Activated cGAS/STING signaling elicits endothelial cell senescence in early diabetic retinopathy

Abstract

Diabetic retinopathy (DR) is a leading cause of blindness in working-age adults and remains an important public health issue worldwide. Here we demonstrate that the expression of stimulator of interferon genes (STING) is increased in patients with DR and animal models of diabetic eye disease. STING has been previously shown to regulate cell senescence and inflammation, key contributors to the development and progression of DR. To investigate the mechanism whereby STING contributes to the pathogenesis of DR, diabetes was induced in STING-KO mice and STINGGT (loss-of-function mutation) mice, and molecular alterations and pathological changes in the retina were characterized. We report that retinal endothelial cell senescence, inflammation, and capillary degeneration were all inhibited in STING-KO diabetic mice; these observations were independently corroborated in STINGGT mice. These protective effects resulted from the reduction in TBK1, IRF3, and NF-κB phosphorylation in the absence of STING. Collectively, our results suggest that targeting STING may be an effective therapy for the early prevention and treatment of DR.

Keywords: Cellular senescence; Ophthalmology; Retinopathy; Signal transduction.

Figure 1. Elevated cGAS, STING, type I IFN, and free DNA levels in patients with diabetic retinopathy.

Representative immunoblots (A) and densitometry graphs show that the protein levels of cGAS (B) and STING (C) are increased in the retina of human DR donor cadaver tissue (Ctl, n = 8 DR, n = 13). Immunostaining of STING in human PDR donor retinal sections shows that (D) STING is present (arrow) in endothelial cells, (E) intraretinal neovascularization, (F) faintly in an occluded capillary and strongly in surrounding microglia/macrophages, and (G) in endothelial cells and surrounding microglia/macrophages of a hyalinized capillary microaneurysm. Scale bars: 25 μm. The levels of IFN-α (H) and IFN-β (I) in the aqueous humor, and IFN-α (J) and IFN-β (K) in the vitreous humor of patients without (Ctl) and with DR or PDR. (Ctl, n = 20, DR, n = 23; NPDR, n = 10; PDR, n = 13 for H and I; Ctl, n = 13, PDR, n = 30 for J and K). (L) The levels of free DNA in the vitreous humor of patients without (n = 11) and with PDR (n = 17). Data represent mean ± SEM. *P < 0.05; **P < 0.01; ***P < 0.001; ****P < 0.0001 by Mann-Whitney U test for B, C, H (left), I (left), and J–L, or Kruskal-Wallis test for H (right) and I (right). Ctl, control; NPDR, nonproliferative diabetic retinopathy; PDR, proliferative diabetic retinopathy.

Figure 2. cGAS/STING signaling pathway is activated in the retina of diabetic mice.

Representative immunoblots (A) and densitometry graphs show that the protein levels of (B) cGAS and (C) STING expression was increased in the retina of diabetic (D) mice (diabetes induction for 2 months) compared with nondiabetic (N) controls. Immunofluorescence study shows that (D) STING is localized in the retinal nerve fiber layer (asterisk) of both nondiabetic and diabetic mice, and more were observed in retinal endothelial cells in diabetic retina. Red, STING⁺ cells; green, CD31⁺ endothelial cells. Scale bar: 50 μm. ELISA shows that (E) IFN-α (not significant) and (F) IFN-β (significant) were increased in the diabetic retina compared with nondiabetic controls, which is correlated with pericytes loss (G and H; arrowheads indicate pericytes). (I) Flat-mount micrographs from WT diabetic mice (2 months of diabetes, 4 months of age) and age-matched nondiabetic (N) controls showing blood vessel hyperpermeability; arrows indicate leakage sites. Scale bars: 50 μm. (J) Quantification of FITC-BSA leakage into the retina. Representative immunoblots (K) show increased inflammatory factors iNOS (L) and ICAM-1 (M) and superoxide production (N) in diabetic mice. In A–N, n = 6 for each group. Data are expressed as mean ± SD. *P < 0.05, **P < 0.01, ***P < 0.001 versus nondiabetic controls by unpaired, 2-tailed Student’s t test.

Figure 3. STING KO attenuates diabetes-induced retinal capillary degeneration and vascular leakage.

(A) Representative micrographs of retinal vessels from diabetic mice (8 months of diabetes) and age-matched nondiabetic mice. Arrows indicate degenerated capillaries and arrowheads indicate capillary pericytes. Scale bars: 50 μm. (B) Diabetes increased the number of degenerated capillaries and (C) decreased the number of retinal capillary pericytes in WT diabetic mice compared with nondiabetic controls. STING KO attenuated such alterations caused by diabetes. (D) Diabetes-induced accumulation of FITC-BSA in the mouse retina was significantly reduced in STING-KO diabetic mice compared with WT diabetic controls. n = 6 mice for each group; the data are expressed as mean ± SD. Statistical differences were examined by ordinary 1-way ANOVA followed by Tukey’s multiple-comparison test. *P < 0.05; **P < 0.01; ***P < 0.001. N, control; KO, knockout.

Figure 4. STING^GT attenuates diabetes-induced retinal capillary degeneration and vascular leakage.

(A) Representative micrographs of retinal vessels from diabetic mice (8 months of diabetes) and age-matched nondiabetic mice. Arrows indicate degenerated capillaries and arrowheads indicate capillary pericytes. Scale bars: 50 μm. (B) Diabetes increased the number of degenerated capillaries and (C) decreased the number of retinal capillary pericytes in WT diabetic mice compared with nondiabetic controls. In STING^GT mice, such alterations caused by diabetes were attenuated. (D) Diabetes-induced accumulation of FITC-BSA in the mouse retina was significantly reduced in STING^GT diabetic mice compared with WT diabetic controls. n = 6 mice for each group; the data are expressed as mean ± SD. Statistical differences were examined by ordinary 1-way ANOVA followed by Tukey’s multiple-comparison test. *P < 0.05; **P < 0.01; ***P < 0.001. N, control; KO, knockout.

Figure 5. Genetic inhibition of STING inhibits NF-κB signaling and attenuates diabetes-induced increases in inflammatory proteins, production of superoxide, and leukostasis in the mouse retina.

(A) Representative immunoblots and densitometry graphs demonstrating the diabetes-induced increases in the ratios of (B) p-IKK/total IKK, (C) p-IκB/IκB, and (D) p-NF-κB/total NF-κB. Levels of (E) iNOS and (F) ICAM-1 were inhibited in the retina of STING-KO and STING^GT diabetic mice. (G) Retinal superoxide was measured using lucigenin; STING-KO and STING^GT attenuate the retinal production of superoxide caused by diabetes. (H) Representative images and (I) quantification of attached leukocytes (arrows) in the retina blood vessels show that diabetes increased the number of adherent leukocytes in the WT diabetic retina compared with nondiabetic control mice. This number was markedly reduced in STING-KO and STING^GT diabetic mice. Scale bars: 100 μm. In A–I, n = 6 mice for each group, data are expressed as mean ± SD. Statistical differences were examined by ordinary 1-way ANOVA followed by Tukey’s multiple-comparison test. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 versus WT nondiabetic (N) controls. KO, knockout; D, diabetes.

Figure 6. Activation of STING leads to increased secretion of IFN-β, which contributes to cellular senescence in mouse retinal endothelial cells.

(A) The secretion of IFN-β into the medium is higher in mRECs after cGAMP treatment compared with controls. (B) FSC-A versus SSC-A dot plots were gated to eliminate debris, single cells were selected on FSC-A versus FSC-H, and positive senescent cells emit a fluorogenic signal that has absorption/emission maxima of 490/514 nm, which fall within the FITC detection region. (C) Quantification of senescent mRECs after IFN-β and cGAMP treatment for 1, 2, and 4 days. A significantly higher percentage of senescent cells was observed in the IFN-β– and cGAMP-treated groups after 2 and 4 days of incubation compared with nontreated control groups. (D) Representative images for immunocytochemical study. Scale bar: 50 μm. Green represents high β-galactosidase activity; blue, Hoechst. (E) Quantification demonstrating a higher percentage of senescent cells in the IFN-β– and cGAMP-treated groups after 4 days of treatment compared with nontreated controls. (F) Representative images for proliferation assay. Scale bar: 50 μm. Green, EdU; blue, Hoechst. (G) IFN-β– and cGAMP-treated cells exhibit lower rates of proliferation than the control, which was partially rescued by the STING inhibitor. Values are the mean of 4 replicates ± SD. *P < 0.05, **P < 0.01, ***P < 0.001, ****P < 0.0001 for the effects of IFN-β versus control groups. ‡‡‡P < 0.001, ‡‡‡‡P < 0.0001 for the effects of cGAMP versus nontreated control. Significance was examined by Mann-Whitney U test (A) or Kruskal-Wallis test followed by Tukey’s multiple-comparison test (C–G).

Figure 7. cGAMP and IFN-β induce cellular senescence and reduced cell proliferation in human retinal endothelial cells.

(A) Representative images for β-galactosidase assay. Original magnification, ×20; scale bar: 50 μm. (B) Quantification demonstrating a higher percentage of senescent cells in the IFN-β– and cGAMP-treated groups. (C) Representative images for EdU assay. Original magnification, ×10; scale bar: 50 μm. (D) IFN-β– and cGAMP-treated cells exhibit lower rates of proliferation than control. Green, β-galactosidase activity; blue, Hoechst; red, EdU. Percentage of EdU-positive senescent cells and percentage of EdU-positive proliferating cells are expressed as mean ± SD (n = 6). ***P < 0.001; ****P < 0.0001 by Kruskal-Wallis test followed by Tukey’s multiple-comparison test.

Figure 8. Blocking STING inhibits diabetes-induced increases in IFN-β in the retina through STING/TBK1/IRF3 signaling.

(A) Representative immunoblots and quantification of (B) STING, (C) p-TBK1, and (D) IRF3. Induction of diabetes in WT mice resulted in increased STING compared with appropriate controls; STING was not expressed in STING-KO and STING^GT mice. p-TBK1 and p-IRF3 were decreased in STING-KO and STING^GT diabetic mice compared with WT diabetic controls. ELISA analysis (E) shows that IFN-β was increased in WT diabetic retina compared with nondiabetic controls, but this was inhibited in STING-KO and STING^GT diabetic mice. n = 6 mice for each group; the data are expressed as mean ± SD. Statistical differences were examined by ordinary 1-way ANOVA followed by Tukey’s multiple-comparison test. **P < 0.01, ***P < 0.001, ****P < 0.0001 versus nondiabetic WT control. N, nondiabetic; D, diabetic.

Figure 9. Blocking STING inhibits diabetes-induced increases in cellular senescence markers in retinal blood vessels.

Diabetes increased retinal vascular mRNA levels for (A) p16, (B) p21, (C) Igfbp3, and (D) p53 compared with levels in nondiabetic mice. STING-KO and STING^GT significantly inhibited p16, p21, and Igfbp3, but not p53, in diabetic mice. (E) Representative images show a general development of a blue color (β-galactosidase activity, arrows) in the freshly isolated retinal blood vessel from WT diabetic mice but not from STING-KO and STING^GT mice. Scale bar: 50 μm. For A–D, n = 6 mice for each group; the data are expressed as mean ± SD. Statistical differences were examined by ordinary 1-way ANOVA followed by Tukey’s multiple-comparison test. *P < 0.05, **P < 0.01 versus nondiabetic WT control. N, nondiabetic; D, diabetic.

Figure 10. A schematic illustration of the role of STING in diabetic retinopathy (DR).

STING-mediated induction of the retinal type I IFNs secretion through TBK1/IRF3 and inflammatory molecules via the TBK1/NF-κB pathway in diabetes results in increased endothelial cell senescence and production of superoxide. These factors combine to cause retinal microvascular injuries in DR. Hyperglycemia evokes various other pathological mechanisms such as retinal neurodegeneration, accumulation of AGEs, and induction of PKC, the polyol, and the hexosamine pathways also critical in DR, independently of STING. STING may serve as a hub that stimulates multiple pathways (also promoted by these other factors) that collectively promote this multifaceted disease. All processes are interconnected to the development and progression of DR. AGEs, advanced glycation end products; PKC, protein kinase C; ROS, reactive oxygen species.