STAT6^E372K variant is deleterious and may act as GOF mutation.

The STAT6^E372K mutation is associated with dysregulated T_H2 cell responses.

Based on the role of STAT6 in T_H2 cell responses, we analyzed circulating CD4⁺ T-cell subtypes including naïve, memory, T_H1, T_H2, T_H17, Treg, and circulating follicular helper T (cT_FH) cells. The gating strategy is presented in Fig. E4. The patient with the STAT6^E372K variant exhibited an increased frequency of circulating memory CD4⁺ T cells (CD3⁺CD4⁺CD45RA⁻ CD45RO⁺) and a decreased frequency of circulating naïve CD4⁺ T cells (CD3⁺CD4⁺CD45RA⁺CD45RO⁻) (Fig. 2A). More detailed analysis of the CD4⁺ T cells compartment revealed that the patient with the STAT6^E372K variant had an increase of T_H2 CD4⁺ T cells as reflected by the expression of the T_H2 cell markers GATA3, CRTH2 and CCR4 (CD3⁺CD4⁺GATA3⁺, CD3⁺CD4⁺CRTH2⁺ and CD3⁺CD4⁺CCR4⁺). There was also a trend towards decreased expression of T_H1 (CD3⁺CD4⁺CXCR3⁺) and T_H17 (CD3⁺CD4⁺RORγT, CD3⁺CD4⁺CCR6⁺) cell markers (Fig. 2B, C). Treg cell analysis also revealed an increase of T_H2 skewed Treg cells³¹, indicative of their T_H2 cell-like reprogramming (CD3⁺CD4⁺CD127⁻Foxp3⁺GATA3⁺, CD3⁺CD4⁺CD127⁻Foxp3⁺CRTH2⁺ and CD3⁺CD4⁺CD127⁻Foxp3⁺CCR4⁺) and a decrease of T_H1 (CD3⁺CD4⁺CD127⁻Foxp3⁺CXCR3⁺) and T_H17 (CD3⁺CD4⁺CD127⁻ Foxp3⁺RORγt⁺, CD3⁺CD4⁺CD127⁻Foxp3⁺CCR6⁺) Treg cells (Fig. 2D, E). Further analysis also revealed a decrease of cT_FH (CD4⁺CXCR5⁺PD1⁺) cells, skewing of these cells toward a T_H2 phenotype (CD4⁺CXCR5⁺CD45RA⁻CXCR3⁻CCR6⁻) and away from other T_FH cell phenotypes including T_FH1-like (CD4⁺CXCR5⁺CD45RA⁻CXCR3⁺CCR6⁻) and T_FH17-like (CD4⁺CXCR5⁺CD45RA⁻CXCR3⁻CCR6⁺) (Fig. 2F). There was also an increase in circulating T follicular regulatory T cells (cT_FR) (CD4⁺CXCR5⁺PD1⁺CD127⁻Foxp3⁺) (Fig. 2F).

Figure 2. The STAT6^E372K mutation is associated with skewed T_H2 responses.

(A) Flow cytometric analysis and graphical representation of naïve (CD3⁺CD4⁺CD45RO⁻CD45RA⁺) and memory (CD3⁺CD4⁺CD45RO⁺CD45RA⁻) Teff cells in Healthy controls and the patient with STAT6^E372K mutation. (B) Flow cytometric analysis and graphical representation of T_H2 cells (CD3⁺CD4⁺GATA3⁺) in the respective groups. (C) Cells frequencies of CD4⁺CRTH2⁺, CD4⁺CCR4⁺, CD4⁺RORγ⁺, CD4⁺CCR6⁺, CD4⁺CXCR3⁺ T cells in the respective groups. (D) Flow cytometric analysis and graphical representation of Treg GATA3⁺ cells (CD3⁺CD4⁺CD127⁻Foxp3⁺GATA3⁺) in the respective groups. (E) Cells frequencies of Treg CRTH2⁺, Treg CCR4⁺, Treg RORγT⁺, Treg CCR6⁺, and Treg CXCR3⁺ T cells in the respective groups. (F) Cells frequencies of circulating T follicular helper (cT_FH), T follicular helper T_H2 (CD4⁺CXCR5⁺CD45RA⁻CXCR3⁻CCR6⁻), T follicular helper T_H17 (CD4⁺CXCR5⁺CD45RA⁻CXCR3⁻CCR6⁺), and T follicular regulatory cells in the respective groups. Each symbol represents one subject. Numbers in flow plots indicate percentages. Error bars indicate SEM.

The STAT6^E372K mutation is associated with enhanced B cell switching to IgE.

Analysis of patient peripheral blood B cells revealed a similar frequency of CD19⁺CD27⁺IgD⁻ class switched memory B cells but increased frequency of circulating CD19⁺CD20⁻CD38⁺ plasmablasts relative to control subjects (Fig. 3A,B, C). In particular, there was a high frequency of circulating CD19⁺CD27⁺IgE⁺ B cells in patient relative to controls (Fig. 3D). This increase reflected in part heightened expression on circulating patient B cells of the IL-4-inducible low affinity IgE receptor (CD23) (Fig. 3E)⁵¹, which could then bind circulating IgE regardless of the status of the B cell class switch isotype. Further analysis revealed that a subset of circulating B cells that do not express CD23 still exhibited IgE staining, raising the possibility of enhanced switching to IgE (Fig. 3F).

Figure 3. The STAT6^E372K mutation is associated with increased circulating CD23⁺ and IgE⁺ memory B cells.

(A) Gating strategy for B cells analysis. (B) Cell Frequencies of Class switched memory B cells (CD19⁺, CD27⁺ IgD⁻). (C) Flow cytometric analysis and graphical representation of Plasmablasts (CD19⁺ CD20⁻ CD38⁺) in healthy controls and in the patient with STAT6^E372K mutation. (D) Flow cytometric analysis and graphical representation of IgE⁺ circulating B cells in healthy controls and in the patient with the STAT6^E372K mutation. (E, F) Flow cytometric analysis and graphical representation of CD23⁺ (E) and CD23–IgE+ memory (IgD⁻ CD27⁺) B cells (F) in healthy controls and in the patient with STAT6^E372K mutation. Each symbol represents one subject. Numbers in flow plots indicate percentages. Error bars indicate SEM.

To determine the propensity of patient B cells to switch to IgE, we sorted naïve B cells (CD19⁺IgD⁺CD27⁻CD23⁻IgE⁻) to high purity (≥97%) following a gating strategy outlined in (Fig. 4A, B). The cells were cultured in vitro and stimulated with anti-CD40 mAb+IL-4 to induce switching to IgE⁵². Results showed that the patient B cells exhibited markedly increased IgE staining and total IgE in the supernatants at the end of the culture, indicative of enhanced switching to IgE (Fig. 4C, D)

Figure 4. The STAT6^E372K mutation is associated with enhanced B cell class switching to IgE.

(A) Gating strategy for cell-sorting of naïve B cells (CD19⁺IgD⁺CD27⁻CD23⁻IgE⁻). (B) Purity check after cell sorting. (C, D) Flow cytometric analysis, frequencies of IgE⁺ B cells (C), and total IgE in the culture supernatants (D) in cultures of the respective sorted B cell populations that were treated with an isotype control mAb or with anti-CD40 mAb+IL-4. Each symbol represents one subject. Numbers in flow plots indicate percentages. Error bars indicate SEM.

STAT6^E372K is a GOF mutation.

Analysis of primary CD4⁺ T cells revealed a modest increase in the expression of STAT6 in the patient relative to controls (Fig. 5A). To further characterize the functional significance of this mutation, primary T cells were stimulated with IL-4 for 5 to 30-minutes, and analysis by phosphoflow revealed that the STAT6^E372K mutation induces hyper-phosphorylation of STAT6 which was evidenced at baseline and increased dramatically post-IL-4 stimulation (Fig. 5B).

Figure 5. STAT6^E372K is a GOF mutation.

(A) Flow cytometric analysis and mean fluorescence intensity (MFI) of STAT6 in CD4⁺ T cells of Healthy controls and the patient with STAT6^E372K mutation. (B) Flow cytometric analysis and MFI of p-STAT6 (top) and frequencies of p-STAT6 cells (bottom) in CD4⁺ T cells of Healthy controls and the patient with STAT6^E372K mutation after stimulation with IL-4 (20ng/ml) for 5 to 30 minutes (Black line: Healthy control unstimulated, Red line: STAT6^E372K patient unstimulated, Blue: Healthy control stimulated with IL-4 and Red: STAT6^E372K patient stimulated with IL-4). (C) Flow cytometric analysis and MFI of p-STAT5 (top) and of p-STAT3 (bottom) in CD4⁺ T cells of Healthy controls and the patient with STAT6^E372K mutation after stimulation with IL-2 (20ng/ml) or IL-6 (20ng/ml), respectively for 5 to 30 minutes (D) Dephosphorylation kinetics of phospho-STAT6 in response to deprivation of IL-4 and in CD4⁺ T cells represented as absolute MFI (left) and normalized to maximum expression before deprivation (right). (E) MFI expression of STAT6 in HEK293 cell transfected with either STAT6^WT or STAT6^E372K proteins. (F) MFI expression of p-STAT6 in HEK293 cell transfected with either STAT6^WT or STAT6^E372K proteins after IL-4 (20ng/ml) stimulation for 5 to 30 minutes. (G) STAT6 response element-driven luciferase reporter activation by STAT6^WT or STAT6^E372K transfected into HEK293 cells at baseline and following IL-4 treatment for 1 and 4 hr. RLU: Relative Luciferase Unit. (H, I) Confocal microscopic analysis of total STAT6, p-STAT6 and DAPI in HEK293 cells transfected with either STAT6^WT or STAT6^E372K and then either sham-treated or stimulated with IL-4 for 1 and 4 hr, respectively. Numbers in flow plots indicate MFI. Error bars indicate SEM. Statistical tests: *P<0.05, **P<0.01, ****P<0.0001 by two-way ANOVA with Dunnett’s post hoc analysis (F; G, I).

Stimulation with IL-2 or IL-6 resulted in similar levels of STAT5 and STAT3 phosphorylation, respectively, in patient and control T cells (Fig. 5C). The STAT6^E372K mutation did not appear to delay the STAT6 dephosphorylation upon cytokine deprivation as evidenced by a rapid p-STAT6 decline, indicating that its main effect is to enhance the phosphorylation of STAT6 by the JAK kinases (Fig. 3D).

To further assess the impact of this mutation on STAT6 activity, we transfected HEK293 cells with either STAT6^E372K or control STAT6^WT expression plasmids and analyzed STAT6 protein expression and p-STAT6 formation in response to IL-4 treatment. Flow cytometric analysis revealed that the STAT6^E372K mutation did not impact STAT6 protein expression in HEK293 (Fig. 5E). To analyze the role of this mutation on the STAT6 activation, Phosphoflow analysis revealed that treatment of the respectively transfected cells with IL-4 resulted in increased phosphorylation of the STAT6^E372K mutant compared to STAT6^WT (Fig. 5F). This increase persisted when the phosphoflow staining was normalized for STAT6 expression. We further analyzed the capacity of the STAT6^E372K mutant to upregulate STAT6-dependent gene expression compared to STAT6^WT by co-transfecting HEK293 cells with a luciferase gene driven by STAT6 response elements together with the respective STAT6 expressing construct. Results showed that the STAT6^E372K mutant drove higher expression of the luciferase gene early (1 hr) compared to STAT6^WT, but that this increase was tempered at later time points (Fig. 5G). We also analyzed the nuclear translocation of p-STAT6 species in HEK293 cells that were transfected with either STAT6^WT or the STAT6^E372K mutant and then stimulated with IL-4. The p-STAT6^E372K mutant exhibited a modest increase in nuclear translocation at baseline relative to p-STAT6^WT. IL-4 treatment resulted in progressive nuclear translocation of p-STAT6 over the 4 hr duration of the study that was markedly increased for p-STAT6^E372K mutant relative to p-STAT6^WT (Fig. 5H) Overall, these results demonstrated that the STAT6^E372K is a GOF mutation that upregulates phosphorylation and the nuclear translocation of STAT6.

Ruxolitinib therapy suppresses the allergic dysregulation caused by STAT6^E372K.

Based on the immunological studies showing STAT6^E372K as a GOF mutation, and in view of the success of Jakinib therapy in patient with STAT GOF mutations, we initiated therapy of our patient with oral ruxolitinib (10 mg/m²/dose, twice daily). Ruxolitinib rapidly controlled the skin disease with decreased itching noticed within three days of therapy. The SCORAD levels decreased from 96 to 25, and ruxolitinib therapy induced almost normal skin hydration and appearance at one month post therapy (Fig. 6A). A normalization in blood eosinophil number and decrease in serum IgE levels were noticed (Table 1). Ruxolitinib therapy was not associated with drug-related adverse effects or infections. Currently, he continues to do well after six months of therapy with a most recent SCORAD of 8. Furthermore, there was resolution of his eosinophilic esophagitis-associated dysphagia and a catch-up in his weight from 18 kg (≤3^rd percentile) to 24 kg (3^rd percentile). Immunological studies carried out at one month post initiation of therapy revealed decreased frequencies of circulating memory T cells and an increase in those of naïve T cells (Fig. 6B). A reduction of the T_H2 skewing was also noted (Fig. 6C, D), associated with an increased T_H1 and T_H17 cell frequencies (Fig. 6E, F). The percentage of IL-4 producing circulating CD4⁺ T cells, which was increased in the patient pre-ruxolitinib, declined on therapy, while those of other T_H1 and T_H17 cell subsets increased (Fig. 6G, H). Ruxolitinib treatment did not impact T_FH cell frequencies or their T_H2-like skewing but normalized the frequencies of circulating Treg and T_FR cells (Fig. E5 in the Online Repository). Ruxolitinib treatment did not affect the upregulation of activation markers including CD69 and CD25 following in vitro T cell activation nor did it impact T cell proliferation in response to stimulation with T cell mitogens including anti-CD3+anti-CD28 antibodies and PHA (Fig. E1). Finally, treatment with ruxolitinib normalized p-STAT6 induction in patient T cells both at baseline and in response to cytokine treatment (IL-4) (Fig.6I). These results indicated that treatment with the JAK1 inhibitor ruxolitinib effectively suppressed the clinical and immunological attributes of the allergic dysregulation associated with STAT6^E372K mutation.

Figure 6. Ruxolitinib therapy suppresses the allergic dysregulation and inhibits the T_H2 skewing and STAT6 hyperactivation.

(A) Response of patient atopic dermatitis to ruxolitinib therapy. (a-e), Severe atopic dermatitis before ruxolitinib therapy covering the face (a), body (b, c), elbow and hand (d, e) of the patient. (f-i), control of atopic dermatitis after ruxolitinib treatment. (B) Flow cytometric analysis of naïve (CD3⁺CD4⁺CD45RO⁻CD45RA⁺) and memory (CD3⁺CD4⁺CD45RO⁺CD45RA⁻) Teff cells in healthy controls and the patient with STAT6^E372K mutation pre- and post-therapy with ruxolitinib. (C, D) Flow cytometric analysis of CD4⁺CCR4⁺(C) and CD4⁺CRTH2⁺(D) T cells in the respective groups. (E, F) Cells frequencies of CD4⁺CCR6⁺(E) and CD4⁺CXCR3⁺(F) T cells in the respective groups. (G, H) Flow cytometric analysis with representative plots of IL-4-, IL-17-, IFN-γ- and IL-10-producing circulating CD4⁺ T cells pre-and post-ruxolitinib therapy compared with healthy controls. (I) Flow cytometric analysis and MFI of p-STAT6 (left) and p-STAT6 fold change (right) in CD4⁺ T cells of healthy controls and the patient with STAT6^E372K mutation pre- and post-treated with ruxolitinib after stimulation with IL-4 (20ng/ml) for 5 to 30 minutes (Red line: STAT6^E372K patient pre-treatment unstimulated, Red: STAT6^E372K patient stimulated with IL-4, Orange line: STAT6^E372K patient post-treatment with ruxolitinib and orange: STAT6^E372K patient post-treated with ruxolitinib stimulated with IL-4). Each symbol represents one subject. Numbers in flow plots indicate percentages. Error bars indicate SEM.

Discussion

In this study, we describe the identification of a STAT6^E372K GOF mutation in a child with allergic dysregulation characterized by early-onset severe AD accompanied by features mimicking HIES syndrome. The mutation resulted in increased STAT6 activation as evidenced by increased phospho-STAT6 both at baseline and following IL-4-treatment and accelerated p-STAT6 translocation to the nucleus. STAT6-dependent reporter gene activation was increased early (1hr) but not later (4 hr) post IL-4 activation, which may reflect increased p-STAT6 sequestration in the nucleus. These alterations in STAT6 activity were associated with enhanced T_H2 cell responses and augmented B cell isotype switching to IgE. Furthermore, the dysregulated STAT6 activity was associated with T_H2 cell-like skewing of Treg cells, a phenotype conducive to immune dysregulation and exacerbated allergic inflammation³¹. The capacity of ruxolitinib to suppress the STAT6^E372K hyperactivity and ameliorate disease manifestation identify Jakinib inhibitors as appropriate targeted therapy in this novel disorder.

STAT6 is a central transcription factor in the nexus of type 2 immunity by virtue of its activation by IL-4 and IL-13 to promote a variety of allergic responses including T_H2 cell differentiation⁵³, B cell activation and IgE production^{22, 25}, alternate macrophage and epithelial goblet cell differentiation and innate lymphoid cells type 2 (ILC2), mast cells activation^54–57. As such, its dysregulation would be predicted to precipitate a broad allergic inflammatory response, as is reflected in the development in our patient of severe atopic dermatitis, eosinophilic esophagitis, food allergy and hyper IgE⁵⁸. Some of these phenotypes were reflected in related transgenic mouse models of enhanced STAT6 signaling. The first is an engineered STAT6 mutant with two alanine substitutions in the SH2 domain (STAT6VT), whose expression is restricted to lymphoid tissues. The STAT6VT mice develop marked T_H2-cell skewing, severe atopic dermatitis (AD)-like phenotype with blepharitis and allergic lung inflammation^{36, 37}. In a second model, an inactivating mutation (IL-4RαY709→F) was engineered in the immunotyrosine inhibitory motif (ITIM) in the cytoplasmic domain of the IL-4Rα chain, which dampens STAT6 activation by recruiting the phosphotyrosine phosphatase Shp1⁵⁹. Accordingly, the mutant receptor mediated enhanced IL-4- and IL-13-induced STAT6 phosphorylation and activation^{57, 59}. These mice exhibited augmented allergic responses including allergic airway inflammation and food allergy in part by a mechanism involving the Th2 cell like reprogramming of their Treg cells, leading to their failure to effectively control allergic responses^{31, 60, 61}. Of note, STAT6VT T effector cells also appear resistant to Treg cell-mediated suppression³⁰. Consistent with the T_H2 cell like phenotype of the Treg cells of the patient reported herein, one can postulate that the aberrant STAT6^E372K may similarly act in part to impair Treg cell control of the allergic response, giving rise among other things of dysregulated T_H2 and PD1⁺ T_H2-like cT_FH cell responses to drive the allergic inflammation in our patient.

Somatic GOF mutations in STAT6, including STAT6^E372K, have been previously described in follicular lymphomas. However, in that report, the E372K mutation was found associated with largely autonomous transcriptional activity independent of IL-4 or STAT6-Y641 phosphorylation³⁴. In contrast, our own studies showed that the GOF attributes of the STAT6^E372K mutation were overwhelmingly IL-4 dependent in terms of p-STAT6 formation, nuclear translocation and luciferase reporter activation as well as B cell switching to IgE. Furthermore, our patient responded to ruxolitinib, indicative of IL-4 receptor-driven STAT6^E372 activation in the patient. These differences may reflect the methods employed in the respective studies.

The identification of a germline GOF mutation in STAT6 expands the spectrum of GOF mutations in JAK-STAT pathways. In these disorders, the JAK inhibitors are leveraged to control of JAK-STAT pathway and consequently control different disease activities⁶². Recently, data are available on the successful usage of JAK inhibitors in IEI disorders like STAT1, STAT3, STAT5B, and JAK1 GOFs, interferonopathies, and SOCS1 haploinsufficiency^{15, 42, 63–66}. We previously showed that treatment with the JAK inhibitor in patients with STAT1 GOF mutations decreased hyperresponsiveness to type I and II interferons, normalized T_H1 and follicular T helper cell responses, improved T_H17 differentiation, cured mucocutaneous candidiasis, and maintained remission of associated autoimmunity^{41, 42}. The studies herein extend the usage of JAK inhibitors in primary atopic disorders by establishing disease control in STAT6 GOF disease. However, long-term data regarding the outcome of this targeted therapy are needed to understand the full efficacy of these drugs in controlling the disease symptoms.

Finally, the identification of STAT6 GOF mutations as a cause of severe allergic dysregulation expands our knowledge of the genetic architecture of allergic diseases^{1, 12, 67–70}, and highlights the pivotal role of the IL-4/IL-13/IL4R/STAT6 axis in disease pathogenesis. Establishing the molecular underpinning of these diseases enables the use of targeted therapies to reverse the atopic inflammation and restore immune homeostasis in these patients^{21, 71}.

Key messages.

1. A germline STAT6 gain-of-function mutation gave rise to severe allergic inflammation.

2. Therapy with a Janus kinase inhibitor controlled disease manifestations and suppressed the associated immune dysregulation.

Abbreviations.

AD

Atopic dermatitis

APC

Allophycocyanin

CRTH2

Chemoattractant receptor homologous molecule expressed on TH2-type cells

cT_FH

Circulating follicular helper T

GATA3

GATA binding protein 3

GFP

Green fluorescent protein

GOF

Gain-of-function

HIES

Hyper IgE syndrome

IEI

Inborn errors of immunity

IL-4R

IL-4 receptor

JAK

Janus kinase

Jakinibs

JAK inhibitors

p

Phosphorylated

SH2

Src homology 2

STAT

Signal transducer and activator of transcription

Treg

Regulatory T

WES

Whole-exome sequencing

WT

Wild type