Aberrant RNA sensing in regulatory T cells causes systemic autoimmunity

Abstract

Chronic and aberrant nucleic acid sensing causes type I IFN-driven autoimmune diseases, designated type I interferonopathies. We found a significant reduction of regulatory T cells (T_regs) in patients with type I interferonopathies caused by mutations in ADAR1 or IFIH1 (encoding MDA5). We analyzed the underlying mechanisms using murine models and found that T_reg-specific deletion of Adar1 caused peripheral T_reg loss and scurfy-like lethal autoimmune disorders. Similarly, knock-in mice with T_reg-specific expression of an MDA5 gain-of-function mutant caused apoptosis of peripheral T_regs and severe autoimmunity. Moreover, the impact of ADAR1 deficiency on T_regs is multifaceted, involving both MDA5 and PKR sensing. Together, our results highlight the dysregulation of T_reg homeostasis by intrinsic aberrant RNA sensing as a potential determinant for type I interferonopathies.

Fig. 1.. Patients with AGS caused by ADAR1 or IFIH1 mutations have a decreased frequency of peripheral effector T_regs.

(A) Representative flow cytometry (FC) plots of CD25 and CD45RA expression on CD4⁺ T cells from controls or patients with AGS. Fr. I: CD25^lowCD45RA⁺ suppressive resting T_regs, Fr. II: CD25^hiCD45RA⁻ highly suppressive effector T_regs, and Fr. III: CD25^lowCD45RA⁻ (FOXP3^low) nonsuppressive T cells. (B) Summarized percentages of Fr. I, Fr. II, and Fr. III from all controls and patients with AGS analyzed in this study. (C to F) Representative histograms of CTLA-4 and PD-1 expression on Fr. I (solid line) and Fr. II (dotted line) and summarized mean fluorescence intensity (MFI) values from controls (black) and patients with AGS (yellow), relative to Fr. I in healthy controls (HC). The samples from patients with AGS have been analyzed one at a time (in two instances, two at a time), together with control samples from healthy donors. The dot plots shown here contain pooled data from respective analyses. Samples from patients 1 and 4 have been analyzed at four [three for CTLA-4 and PD-1 expression; (D) and (F) and fig. S1D] and two different time points (fig. S1A), and the mean is represented as one symbol in pooled-data dot plots (B, D, and F); otherwise, each symbol represents one individual. Statistics were calculated using Student’s t test with Welch’s correction (B) and one-way ANOVA (D and F); *P ≤ 0.05; **P ≤ 0.01; ****P ≤ 0.0001; ns, not significant, P > 0.05.

Fig. 2.. Adar1 deletion in T_regs causes T_reg loss and a scurfy-like phenotype in mice.

(A and B) Pictures and body weight measurements of 3-week-old Foxp3-WT (n = 44) and Foxp3^ΔAdar1 (n = 36) mice. (C) Survival graphs of Foxp3-WT and Foxp3^ΔAdar1 mice (n = 7, but all Foxp3^ΔAdar1 mice developed a severe phenotype and were sacrificed at latest 4 weeks after birth). (D) Representative H&E staining images. S. intestine, small intestine. Scale bars, 100 μm. (E to H) Representative FC plots and summarized percentages (%) of CD4⁺FOXP3⁺ T_regs in the spleens. (I and J) Representative FC plots and summarized percentages of naïve (CD44⁻CD62L⁺), effector (CD44⁺CD62L⁻), and memory (CD44⁺CD62L⁺) CD4⁺ T cells in the spleens. (K) ELISA heatmap of indicated cytokines measured in supernatants from enriched CD4⁺ T cells (pg/ml), stimulated overnight with anti-CD3/anti-CD28 antibody–coated beads. (L) Representative FC plots of FOXP3⁺ versus GATA3⁺CD4⁺ T cells after overnight culture (without stimulation). Panels (E) to (J) are representative of ≥3 independent experiments with ≥3 mice per group. Panels (K) and (L) are representative of two experiments with two mice per group. In dot plots, each symbol indicates an individual mouse. Statistics were calculated using Student’s t test; ****P ≤ 0.0001; ns, not significant, P > 0.05.

Fig. 3.. Constitutive MDA5 activation in T_regs causes T_reg loss and autoimmunity in mice.

(A and B) Pictures and body weight measurements of 4-week-old Foxp3-WT (n = 12) and Foxp3-GS (n = 9) mice. (C) Survival graph of Foxp3-WT (n = 12) and Foxp3-GS (n = 14) mice. (D) Representative H&E staining images of indicated organs. Scale bars, 100 μm. (E) Representative H&E staining images of kidneys, immunofluorescence staining images of IgG (green) in the kidneys (4′,6-diamidino-2-phenylindole (DAPI), blue), and immunofluorescence staining of L929 cells using sera from Foxp3-WT and Foxp3-GS mice. (F) Relative mRNA expression of indicated genes. (G and H) Representative FC plots, summarized percentages (%), and total numbers (#) of CD4⁺YFP⁺(FOXP3⁺) T_regs in spleens and small intestines. Panels (F) to (H) are representative of ≥3 independent experiments with ≥3 mice per group. In dot plots, each symbol indicates an individual mouse. Statistics were calculated using Student’s t test; *P ≤ 0.05; **P ≤ 0.01; ****P ≤ 0.0001; ns, not significant, P > 0.05.

Fig. 4.. Adar1 deletion in T_regs activates both the MDA5/MAVS and PKR/eIF-2α pathways contributing to cell death.

(A) Annexin V⁺7AAD⁺ T_reg percentages in the spleens and mesenteric lymph nodes (MLNs). (B) Relative mRNA expression of Noxa in sorted CD4⁺YFP⁺(FOXP3⁺) T_regs. (C and D) Percentages of CD4⁺FOXP3⁺ ex vivo induced T_regs (iT_regs) at day 3 and their viability at day 5. iT_regs were induced from Foxp3-WT (n = 3) or Foxp3^ΔAdar1 (n = 5) naïve CD4⁺ T cells and treated or not with caspase inhibitor. (E and F) Representative FC plots and summarized percentages of CD4⁺FOXP3⁺ T_regs in the spleens of 30-week-old Cx3cr1^ΔAdar1, 10-week-old Cx3cr1-GS, and respective Cx3cr1-WT control mice. (G) Body weight measurements of 3-week-old Foxp3-WT (n = 44), Foxp3^ΔAdar1 (n = 36), and Mavs^−/−Foxp3^ΔAdar1 (n = 20) mice. (H) Survival graphs of Foxp3-WT (gray), Foxp3^ΔAdar1 (blue, n = 7), and Mavs^−/−Foxp3^ΔAdar1 mice (orange, n = 5). (All Foxp3^ΔAdar1 and Mavs^−/−Foxp3^ΔAdar1 mice to date developed a severe phenotype and were sacrificed at latest 4 or 8 weeks old, respectively.) (I to L) Representative FC plots and summarized percentages of CD4⁺FOXP3⁺ T_regs in the spleens. (M to P) Representative histograms and plots of phospho–eIF-2α mean fluorescence intensity in CD4⁺FOXP3⁺ T_regs from spleens. Panels (A), (B), (I), (J), (M), and (N) are representative of ≥3 independent experiments with ≥3 mice per group; panels (C) to (F) are representative of two independent experiments with ≥2 mice per group. In dot plots, each symbol indicates an individual mouse. Statistics were calculated using Student’s t test or one-way ANOVA; *P ≤ 0.05; **P ≤ 0.01; ***P ≤ 0.001; ****P ≤ 0.0001; ns, not significant, P > 0.05.