Functional IRF3 deficiency in a patient with herpes simplex encephalitis

Abstract

Herpes simplex encephalitis (HSE) in children has previously been linked to defects in type I interferon (IFN) production downstream of Toll-like receptor 3. Here, we describe a novel genetic etiology of HSE by identifying a heterozygous loss-of-function mutation in the IFN regulatory factor 3 (IRF3) gene, leading to autosomal dominant (AD) IRF3 deficiency by haploinsufficiency, in an adolescent female patient with HSE. IRF3 is activated by most pattern recognition receptors recognizing viral infections and plays an essential role in induction of type I IFN. The identified IRF3 R285Q amino acid substitution results in impaired IFN responses to HSV-1 infection and particularly impairs signaling through the TLR3-TRIF pathway. In addition, the R285Q mutant of IRF3 fails to become phosphorylated at S386 and undergo dimerization, and thus has impaired ability to activate transcription. Finally, transduction with WT IRF3 rescues the ability of patient fibroblasts to express IFN in response to HSV-1 infection. The identification of IRF3 deficiency in HSE provides the first description of a defect in an IFN-regulating transcription factor conferring increased susceptibility to a viral infection in the CNS in humans.

Figure 1.

Identification of a heterozygous mutation in IRF3 by WES. (a) Summary of information on the IRF3 mutation in P1. (b) Sanger sequencing of the IRF3 gene in P1 and healthy control. (c) Schematic diagram of the IRF3 protein consisting of an N-terminal DNA-binding domain (DBD), a central regulatory domain (RD), and C-terminal serine-rich region (SRR). The identified R285Q mutation is localized in the RD. (d) Family pedigree with allele segregation. Family members heterozygous for the mutation are indicated by a bold vertical line (e) Whole-cell lysates from PBMCs from P1 and a healthy control were subjected to Western blotting and probed with anti-IRF3 and anti-GAPDH.

Figure 2.

Impaired IFN induction through nucleic acid–sensing pathways in patient cells. (a–o and q–t) PBMCs from the patient and controls were stimulated with 4 µg/ml poly(dA:dT) (a–e), 4 µg/ml HSV-1–derived dsDNA (f–j), 50 µg/ml extracellular poly(I:C) (k–o), 1 µg/ml R848, 5 µM ODN2216, or 4 µg/ml transfected poly(I:C) (q–t). Total RNA was harvested 6 h later and subjected to RT-qPCR for measurement of IFN-β (a, f, k, and q), IFNα2 (b, g, l, and r), IFNλ1 (c, h, m, and s), CXCL10, (d, i, and n), or TNF (e, j, o, and t). Cytokine mRNA levels were normalized and compared with an age- and gender-matched control or the pooled results of a total of 12 controls. (p) Fibroblasts from the patient and controls were stimulated with extracellular poly(I:C). Total RNA was harvested 6 h later and subjected to RT-qPCR for measurement of IFN-β. Data are shown as box plots with median, first, and third quartiles. Error bars represent minimum and maximum values. The pooled controls are illustrated as the 5–95% population, with outliers shown as independent dots. For all data, similar results were obtained in at least two independent experiments. Nonparametric Mann-Whitney ranked sum test was used for statistical analysis. *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001; ****, P ≤ 0.0001. C, control. P, patient. t-, transfected.

Figure 3.

Abolished IFN responses to HSV-1 in patient cells. (a–e and g–j) PBMCs from the patient and controls were infected with HSV-1 (MOI 9), HSV-2 (MOI 9), HHV8 (30 genomes per cell), or IAV (MOI 3) as indicated for 6 h, and RNA was isolated for measurement of IFN-β (a and g), IFNα2 (b and h), IFNλ1 (c and i), CXCL10 (d), and TNF expression (e and j). Cytokine mRNA levels were normalized and compared with an age- and gender-matched control or the pooled results of a total of 12 healthy controls. (f) Fibroblasts were infected with HSV-1 (MOI 9). Total RNA was harvested 6 h later and subjected to RT-qPCR for measurement of IFN-β. Data are shown as box plots with median, first, and third quartiles. Error bars represent minimum and maximum values. The pooled controls are illustrated as the 5–95% population, with outliers shown as independent dots. For all data, similar results were obtained in at least two independent experiments. Non-parametric Mann-Whitney ranked sum test was used for statistical analysis. *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001. (g–j) C, control; P, patient.

Figure 4.

The R285Q IRF3 mutant is not phosphorylated, dimerized, or transcriptionally activated upon stimulation. (a, b, and f) HEK293T-derived, IRF3-deficient cells were transiently transfected with IFN-β promoter luciferase reporter, β-actin promoter, Renilla luciferase reporter, and plasmids encoding WT IRF3, R285Q IRF3, MAVS, STING, or TRIF as indicated. (a and f) Cells were infected with SeV (20 HAU/well) and luciferase activities were measured 16 h later (a, b, and f). Firefly luciferase activity was normalized to Renilla luciferase activity and presented as either means from triplicate cultures ± SD or percentage stimulation in cells expressing R285Q IRF3 relative to cells expressing WT IRF3 ± SD. Data from 2–3 independent experiments are shown. (c) IRF3-deficient THP1-derived monocytes transduced with WT and R285Q IRF3 were infected for 6 h with HSV-1 (MOI 3) or SeV (5 HAU/well). (d and e) L929 cells were transfected with empty vector, IRF3 WT or IRF3 R285Q as indicated. The cells were stimulated by infection with Newcastle disease virus (MOI 100) or cotransfection with human TBK1 or TRIF. Cells were lysed and subjected to native-PAGE followed by immunoblotting with antibodies against IRF3 and IRF3 S386-P. (g and h) Fibroblasts from P1 and a control were transduced with lentiviral vectors encoding eGFP, WT IRF3, or R285Q IRF3. The cells were infected with HSV-1 (g; MOI 9) or stimulated with extracellular poly(I:C) (h; 50 µg/ml). Total RNA was harvested 6 h later and subjected to RT-qPCR for measurement of IFN-β. Data are shown as means of four measurements ± SD. Data are from at least two independent experiments. Nonparametric Mann-Whitney ranked sum test was used for statistical analysis. *, P ≤ 0.05; **, P ≤ 0.01; ***, P ≤ 0.001. RQ, R285Q.