Identification of a Novel Mutation in TNFAIP3 in a Family With Poly-Autoimmunity

Abstract

Haploinsufficiency of A20 (HA20) is an inflammatory disease caused by mutations in the TNFAIP3 gene classically presenting with Behcet's-like disease. A20 acts as an inhibitor of inflammation through its effect on NF-kB pathway. Here we describe four consanguineous patients (three sisters and their mother) with a predominantly autoimmune phenotype, including thyroiditis, type I diabetes, hemolytic anemia and chronic polyarthritis. All patients had recurrent oral ulcers, with only 1 patient presenting also recurrent fever episodes, as a classical autoinflammatory feature. Next generation sequencing identified a novel heterozygous frameshift mutation (p.His577Alafs*95) that causes a premature stop codon in the zinc finger domain of A20, leading to a putative haploinsufficiency of the protein. Functional analyses confirmed the pathogenicity of the mutation. The variant was associated with decreased levels of A20 in blood cells. Accordingly, ex-vivo lipopolysaccharide (LPS)-stimulated patients' peripheral blood mononuclear cells (PBMCs) showed higher levels of p65 NF-kB phosphorylation, as well as increased production of the proinflammatory cytokines IL-1β, IL-6 and TNF-α. Moreover, in agreement with recent observations, demonstrating a role for A20 in inhibiting STAT1 and IFNγ pathways, markedly higher circulating levels of the two IFNγ-inducible chemokines CXCL9 and CXCL10 were detected in all patients. Supporting the findings of a hyperactivation of IFNγ signaling pathway in HA20 patients, patients' monocytes showed higher levels of STAT1 without stimulation, as well as higher phosphorylated (active) STAT1 levels following IFNγ stimulation. In conclusion, our study show that in the clinical spectrum of HA20 autoimmune features may predominate over autoinflammatory features and demonstrate, from a molecular point of view, the involvement of A20 in modulating not only the NF-kB, but also the IFNγ pathway.

Keywords: HA20; IFNγ; NF-kB; TNFAIP3; poly-autoimmunity.

Figure 1

Family pedigree and genetic analyses. (A) Pedigree of the family with the heterozygous mutation in TNFAIP3 gene. (B) Schematic diagram of A20 protein and the location of the new identified mutation. OTU, ovarian tumor domain; ZnF 1-7, zinc finger domains. (C) Whole exome sequencing identified a novel heterozygous mutation p.His577Alafs*95 in the TNFAIP3 gene. (D) Confirmation of the presence of the mutation in Pt2 and Pt3 by Sanger Sequencing.

Figure 2

Reduced A20 function in vivo. (A) Western Blot and densitometric analysis of A20 protein levels in PBMCs isolated from patients (Pt1-4) and 2 healthy donors (HD). GAPDH was used as loading control. Densitometric quantification of A20 was reported (right graph). (B) PBMCs isolated from Pt1, Pt4 and one healthy subject (HD) were starved for 2 h in media with 0.5% of FBS and lysed or stimulated for 30 minutes with 0.01 μg/ml LPS and then lysed. Phosphorylated (S536) p65 NF-kB (P-p65) and total p65 NF-kB protein levels were assessed by Western blot analyses. GAPDH was used as loading control. The phosphorylated-p65 NF-kB/total p65 NF-kB densitometry comparative ratio was also reported (right graph). Similar results were obtained in two independent experiments.

Figure 3

PBMCs from HA20 patients express markedly higher levels of pro-inflammatory cytokines. (A) The mRNA expression levels of IL1B, IL6 and TNF were evaluated by qPCR analysis in PBMCs from patients (Pt1-4), a relative (father) and 2 healthy donors (HDs) unstimulated (US) or stimulated for the indicated hours h with 0.01 μg/ml of LPS. Results were obtained after normalization with the housekeeping gene HPRT1 and were expressed as arbitrary units (AU). (B) PBMCs isolated from patients (Pt1-4), a relative (father) and 4 healthy donors (HDs) were unstimulated (US) or stimulated for 24 hours with 0.01, 0.1 and 1 μg/ml of LPS and IL-1β, IL-6 and TNF-α protein levels were measured in the conditioned media by ELISA.

Figure 4

Interferon gamma pathway is upregulated in HA20 patients. (A) CXCL9 and CXCL10 levels were measured in plasma samples collected from patients (Pt1-4) during different hospitalizations, from the father and from healthy donors (HD; n=15) by ELISA. Red bars indicate sample median. Statistical analyses were performed with Mann-Whitney test comparing each patient with HDs. **p<0,01; ***p<0,001. (B) Unstimulated PBMCs from patients (Pt1-4) and two healthy donors (HDs) were stained for total STAT1 levels. Results are reported as STAT1 mean fluorescence intensity (MFI) in monocytes (CD14+ cells). (C) PBMCs from patients (Pt1-4) and two healthy donors (HDs) were stimulated for 10 minutes with 10 ng/ml of IFNγ and phosphorylated STAT1 (pSTAT1) levels were detected by flow cytometry. Results were reported as % of pSTAT1 positive monocytes (CD14+ cells). (D) PBMCs from patients (Pt1-4) and one healthy donor (HD) were left unstimulated (US, white columns) or stimulated for 2 hours with 10 ng/ml of IFNγ (IFNγ, black columns) and CXCL9 and CXCL10 mRNA levels were analyzed by qPCR. Results were obtained after normalization with the housekeeping gene HPRT1 and were expressed as arbitrary units (AU). Similar results were obtained in two independent experiments.