The phenotype of the most common human ADAR1p150 Zα mutation P193A in mice is partially penetrant

Abstract

ADAR1 -mediated A-to-I RNA editing is a self-/non-self-discrimination mechanism for cellular double-stranded RNAs. ADAR mutations are one cause of Aicardi-Goutières Syndrome, an inherited paediatric encephalopathy, classed as a "Type I interferonopathy." The most common ADAR1 mutation is a proline 193 alanine (p.P193A) mutation, mapping to the ADAR1p150 isoform-specific Zα domain. Here, we report the development of an independent murine P195A knock-in mouse, homologous to human P193A. The Adar1^P195A/P195A mice are largely normal and the mutation is well tolerated. When the P195A mutation is compounded with an Adar1 null allele (Adar1^P195A/- ), approximately half the animals are runted with a shortened lifespan while the remaining Adar1^P195A/- animals are normal, contrasting with previous reports. The phenotype of the Adar1^P195A/- animals is both associated with the parental genotype and partly non-genetic/environmental. Complementation with an editing-deficient ADAR1 (Adar1^P195A/E861A ), or the loss of MDA5, rescues phenotypes in the Adar1^P195A/- mice.

Keywords: A-to-I RNA editing; ADAR1; MDA5; P193A mutation; Zα domain.

Figure 1. Generation and characterisation of an Adar1 ^P195A knock‐in allele

1. A

   Schematic of the murine wild‐type ADAR1 isoforms and the location of the P195A mutation.

2. B

   Genomic alignment and translation of the WT and P195A allele.

3. C

   Sanger sequencing traces and alignments of genomic DNA isolated from animals from Adar1 ^+/+ (WT) and Adar1 ^P195A/P195A .

4. D

   Western blot analysis of ADAR1 expression in E13.5‐immortalised MEFs of the indicated genotypes +/− interferon‐β (IFNβ). Data from two independent cell lines per genotype.

5. E

   Results from inbreeding of Adar1 ^P195A/+ animals.

6. F

   Weaning weights of mice of the indicated genotypes;

7. G

   Survival analysis of mice of the indicated genotypes; number as indicated for each genotype. Statistical difference determined by Log‐rank (Mantel–Cox) test.

8. H

   Total white blood cell counts (WBC) in peripheral blood (PB).

9. I

   Absolute numbers of each lineage in PB. *P < 0.05. Significance was determined by a two‐way ANOVA test with multiple comparisons.

10. J–L

    Peripheral blood, (J) red blood cell (RBC), (K) haematocrit (HCT) and (L) platelets.

Data information: Data in (F), (H), (J), (K) and (L) represent the mean ± SEM. *P < 0.05; **P < 0.01; Significance determined by ordinary one‐way ANOVA with Tukey's multiple‐comparison tests (adjusted P‐value). Source data are available online for this figure.

Figure 2. In vivo expression of P195A is well tolerated and results in a modest induction of interferon‐regulated gene expression

1. A

   Schematic outline of the experiment.

2. B

   Representative genotyping of recombination of the Adar1 floxed allele at day 28 using genomic DNA isolated from whole bone marrow cells. Recombination percentage was calculated using LabChip (PerkinElmer)‐based quantitation of band intensity compared to the WT/P195A allele and known standard/marker.

3. C

   Survival analysis of mice of the indicated genotypes; number as indicated for each genotype. No statistical difference between genotypes by Log‐rank (Mantel–Cox) test.

4. D

   Percentage change in body weight of each cohort based on comparison of the weight at day 28 of tamoxifen food compared to day 0 (prior to initiation of tamoxifen‐containing diet).

5. E

   Peripheral blood leukocyte populations (by lineage) between genotypes at day 0 and day 28.

6. F

   Bone marrow (BM) cellularity.

7. G

   Differential analysis of leukocyte populations in the BM.

8. H

   BM erythroid cells.

9. I

   Splenic cellularity.

10. J

    Differential analysis of leukocyte populations in the spleen.

11. K

    Thymic cellularity.

12. L

    Differential analysis of thymocyte populations in the thymus.

13. M

    Representative flow cytometry histography of Sca‐1 expression between control (Δ/+; grey) and P195A only (Δ/P195A; blue) expressing BM sample. Quantitation of mean Sca‐1 fluorescence intensity within the lineage negative fraction of whole bone marrow.

14. N

    qPCR (SYBR green)‐based analysis of indicated gene expression in BM. Data expressed as mean ± SEM gene expression related to Ppia expression.

Data information: Unless otherwise stated, data expressed as mean ± SEM; R26‐CreER^ki/+ Adar1 ^fl/+ (control; n = 3) and R26‐CreER^ki/+ Adar1 ^fl/P195A (P195A; n = 4). All BM, spleen and thymic analysis at day 28 post‐tamoxifen treatment; statistical comparison in plots was done by T‐test (panels D, F, I, K and N) or two‐way (panels E, G, H, J, M and L) ANOVA tests with multiple comparisons with statistical significance of **P < 0.01 and ***P < 0.001. Source data are available online for this figure.

Figure 3. The P195A mutation has a minimal impact on physiological A‐to‐I editing

1. A

   Analysis of differential gene expression from mouse brains of Adar1 ^P195A/P195A (P195A/P195A) compared to Adar1 ^+/+ (WT). Gene expression changes that result in an increased relative gene expression in the Adar1 ^P195A/P195A genotype are above the mid‐point. Red indicates genes that are significantly different based on abs(log₂FC) > 1, FDR < 0.05.

2. B

   Murine modified version of the Alu editing index (AEI) of all samples of the indicated genotype.

3. C

   Editing frequency of known sites in P195A/P195A compared to WT by JACUSA. Sites with at least 50 reads and with an editing rate of not < 0.01 (1%). Red indicates the numbers that are significantly different based on the JACUSA call‐2 statistic > 5.

4. D

   Summary of the total A‐to‐I editing sites shared by both WT and P195A/P195A mice and the number of up‐edited or down‐edited sites in P195A/P195A compared to WT.

5. E

   Total genomic locations of the edited sites in WT or P195A/P195A.

6. F

   The genomic distribution of the significant sites in P195A/P195A compared to WT.

7. G

   Editing frequency of sites and sites in genes reported to be subject to Z‐RNA formation in P195A/P195A compared to WT by JACUSA. Light blue indicates sites that pass both editing and the Z‐RNA threshold, and dark blue dots represent the significantly differentially edited sites in Z‐RNA formation genes.

Figure 4. Weaning weights and survival of Adar1 ^P195A/E861A and Adar1 ^P195A/− animals

1. A–C

   (A) Results from inbreeding of Adar1 ^P195A/E861A animals, (B) weaning weights and (C) survival analysis of mice of the indicated genotypes.

2. D, E

   (D) Weaning weights and (E) survival of mice from inbreeding to generate P195A/− animals.

3. F–H

   (F) Results from breeding of Adar1 ^P195A/P195A animals with Adar1 ^+/− animals, (G) weaning weights and (H) survival of pups derived from breeding of an Adar1 ^P195A/P195A to an Adar1 ^+/− animal; number as indicated for each genotype.

4. I–K

   (I) Results from breeding of Adar1 ^P195A/− animals with Adar1 ^P195A/+ animals, (J) weaning weights and (K) survival of pups derived from breeding an Adar1 ^P195A/− to an Adar1 ^P195A/+ genotype; number indicated for each genotype. Inset photo: 35‐day‐old Adar1 ^P195A/− male bred from Adar1 ^P195A/P195A × Adar1 ^+/− parents (data in Panel F); and sibling 29‐day‐old Adar1 ^P195A/+ and Adar1 ^P195A/− male bred from an Adar1 ^P195A/− × Adar1 ^P195A/+ (data in Panel I).

Data information: The significance of the frequency of genotype (A, F and I) was performed by Chi‐square test and weaning weights in (B), (D) and (J) by ordinary one‐way ANOVA with Tukey's multiple‐comparison test (adjusted P‐value). Statistical analysis of weights in panel G used an unpaired t‐test (two‐sided). Significance difference in survival plots used Log‐rank (Mantel–Cox) test. Weaning weights are presented as mean ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001.

Figure 5. Loss of MDA5 rescues both the weight and viability of Adar1 ^P195A/−

1. A, B

   (A) Weaning weights and (B) survival of Adar1 ^P195A/E861A Ifih1 ^−/− and littermates (all Ifih1 ^−/−).

2. C, D

   (C) Weaning weights and (D) survival of Adar1 ^P195A/‐ Ifih1 ^−/− and littermates (all Ifih1 ^−/−).

Data information: Weaning weights are presented as mean ± SEM. No statistically significant difference across any comparison (ordinary one‐way ANOVA with Tukey's multiple‐comparison test) in (A) and (C). No statistical difference between genotypes by Log‐rank (Mantel–Cox) test in (B) and (D).

Figure 6. Upregulation of ISGs and tissue‐specific elevation of ISR in Adar1 ^P195A/− are MDA5 dependent

1. A

   QuSAGE analysis using ISGs and ISR gene signatures from previous publications (Material and Methods; Gene set testing (QuSAGE and heatmap analysis)) as the gene set. Every single curve depicts a gene within the barcode. Data from brain samples of Adar1 ^P195A/P195A (P195A/P195A) compared to Adar1 ^+/+ (WT) or Adar1 ^P195A/−(P195A/−; non‐runted) compared to WT (this study).

2. B

   QuSAGE analysis of ISGs and ISR genes in the whole‐brain RNA‐seq data of homozygous Adar1 ^W197A/W197A compared to WT models published by Nakahama et al, .

3. C

   QuSAGE analysis of ISGs and ISR genes in the spleen RNA‐seq data of Adar1 ^{P195A/p150‐} or Adar1 ^{P195A/ ‐} Eif2ak2 ^−/− published by Maurano et al (2021).

4. D

   Heatmap representation of the normalised expression of interferon‐stimulated genes (Oas1a, Ifi27 and Irf7) and the integrated stress response genes (Asns, Cdkn1a and Hmox) in the brain of the indicated genotypes using Taqman‐based qPCR.

5. E

   Heatmap representation of the normalised expression of ISGs and ISR genes in the liver of the indicated genotypes using Taqman‐based qPCR.

6. F

   Individual plots of the normalised expression of ISGs and ISR genes in the kidney of the indicated genotypes using Taqman‐based qPCR.

7. G

   Western blot of PKR, phospho‐eIF2α and total eIF2α levels in whole kidney lysates derived from the indicated genotypes; same mice as used for panel F. Relative phospho‐eIF2α expression was calculated against total eIF2α normalised to the average of P195A/+ samples using ImageJ Fiji. All images are from the same gel and re‐probed for different antibodies.

Data information: Data presented in panels (D–F) were relative to Hprt expression and normalised to the average value of the WT samples. Statistical analysis for all qPCR was two‐way ANOVA (D, E) or ordinary one‐way ANOVA with multiple‐comparison test (F); *P < 0.05; ***P < 0.001, ****P < 0.0001. Source data are available online for this figure.

Figure 7. P195A/Δ sensitises to cell death following treatment with IFNβ

1. A

   Schematic outline of how the HOXA9‐immortalised myeloid cell lines are derived.

2. B

   Experimental outline.

3. C

   Genomic DNA genotyping demonstrating efficient recombination of the floxed Adar1 allele following tamoxifen treatment.

4. D, E

   (D) Average proliferation and (E) viability of cell lines with and without tamoxifen treatment (isogenic pairs) over 14 days of treatment (n = 3 independent lines per genotype; biological replicates).

5. F

   qPCR (SYBR green) analysis of Ifit1 (left) and Irf7 (right) expression on day 14 of analysis.

6. G

   Cell viability at IFNβ dosages 0, 50 and 1,000 U/ml after 24 h. Statistical analysis was two‐way ANOVA with multiple‐comparisons correction.

7. H

   qPCR (SYBR green) of Ifit1 (left) and Irf7 (right) expression in the cell lines collected at 3 or 24 h post‐treatment with 50 or 1,000 U/ml of IFNβ. No statistical difference between genotypes in Ifit1 (left) and Irf7 (right) expression.

Data information: For data in panels (D) through (H), the experiments were performed with three independently derived cell lines per genotype (biological replicates). The qPCR data are expressed as mean ± SEM gene expression relative to Ppia expression. *P < 0.05; ***P < 0.001, ****P < 0.0001. Source data are available online for this figure.