Targeted deep sequencing identifies rare loss-of-function variants in IFNGR1 for risk of atopic dermatitis complicated by eczema herpeticum

Li Gao, Lianghua Bin, Nicholas M Rafaels, Lili Huang, Joseph Potee, Ingo Ruczinski, Terri H Beaty, Amy S Paller, Lynda C Schneider, Rich Gallo, Jon M Hanifin, Lisa A Beck, Raif S Geha, Rasika A Mathias, Kathleen C Barnes, Donald Y M Leung, Li Gao, Lianghua Bin, Nicholas M Rafaels, Lili Huang, Joseph Potee, Ingo Ruczinski, Terri H Beaty, Amy S Paller, Lynda C Schneider, Rich Gallo, Jon M Hanifin, Lisa A Beck, Raif S Geha, Rasika A Mathias, Kathleen C Barnes, Donald Y M Leung

Abstract

Background: A subset of atopic dermatitis is associated with increased susceptibility to eczema herpeticum (ADEH+). We previously reported that common single nucleotide polymorphisms (SNPs) in the IFN-γ (IFNG) and IFN-γ receptor 1 (IFNGR1) genes were associated with the ADEH+ phenotype.

Objective: We sought to interrogate the role of rare variants in interferon pathway genes for the risk of ADEH+.

Methods: We performed targeted sequencing of interferon pathway genes (IFNG, IFNGR1, IFNAR1, and IL12RB1) in 228 European American patients with AD selected according to their eczema herpeticum status, and severity was measured by using the Eczema Area and Severity Index. Replication genotyping was performed in independent samples of 219 European American and 333 African American subjects. Functional investigation of loss-of-function variants was conducted by using site-directed mutagenesis.

Results: We identified 494 single nucleotide variants encompassing 105 kb of sequence, including 145 common, 349 (70.6%) rare (minor allele frequency <5%), and 86 (17.4%) novel variants, of which 2.8% were coding synonymous, 93.3% were noncoding (64.6% intronic), and 3.8% were missense. We identified 6 rare IFNGR1 missense variants, including 3 damaging variants (Val14Met [V14M], Val61Ile, and Tyr397Cys [Y397C]) conferring a higher risk for ADEH+ (P = .031). Variants V14M and Y397C were confirmed to be deleterious, leading to partial IFNGR1 deficiency. Seven common IFNGR1 SNPs, along with common protective haplotypes (2-7 SNPs), conferred a reduced risk of ADEH+ (P = .015-.002 and P = .0015-.0004, respectively), and both SNP and haplotype associations were replicated in an independent African American sample (P = .004-.0001 and P = .001-.0001, respectively).

Conclusion: Our results provide evidence that both genetic variants in the gene encoding IFNGR1 are implicated in susceptibility to the ADEH+ phenotype.

Keywords: IFNGR1; atopic dermatitis; eczema herpeticum; genetic variants.

Copyright © 2015 American Academy of Allergy, Asthma & Immunology. Published by Elsevier Inc. All rights reserved.

Figures

Fig. 1
Fig. 1
(A) Combined site-frequency spectrum of ADEH+ and ADEH− patients for the 4 interferon-pathway genes. (B) Proportions of all variants identified in the 4 interferon pathway genes according to major functional categories. (C) Proportions of variants identified in ADEH+ and (D) ADEH− patients.
Fig. 2
Fig. 2
Schematic presentation of 6 rare missense variants in IFNGR1. On the top, the variations that identified by targeted sequencing are illustrated with allele frequencies (in ADEH+/ADEH− patients), 3 known functional variations are highlighted in bold and 2 novel variants are in red. On the bottom, the various domain and 7 exons in the 5′ to 3′ direction are indicated. SP=signal peptide, TM=transmembrane domain.
Fig. 3
Fig. 3
Single marker and haplotype association tests for European Americans and ENCODE regulation tracks on IFNGR1 region (chr6: 137,516,621-137,542,567). SNVs identified by targeted sequencing significantly associated with ADEH+ phenotype (with P values great than 0.01) were illustrated (top panel). The y axis indicates the value of −log10 (P), and the x axis indicates the relative position for each SNV locus in IFNGR1, 3′ to 5′ direction. The vertical lines (green for intron, red for coding-nonsynonymous, purple for coding-synonymous, grey for near-gene-5 and blue for utr-5) represent the position of each SNV on the x axis, and height of the line on the y axis indicates the value of −log10 (P). Similarly, the most significant haplotype results from two to seven-marker windows across the IFNGR1gene were also illustrated by horizontal lines in black (except red color was used for the 7-SNP window) and a P value cutoff of 0.05 was depicted by the grey horizontal line. The bottom panel illustrates the IFNGR1 structure, position of seven SNVs within IFNGR1 gene region (User Track) and regulatory regions with ENCODE regulation tracks including UCSC Gene (IFNGR1), DNase Clusters, Transcription Factor ChIP-seq, Layered H3K27Ac and Transcription Levels on 3 cell lines (NHLF (in pink color), NHEK (in purple color), and K562 (in blue color).
Fig. 4
Fig. 4
Reduced GAS activation by IFNgR1V14M and IFNgR1Y397C variants. Plasmids of empty-vector, IFNgR1WT, IFNgR1V14M, IFNgR1Y397C, and IFNgR1V14MY397C were cotransfected with GAS luciferase reporter plasmids into 293FT cells for overnight. A portion of cells was stimulated with IFN-γ (10ng/ml) for an additional 6 hours. The upper panel shows the luciferase intensity of GAS reporter; the lower panel shows the receptor levels corresponding to the samples shown in the upper panel. This is the representative result of three independent experiments.
Fig. 5
Fig. 5
IFN-γ induced phosphorylation of STAT1 is reduced in IFNgR1(−/−). EBV cells reconstituted with IFNgR1V14M and IFNgR1Y397C variants. Plasmids of IFNgR1WT, IFNgR1V14M, IFNgR1Y397C, and IFNgR1V14MY397C were transfected into IFNgR1 deficient cells. The cells were then stimulated with recombinant human IFN-γ at concentration of 0, 10 ng/ml and 50ng/ml for 30 min. 100μg of protein per lane were loaded for the detection of pSTAT1, STAT1, IFNgR1 wildtype and mutants.

Source: PubMed

3
Abonneren