Genetic variations in A20 DUB domain provide a genetic link to citrullination and neutrophil extracellular traps in systemic lupus erythematosus

Abstract

Objectives: Genetic variations in TNFAIP3 (A20) de-ubiquitinase (DUB) domain increase the risk of systemic lupus erythematosus (SLE) and rheumatoid arthritis. A20 is a negative regulator of NF-κB but the role of its DUB domain and related genetic variants remain unclear. We aimed to study the functional effects of A20 DUB-domain alterations in immune cells and understand its link to SLE pathogenesis.

Methods: CRISPR/Cas9 was used to generate human U937 monocytes with A20 DUB-inactivating C103A knock-in (KI) mutation. Whole genome RNA-sequencing was used to identify differentially expressed genes between WT and C103A KI cells. Functional studies were performed in A20 C103A U937 cells and in immune cells from A20 C103A mice and genotyped healthy individuals with A20 DUB polymorphism rs2230926. Neutrophil extracellular trap (NET) formation was addressed ex vivo in neutrophils from A20 C103A mice and SLE-patients with rs2230926.

Results: Genetic disruption of A20 DUB domain in human and murine myeloid cells did not give rise to enhanced NF-κB signalling. Instead, cells with C103A mutation or rs2230926 polymorphism presented an upregulated expression of PADI4, an enzyme regulating protein citrullination and NET formation, two key mechanisms in autoimmune pathology. A20 C103A cells exhibited enhanced protein citrullination and extracellular trap formation, which could be suppressed by selective PAD4 inhibition. Moreover, SLE-patients with rs2230926 showed increased NETs and increased frequency of autoantibodies to citrullinated epitopes.

Conclusions: We propose that genetic alterations disrupting the A20 DUB domain mediate increased susceptibility to SLE through the upregulation of PADI4 with resultant protein citrullination and extracellular trap formation.

Keywords: NET; PAD4; PADI4; peptidyl arginine deiminase; rs2230926.

Figure 1

Genetic disruption of A20 de-ubiquitinase (DUB) domain does not lead to enhanced NF-ĸB activation. (A) Generation of A20 C103A knock-in (KI) mutation in U-937 cell line using CRISPR/Cas9. Three different homozygote C103A KI clones were generated and Sanger sequencing of the homozygote clones is shown to the right. (B) RNA sequencing data from lipopolysaccharide (LPS)-stimulated human U-937 cells (A20 WT, knock-out (KO) and C103A KI) were analysed using gene set enrichment analysis (GSEA). NF-ĸB target genes were significantly enriched in A20 KO cells (left), but not in A20 C103A KI cells (right), when compared with their wild-type counterparts. (C) Peripheral blood mononuclear cells (PBMCs) from individuals with A20 DUB polymorphism rs2230926 (n=10, CA or CC) or without (n=7, AA), were treated with LPS for 6 hours and intracellular staining of tumour necrosis factor-alpha and IL-6 was measured by flow cytometry in monocytes (CD14 Mo and CD16 Mo) and myeloid dentritic cells (mDC). No significant differences were found between genotypes using a two-tailed t-test. CA represents individuals that are heterozygous for rs2230926 risk polymorphism, CC are homozygous and AA represents individuals without rs2230926 risk allele. FDR, false discovery rate; NES, normalised enrichment score; OTU, ovarian tumour domain; ZF, zinc finger.

Figure 2

A20 de-ubiquitinase (DUB) domain disruption leads to PADI4 over expression. (A) Confirmation of RNAseq data using qPCR analysis of the expression of PADI4 mRNA in U937 wild-type (WT) (n=4), A20 knock-out (KO) (n=4) and A20 C103A knock-in (KI) cells (n=3). Individual squares or circles represent individual CRISPR/Cas9-generated cell clones. (B) PAD4 protein expression measured by western blot in DMSO-treated individual U937 WT and C103A KI clones. (C) qPCR analysis of the expression of murine PADI4 mRNA in BMDMs from WT and A20 C103A KI mice (n=3 per group). (D) qPCR analysis of the expression of PADI4 mRNA in human PBMCs (n=13 AA, n=13 CA). (E) qPCR analysis of the expression of PADI4 in PBMC-derived human monocytes (left graph, n=9 AA, n=13 CA). The right graph represent Western blot quantification of PAD4 protein relative to glyceraldehyde-3-phosphate dehydrogenase (GAPDH) in PBMC-derived human monocytes (n=7 AA, n=6 CA). CA represents individuals that are heterozygous for rs2230926 risk polymorphism and AA represents individuals without rs2230926 risk allele. (A,C,D,E) PADI4 expression levels are normalised to house-keeping genes GUSB and ACTB and are represented as relative mRNA expression to average PADI4 expression in controls (WT or AA). Statistical differences between groups were calculated using two-tailed t-test with Welch correction (D) or non-parametric one-tailed Mann-Whitney test (A,C,E). Bars represent mean+SD. *p<0.05, **p<0.01. BMDM, bone marrow derived macrophage.

Figure 3

Disruption of A20 de-ubiquitinase (DUB) domain is associated with PAD4-dependent increase in protein citrullination and release of nuclear antigens. (A–B) Total intracellular protein citrullination was measured in wild-type (WT) and C103A knock-in (KI) U937 cells using flow cytometry after stimulation with ionomycin. The effect of PAD4 inhibitor GSK484 on protein citrullination was assessed in U937 A20 C103A KI cells (B). Data points represent experimental duplicates in three independent experiments. (C) Western blot performed on supernatants from 1 million WT or A20 C103A KI U937 cells stimulated with ionomycin for 6 hours, with or without GSK484. GAPDH expression from the cell lysate is represented as an input control. representative blot from one out of four experiments. (D, E) Quantification of H3cit Western blot signal from four independent experiments is represented as mean H3cit density +SD. (F) the presence of cell free double stranded DNA (dsDNA) in the supernatants of U937 WT and KI cells were measured after 6 hours of ionomycin treatment. Results from four independent experiments are represented as mean % increase compared with unstimulated cells +SD. (G) The presence of dsDNA in the supernatants of unstimulated PBMCs derived from healthy individuals with the rs2230926 risk variant (CA) or without (AA). Statistical differences between groups were performed using paired two-tailed t-test (D–F), two-tailed t-test (G) or paired two-tailed non-parametric Wilcoxon test (A,B). Unless different concentrations are stated, 4 µM ionomycin was used in the presence of 2 mM CaCl2. GSK484 was used at 10 µM. *p<0.05, **p<0.01, ***p<0.001.

Figure 4

Disruption of A20 de-ubiquitinase (DUB) domain is associated with PAD4-dependent increase in NET formation. (A–C) Immunostaining and confocal microscopy of citrullinated H3 (H3cit, red) and DNA (SytoxGreen in green or Hoechst in blue) in murine and human neutrophils. Scale bar=50 µm. (A) Comparison of H3cit expression in Ionomycin-stimulated neutrophils from wild-type (WT) and C103A Ki mice. Results are expressed as mean percentage H3cit-postitive cells compared with total cell number +SD, and are pooled values from three independent experiments including totally 13 WT and 11 C103A KI mice. Each data point represents the mean of one experiment. (B) Neutrophils from C103A KI mice were pretreated with PAD4 inhibitor GSK484 or vehicle (DMSO) before NET-induction by ionomycin. Representative image. (C–E) Neutrophils from human systemic lupus erythematosus (SLE)-patients that were homozygous (CC, n=1), heterozygous (CA, n=5), or lacked (AA, n=6) rs2230926, were induced to form NETs in vitro. (C,D) Human SLE-derived neutrophils were pretreated with GSK484 or vehicle (DMSO) and NET formation was induced with 4 µM ionomycin. Representative images show reduced H3cit expression in GSK484-treated cells (C). H3cit was released in the supernatant on ionomycin stimulation of SLE-neutrophils and could be blocked by PAD4 inhibition (D). (E) Neutrophils from SLE-patients were treated with 1 µM ionomycin and the number of H3cit-positive cells was counted and represented as percentage compared with total cell number. Samples analysed on the same day are connected with a line. Statistical differences between groups were performed using a two-tailed t-test (A) or paired two-tailed t-test (based on age-matched patients from the same experiment) (H). Unless different concentrations are stated, 4 µM ionomycin was used in the presence of 2 mM CaCl2. GSK484 was used at 10 µM. *p<0.05, **p<0.01.