The majority of A-to-I RNA editing is not required for mammalian homeostasis

Abstract

Background: Adenosine-to-inosine (A-to-I) RNA editing, mediated by ADAR1 and ADAR2, occurs at tens of thousands to millions of sites across mammalian transcriptomes. A-to-I editing can change the protein coding potential of a transcript and alter RNA splicing, miRNA biology, RNA secondary structure and formation of other RNA species. In vivo, the editing-dependent protein recoding of GRIA2 is the essential function of ADAR2, while ADAR1 editing prevents innate immune sensing of endogenous RNAs by MDA5 in both human and mouse. However, a significant proportion of A-to-I editing sites can be edited by both ADAR1 and ADAR2, particularly within the brain where both are highly expressed. The physiological function(s) of these shared sites, including those evolutionarily conserved, is largely unknown.

Results: To generate completely A-to-I editing-deficient mammals, we crossed the viable rescued ADAR1-editing-deficient animals (Adar1^E861A/E861AIfih1^-/-) with rescued ADAR2-deficient (Adarb1^-/-Gria2^R/R) animals. Unexpectedly, the global absence of editing was well tolerated. Adar1^E861A/E861AIfih1^-/-Adarb1^-/-Gria2^R/R were recovered at Mendelian ratios and age normally. Detailed transcriptome analysis demonstrated that editing was absent in the brains of the compound mutants and that ADAR1 and ADAR2 have similar editing site preferences and patterns.

Conclusions: We conclude that ADAR1 and ADAR2 are non-redundant and do not compensate for each other's essential functions in vivo. Physiologically essential A-to-I editing comprises a small subset of the editome, and the majority of editing is dispensable for mammalian homeostasis. Moreover, in vivo biologically essential protein recoding mediated by A-to-I editing is an exception in mammals.

Keywords: A-to-I editing; ADAR1; ADAR2; Epitranscriptome; RNA editing; RNA modification.

Fig. 1

A-to-I editing-deficient mice are viable with a normal lifespan. a Breeding data from intercrosses of Adar1^E861A/+Ifih1^−/−Adarb1^−/−Gria2^R/R males and females. b Breeding data from intercrosses of Adar1^E861A/E861AIfih1^−/−Adarb1^−/−Gria2^R/R males with Adar1^E861A/+Ifih1^−/−Adarb1^−/−Gria2^R/R females. c Survival data for all genotypes; numbers per genotype and statistical comparison across all genotypes (pairwise log-rank (Mantel-Cox) test); P value as indicated or ***P < 0.001. d Comparison of survival of Adarb1^−/−Gria2^R/+ animals with either heterozygous Adar1^E861A/+ (purple and pink lines) or homozygous Adar1^E861A/E861A (yellow) and calculated median survival in days. e Comparison of survival of Adar1^E861A/+Ifih1^−/−Adarb1^+/−Gria2^R/R (double heterozygous) and Adar1^E861A/E861AIfih1^−/−Adarb1^−/−Gria2^R/R. f Weaning weight (~ 20 days of age) of the indicated genotypes; both males and females included. *P < 0.05, **P < 0.01, ***P < 0.001 (ordinary one-way ANOVA with multiple comparisons correction (Tukey’s)). g Body weight and body mass composition of 12-week-old males of the indicated genotypes; n per genotype: C57Bl/6 = 8 (white circle); dHet (Adar1^E861A/+Ifih1^−/−Adarb1^+/−Gria2^R/R) = 5 (gray); Adar1^E861A/E861AAdarb1^+/+ (Ifih1^−/−Gria2^+/+) = 3 (red); Adar1^+/+Adarb1^−/− (Ifih1^−/−Gria2^R/R) = 3 (green); Adar1^E861A/E861AAdarb1^−/− (Ifih1^−/−Gria2^R/R) = 3 (blue); *P < 0.05 (ordinary one-way ANOVA with multiple comparisons correction (Tukey’s))

Fig. 2

Peripheral blood and hematopoietic parameters of Adar1^E861A/E861AAdarb1^−/− mice. a Red blood cell counts. b Hemoglobin. c Hematocrit. d Mean corpuscular volume. e Platelet count. f Peripheral blood leukocyte numbers and lineage distribution from the indicated genotypes, C57Bl/6 = 8 (white circle); dHet (Adar1^E861A/+Ifih1^−/−Adarb1^+/−Gria2^R/R) = 8 (gray); Adar1^+/+Adarb1^−/− (Ifih1^−/−Gria2^R/R) = 3 (green); Adar1^E861A/E861AAdarb1^−/− (Ifih1^−/−Gria2^R/R) = 7 (blue), *P < 0.05, **P < 0.01. g Cellularity of the femurs, spleen, and thymus from the indicated genotypes, n = 3 per genotype. All counts were performed on peripheral blood from 12–18-week-old male animals of the indicated genotypes. Number of animals in each genotype indicated in panel a. Statistical analysis: one-way ANOVA with correction for multiple comparisons; *P < 0.05; **P < 0.01; ***P < 0.001

Fig. 3

ADAR2 loss does not modify the transcriptional signature associated with the loss of Adar1-mediated RNA editing. Analysis of differential gene expression from 12-week-old male brains of a Adar1^+/+Ifih1^−/− (WT) compared to Adar1^E861A/E861AIfih1^−/− (E861A); b Adar1^+/+Ifih1^−/−Adarb1^−/−Gria2^R/R (ADAR2 null) compared to Adar1^E861A/+Ifih1^−/−Adarb1^+/−Gria2^R/R (dHet); c Adar1^E861A/E861AIfih1^−/−Adarb1^−/−Gria2^R/R (dKO) compared to Adar1^E861A/+Ifih1^−/−Adarb1^+/−Gria2^R/R (dHet); and d Adar1^E861A/E861AIfih1^−/−Adarb1^−/−Gria2^R/R (dKO) compared to Adar1^+/+Ifih1^−/−Adarb1^−/−Gria2^R/R (Adar2^−/−); n = 3 per genotype; red indicated FDR < 0.05. e Comparison of the differential gene expression signatures of the E861A (a) and dKO (c) samples. The increased expression of the transcripts highlighted in blue is shared between murine and human ADAR1 mutants. Top panel: y-axis has the gene expression comparison of the Adar1^E861A/E861A vs WT; x-axis has the gene expression comparison of the Adar1^E861A/E861AAdarb1^−/− (dKO) vs dHet. Lower panel: Adar1^E861A/E861A compared to Adar1^E861A/E861AAdar2^−/− (dKO) with expanded view of the upper right quadrant. f QuSAGE pathway analysis of the consensus interferon-stimulated gene (ISG)/cytokine signature defined by Liu et al. [20] for the Adar1^E861A/E861A compared to Adar1^+/+Ifih1^−/− (left panel), Adarb1^−/− compared to dHet (center panel), and Adar1^E861A/E861AAdarb1^−/− compared to dHet (dKO; right panel); log₂FC, P value and FDR as indicated on each panel

Fig. 4

Adar1^E861A/E861AAdarb1^−/− have lost A-to-I editing across the transcriptome. a Analysis of A-to-I editing of the Htr2c receptor at the known sites A–D by Sanger sequencing. Genotypes as indicated. b Analysis of editing sites across the genotypes. A dataset of 57,077 murine editing sites was compiled and the datasets assessed for editing at these sites. Sites required ≥ 50 read coverage and an editing rate of ≥ 0.01 (≥ 1%) to be included. The number that passed this threshold for each comparison is listed, and the numbers that are significantly different based on the z factor (z ≥ 5; Jacusa analysis method) are indicated in red. c Editing frequency across coding/site-selective and repetitive/hyperediting sites in the transcriptome in the individual samples from the WT, dHet, Adar1^E861A/E861A, and Adarb1^−/−. Sites required ≥ 50 read coverage and an editing rate of ≥ 0.01 (≥ 1%) to be considered. Boxplot represents the 25% quantile to 75% quantile with the median indicated. d Editing of the 3′UTR of Rpa1 transcript in each of the indicated genotypes

Fig. 5

ADAR1- and ADAR2-specific sites are comparable. a Analysis of evolutionarily conserved A-to-I editing events across the genotypes. Average editing for each site was calculated plotted with reference to the levels at each site identified by Pinto et al. [29]. ADAR1/ADAR2 shared sites are defined as having > 10% and < 150% editing compared to the average editing rate of the WT and Adar1^E861A/+Adarb1^+/− (dHet) samples combined (WT+dHet); ADAR1-specific sites have < 10% editing of this site in the Adar1^E861A/E861A samples and unchanged editing in the Adarb1^−/− compared to WT+dHet; ADAR1-specific/ADAR2 inhibits sites have < 10% editing of this site in the Adar1^E861A/E861A samples and > 150% editing of WT+dHet levels in the Adarb1^−/−; ADAR2-specific sites have < 10% editing of this site in the Adarb1^−/− samples and unchanged editing in the Adar1^E861A/E861A compared to WT+dHet; ADAR2-specific/ADAR1 inhibits sites have < 10% editing of this site in the Adarb1^−/− samples and > 150% editing of WT+dHet levels in the Adar1^E861A/E861A. b Quantitation of the numbers of sites (≥ 50 read coverage and an editing rate of ≥ 0.01 (≥ 1%)) and genomic location across genotypes. ADAR1- or ADAR2-specific sites were defined as having < 10% editing of a site in the one genotype and retained editing in the alternative genotype. The percentage of sites that are ADAR1 or ADAR2 specific is indicated in brackets as a percent of the total number of sites for each location. The sequence context of the editing sites for each classification was derived with Seqlogo. The distribution of editing in B1 and B2 SINEs was mapped from the total sites identified in each genotype. c The genomic distribution/repeat type and average editing level for the ADAR1 and ADAR2 sites compared to the all sites observed in the control (WT+dHet genotype combined). Box and whiskers plot with 5–95 percentile shown. No significant difference between genotypes or P value as indicated (ANOVA with multiple comparisons correction)