ADAR1-dependent editing regulates human β cell transcriptome diversity during inflammation

Abstract

Introduction: Enterovirus infection has long been suspected as a possible trigger for type 1 diabetes. Upon infection, viral double-stranded RNA (dsRNA) is recognized by membrane and cytosolic sensors that orchestrate type I interferon signaling and the recruitment of innate immune cells to the pancreatic islets. In this context, adenosine deaminase acting on RNA 1 (ADAR1) editing plays an important role in dampening the immune response by inducing adenosine mispairing, destabilizing the RNA duplexes and thus preventing excessive immune activation.

Methods: Using high-throughput RNA sequencing data from human islets and EndoC-βH1 cells exposed to IFNα or IFNγ/IL1β, we evaluated the role of ADAR1 in human pancreatic β cells and determined the impact of the type 1 diabetes pathophysiological environment on ADAR1-dependent RNA editing.

Results: We show that both IFNα and IFNγ/IL1β stimulation promote ADAR1 expression and increase the A-to-I RNA editing of Alu-Containing mRNAs in EndoC-βH1 cells as well as in primary human islets.

Discussion: We demonstrate that ADAR1 overexpression inhibits type I interferon response signaling, while ADAR1 silencing potentiates IFNα effects. In addition, ADAR1 overexpression triggers the generation of alternatively spliced mRNAs, highlighting a novel role for ADAR1 as a regulator of the β cell transcriptome under inflammatory conditions.

Keywords: RNA editing; T1D (type 1 diabetes); beta cell (β cell); inflammation; transcriptome.

Figure 1

The proinflammatory cytokines IFNα and IFNɣ+IL-1β induce a partially shared gene signature in EndoC-βH1 and human islets Venn diagrams of up-regulated (log₂FC > 0.58 and adj. P value < 0.05) genes in EndoC-βH1 and human islets after exposure to IFNɣ+IL-1β (A), top) and IFNα (A), bottom). Common genes have been tested for enrichment – using REACTOME as the reference – and significantly enriched pathways are represented as a dot plot: the x-axis represents the gene ratio and the y-axis the enriched pathways. (B) Heatmap representing the log₂ fold change of the 128 up-regulated genes in all 4 datasets (|log₂FC| > 0.58 and adj. P value < 0.05). (C) ADAR1 p150 and STAT1 gene expression in EndoC-βH1 were assessed by qPCR after cytokine treatment. n=3 independent experiments (D) ADAR1 p150 ADAR1 (detected using Anti-ADAR1 #ab168809, Abcam), STAT1 and STAT2 protein expression determined by western blot. β-actin expression was used as loading control.

Figure 2

Cytokine treatment leads to A- to I- mutation in EndoC-βH1 and primary human islets. (A) Global A-to-I RNA editing index across Alu elements (short interspersed nuclear elements) in RNA-seq data demonstrates a higher A-to-I editing signal in IFNα or IFNɣ/IL1β stimulated samples after 8 hours and 48 hours, respectively. Student’s paired two-tailed t-test; *P<0.05, **P<0.005, ****P<0.0001. (B) Noise levels (non-A-to-G mismatches) are notably lower than seen in the global editing index’s biological signal (A-to-G mismatch).

Figure 3

ADAR1 overexpression inhibits the type I IFN response. (A) ADAR1 p150 expression in EndoC-βH1 and EndoC- βH1 (ADAR1) cells determined by qPCR (upper panel) and western blot analysis using Anti-ADAR1 #ab168809, Abcam (lower panel). (B) Volcano plot on differential expressing genes after ADAR1 overexpression. Dashed lines show log₂FC ≤ -2 or log₂FC ≥2 and adj. P value < 0.05. Plot was generated using Enhanced Volcano (C) Pathway analysis on downregulated (left, log₂FC ≤ -2, adj. P value < 0.05) and upregulated (right, log₂FC ≥2, adj. P value < 0.05) genes. Plots were generated using Revigo. (D) Gene expression of ADAR1 p150, MDA5, IFNβ, IFITH1, CXCL10 and STAT1 after ADAR1 overexpression in the presence or absence of polyI:C. N=3 independent experiments. Data are expressed as means of independent experiments ± SEM. Differences between groups were evaluated using one-way ANOVA or linear mixed model in case of missing values, followed by Bonferroni post-hoc test. **P<0.01 and ****P<0.0001.

Figure 4

ADAR1 silencing exacerbates the effects of IFNα in human β cells. EndoC-βH1 cells were transfected with an siRNA control (siCTL: grey bars) or with an siRNA targeting ADAR (white bars) and left to recover for 48h. After this period, cells were left untreated (NT) or were treated with IFNα (2000 U/ml) for 24h. (A) Protein expression was measured by western blotting using Anti-ADAR1 antibody #14175 (Cell Signaling Technology) and representative images of 4 independent experiments are shown. Densitometry results are shown for ADAR p110 (B) and ADAR p150 (C). mRNA expression of ADAR (D), STAT1 (E), MDA5 (F) and MX1 (G) was analyzed by RT-qPCR and normalized by β-actin. Values of siCTL + IFNα were considered as 1. (H) Cell death was evaluated using HO/PI staining. Data are expressed as means of independent experiments (shown as individual data points, n=4-6) ± SEM. Differences between groups were evaluated using one-way ANOVA or linear mixed model in case of missing values, followed by Bonferroni post-hoc test. *p<0.05, **p<0.01 and ****p<0.0001.

Figure 5

ADAR1 overexpression triggers alternative splicing in EndoC-βH1. (A) Global A-to-I RNA editing index across Alu elements (short interspersed nuclear elements) in Puro modified EndoC-βH1 and ADAR1 modified EndoC-βH1 cells (left panel), ADAR3 expression in Puro and ADAR1 modified EndoC-βH1. Data are shown as Transcript Per Million. (B) Donut charts representing the cumulated number of known and de novo alternative splicing events (top), and known events only (bottom). Events displayed have |ΔPSI| >0.10 and FDR <0.05. (C) Volcano plot of the inclusion and exclusion SE (skipped exon) events in EndoC-βh1 cells overexpressing ADAR1. Negative PSI indicates the inclusion of the event in ADAR1-overexpressing EndoC-βH1 cells whereas positive PSI indicates exclusion. Each dot represents an event with its ΔPSI (x-axis) associated to its P-value (y-axis). Colored dots (blue and red) represent genes with FDR<0.05. (D) Dot plot of the enrichment analysis, using Gene Ontology as reference, shows gene sets affected by known splicing events in cells overexpressing ADAR1. Gene ratio (x-axis) refers to the percentage of total genes input with alternative splicing events in selected GO terms (y-axis). SE, Spliced Exon; RI, Retain Intro; MXE, Mutually Exclusive Exon; A3SS, 3’ Alternative Splicing Site; A5SS, 5’ Alternative Splice Site.