Activated STING in the thymic epithelium alters T cell development and selection leading to autoimmunity

Abstract

Coatomer protein complex subunit α (COPA) syndrome is a monogenic disorder of immune dysregulation that leads to interstitial lung disease and high-titer autoantibodies. Constitutive activation of the innate immune molecule stimulator of interferon genes (STING) is centrally involved in disease. However, the mechanisms by which STING results in autoimmunity are not well understood in COPA syndrome and other STING-associated diseases. Prior studies showed a cell autonomous role for STING in thymocyte development. Single-cell data of human thymus demonstrated that STING is highly expressed in medullary thymic epithelial cells (mTECs) and at levels much greater than in T cells. Here, we show that in certain contexts, activated STING exerts a functional role in the thymic epithelium to alter thymocyte selection and predisposes to autoimmunity. In CopaE241K/+ mice, activated STING in mTECs amplified IFN signaling, impaired macroautophagy, and caused a defect in negative selection of T cell precursors. WT mice given a systemic STING agonist phenocopied the selection defect and showed enhanced thymic escape of a T cell clone targeting a self-antigen also expressed in melanoma. Our work demonstrates that STING activation in TECs shapes the T cell repertoire and contributes to autoimmunity, findings that are important for conditions that activate thymic STING.

Keywords: Autoimmune diseases; Autoimmunity; Immunology; Innate immunity; Tolerance.

Figure 1. Activated STING in thymic stroma upregulates IFN signaling.

(A) Representative immunoblot of pSTING in TECs versus thymocytes with relative band density quantification. Each lane represents 1 biological replicate. Data represent 3 independent experiments. (B) Immunofluorescence stain of keratin 5 (KRT5) and pSTING expression on thymic sections from Copa^E241K/+ mice. Scale bar: 100 μm. (C) Volcano plot of RNA sequencing analysis of sorted mTECs (mTEC^hi) from WT and Copa^E241K/+ mice (n = 3 per genotype). (D) Immunofluorescence stain of KRT5 and KRT8 on thymic sections from Copa^E241K/+ and WT littermates. Scale bar: 250 μm. Data represent 3 independent experiments. (E) Left: representative flow cytometry analysis of TECs via MHC-II and Ly51 staining. Right: total mTEC numbers (WT, n = 3; Copa^E241K/+, n = 4; Copa^E241K/+/Sting^gt/gt, n = 3). Data represent 2 independent experiments. Data are mean ± SD. One-way ANOVA with Bonferroni’s multiple-comparison test was used for statistical analysis. A P value of less than 0.05 was considered statistically significant.

Figure 2. Activated STING in thymic stroma increases SP thymocytes.

(A) Left: representative flow plots of CD4⁺ and CD8⁺ on reconstituted thymocytes in bone marrow chimeras. Right: percentages of CD4⁺ SP and CD8⁺ SP thymocytes among the reconstituted thymocytes (WT→WT, n = 6; WT→Copa^E241K/+, n = 6; WT→Sting^gt/gt, n = 5; WT→Copa^E241K/+/Sting^gt/gt, n = 5). (B) Left: representative flow analysis of CD69 and TCR-β on reconstituted thymocytes in bone marrow chimeras. Right: percentages of CD69^hi TCRβ^hi and CD69^lo TCRβ^hi among the reconstituted thymocytes. (C) Percentages of CD69^lo MHC-II^hi (mature stage 2) among the reconstituted CD4⁺ and CD8⁺ SP thymocytes. Flow gating strategy is in Supplemental Figure 5A. WT→WT, n = 7; WT→Copa^E241K/+, n = 6; Copa^E241K/+→WT, n = 6; Copa^E241K/+→Copa^E241K/+, n = 4; WT→Sting^gt/gt, n = 4; WT→Copa^E241K/+/Sting^gt/gt, n = 3. Data in A–C were pooled from at least 2 independent experiments and are mean ± SD. Two-way ANOVA with Šidák’s multiple-comparison test was used for statistical analysis in A and B. One-way ANOVA and Bonferroni’s multiple-comparison test were used in C. A P value of less than 0.05 was considered statistically significant.

Figure 3. Constitutive activation of STING in the thymus impairs autophagic flux.

(A) Left: flow analysis of autophagosome-associated LC3 (LC3II) in mTEC^hi from GFP-LC3 and GFP-LC3 × Copa^E241K/+ mice. Right: percentage of LC3II-GFP⁺ population among total mTEC^hi and GFP median fluorescence intensity in mTEC^hi (GFP-LC3 × WT, n = 4; GFP-LC3 × Copa^E241K/+, n = 4). Data were pooled from 2 independent experiments. (B) Representative immunoblot and densitometric analysis of LC3II following transient transfection of WT or E241K COPA-expressing plasmid into HEK293T cells that stably express STING. Data represent 3 independent experiments. (C) Quantitation of autophagic flux in mTECs of CAG-RFP-GFP-LC3 tandem reporter mice. Left: flow cytometry of autophagosome-associated (RFP≈GFP) and autolysosome-associated (GFP<RFP) LC3 in mTECs. Right: percentage of mTEC^hi with reduced autophagic flux (RFP-GFP-LC3 × WT, n = 5; RFP-GFP-LC3 × Copa^E241K/+, n = 5; RFP-GFP-LC3 × WT × Sting^gt/gt, n = 3; RFP-GFP-LC3 × Copa^E241K/+ × Sting^gt/gt, n = 3). (D) Left: ratio of RFP/GFP fluorescence histogram in mTECs expressing LC3 tandem reporter. Right: mean RFP/GFP ratio in mTECs (RFP-GFP-LC3 × WT, n = 5; RFP-GFP-LC3 × Copa^E241K/+, n = 5; RFP-GFP-LC3 × WT × Sting^gt/gt, n = 3; RFP-GFP-LC3 × Copa^E241K/+ × Sting^gt/gt, n = 3). Data in C and D were pooled from 3 independent experiments and are presented as mean ± SD. Unpaired, parametric, 2-tailed Student’s t test was used for statistical analysis in A. One-way ANOVA with Bonferroni’s multiple-comparison test was used in C and D. A P value of less than 0.05 was considered statistically significant.

Figure 4. Activated STING impairs negative selection of T cells and alters the T cell repertoire.

(A) Left: representative flow analysis of CD4⁺ and CD8⁺ on reconstituted thymocytes in bone marrow chimeras. Right: percentage of CD4⁺ SP thymocytes among the reconstituted thymocytes (OT-II→Rip-mOVA × WT, n = 11; OT-II→Rip-mOVA × Copa^E241K/+, n = 5; WT→Rip-mOVA × WT × Sting^gt/gt, n = 10; WT→Rip-mOVA × Copa^E241K/+ × Sting^gt/gt, n = 4). (B) Left: flow analysis of TCR-β and Nur77 expression in the reconstituted CD4⁺ SP in the bone marrow chimeras. Right: percentage of Nur77⁺ population among CD4⁺ SP thymocytes (OT-II→Rip-mOVA × WT, n = 7; OT-II→Rip-mOVA × Copa^E241K/+, n = 5; WT→Rip-mOVA × WT × Sting^gt/gt, n = 8; WT→Rip-mOVA × Copa^E241K/+ × Sting^gt/gt, n = 4). (C) Left: flow analysis of Vβ5 and Vα2 in CD4⁺ SP thymocytes in the bone marrow chimeras shown in (A). Right: ratio of TCR^hi versus TCR^lo among CD4⁺ SP thymocytes. (D) Left: flow analysis of CD25 and Foxp3 in CD4⁺ thymocytes in the bone marrow chimeras. Right: percentage of thymic Tregs among CD4⁺ thymocytes (OT-II→Rip-mOVA × WT, n = 7; OT-II→Rip-mOVA × Copa^E241K/+, n = 5; WT→Rip-mOVA × WT × Sting^gt/gt, n = 8; WT→Rip-mOVA × Copa^E241K/+ × Sting^gt/gt, n = 4). Data were pooled from 3 independent experiments. Data are mean ± SD. One-way ANOVA and Bonferroni’s multiple-comparison test (log normal distribution in D) were used for statistical analysis. A P value of less than 0.05 was considered statistically significant. All host mice are on Rip-mOVA background.

Figure 5. A systemic STING agonist increases autoreactive T cells in the thymus.

(A) Left: CD4⁺ and CD8⁺ profile of thymocytes and CD3⁺ and Vβ14⁺ expression of CD4⁺ SP thymocytes in Rag1^–/– Tyrp1^B-w/wt TCR mice treated with diABZI STING agonist or vehicle. Right: change in percentage of total thymic CD4⁺ SP and Vβ14⁺CD4⁺ SP following STING agonist treatment (vehicle, n = 14; agonist, n = 21; compiled from 3 independent experiments). (B) Left: cleaved caspase 3 on CD5^hi TCRβ^hi thymocytes in vehicle- and agonist-treated mice. Right: percentage of thymocytes undergoing clonal deletion in agonist-treated mice relative to vehicle treatment (vehicle, n = 13; agonist, n = 14; compiled from 3 independent experiments). Unpaired, parametric, 2-tailed Student’s t test was used for statistical analysis. (C) Left: flow analysis of splenic CD4⁺ SP Vβ14⁺ autoreactive T cells in vehicle- and agonist-treated mice. Right: absolute number of CD4⁺ SP Vβ14⁺ autoreactive T cells. (D) Absolute number of splenic and inguinal CD4⁺ SP and Vβ14⁺ autoreactive T cells in vehicle- and agonist-treated mice (vehicle, n = 14; agonist, n = 21; 3 independent experiments). LN, lymph node. (E) B16 melanoma growth in Rag1^–/– mice that received x-ray radiation, PD-1 antibody, and adoptively transferred splenocytes from Rag1^–/– Tyrp1^B-w/wt TCR mice treated with STING agonist (n = 5) or vehicle (n = 4) (data represent 3 independent experiments). Locally estimated scatter plot smoothing with 95% confidence interval of B16 tumor growth over time. (F) Quantitation of CD3⁺ CD45.1 donor cells within the periphery after transfer into CD45.2 host and treatment with agonist (n = 8) or vehicle (n = 9). Data were pooled from 3 independent experiments. (G) Mean expression of STING1 transcript in select cell types in human thymus. Data are mean ± SD. Two-tailed Mann-Whitney U test was used for statistical analysis unless indicated above. A P value of less than 0.05 was considered statistically significant.