(a) Left and top quadrants, immunoblots for pro-caspase-4 (pro-Casp4) and the p30 cleavage product of caspase-4 (Casp4 p30) in the RPE of human eyes with geographic atrophy (dry AMD) as compared to unaffected controls (Ctr). Specific bands of interest are indicated by arrowheads. Lower right quadrant, densitometry of the bands corresponding to caspase-4 p30 normalized to loading control. The molecular weight markers are indicated on the left side of the blot (Data are presented as mean ± SD; n = 3 control eyes; n = 6 dry AMD eyes; *P = 0.002, two-tailed t test). (b) Immunoblots for pro-Casp4 and Casp4 p30 in human RPE cells mock transfected (just transfection mixture) or transfected with Alu RNA; Alu expression plasmid (pAlu) or empty vector (pNull); or DICER1 or control (Ctr) anti-sense oligonucleotides (AS). Specific bands of interest are indicated by arrowheads. (c) Immunoblot for pro-caspase-11 (Pro-Casp11) and the p30 cleavage product of caspase-11 (Casp11 p30) in RPE tissue of WT mice injected subretinally with Alu RNA or vehicle (Vehi). n = 3 mice per group. (d,e) Top, fundus photographs of the retinas of WT (n = 8 eyes) and Casp11^−/− (n = 10 eyes) mice, (d) and Casp11^−/− (n = 8 eyes) mice expressing a human caspase-4 transgene (Casp11^−/− hCasp4^Tg) (e) injected with vehicle or Alu RNA. The degenerated retinal area is outlined by blue arrowheads. Bottom, immunostaining with zonula occludens-1 (ZO-1) antibody to visualize RPE cellular boundaries; loss of regular hexagonal cellular boundaries is indicative of degenerated RPE. (f) Immunoblots of pro-caspase-1 (pro-Casp1) and the p20 cleavage product of caspase-1 (Casp1 p20) in RPE tissue of WT and Casp11^−/− mice injected subretinally with vehicle (Vehi) or Alu RNA. n = 3 mice per group. (g) Immunoblots of pro-caspase-1 and the p20 cleavage product of caspase-1 in WT and Casp11^−/− mouse RPE cells treated with Alu RNA. (h) IL-18 secretion by WT and Casp11^−/− mouse RPE cells mock transfected or transfected with Alu RNA. n = 3 independent experiments. Data presented are mean ± SD; *P = 0.014, two-tailed t test. (i) Top, fundus photographs of the retinas of WT (n = 8 eyes), caspase-1 and caspase-11 deficient (n = 7 eyes) mice (Casp1^−/− Casp11^129mt/129mt) as well as Casp1^−/− Casp11^129mt/129mt (n = 8 eyes) mice expressing functional mouse caspase-11 from a bacterial artificial chromosome transgene (Casp1^−/− Casp11^129mt/129mt Casp11^Tg) subretinally injected with vehicle or Alu RNA. For all immunoblots, cropped gel image of bands of interest of representative immunoblots of three independent experiments and densitometric analysis (mean (SEM)) are shown. Tubulin or β-actin or Vinculin was as a loading control as indicated in each blot. In d, e and i, binary (Healthy %) and morphometric quantification (PM, polymegethism (mean (SEM))) of RPE degeneration are shown (Fisher’s exact test for binary; two-tailed t test for morphometry; *P < 0.05; **P < 0.01; ***P < 0.001). The degenerated retinal area is outlined by blue arrowheads in the fundus images. Loss of regular hexagonal cellular boundaries in ZO-1 stained flat mounts is indicative of degenerated RPE.

(a) Top, fundus photographs of eyes of WT (n = 6 eyes), and Gsdmd^−/− (n = 10 eyes) mice subretinally injected with vehicle or Alu RNA. Bottom, immunofluorescence staining of zonula occludens-1 (ZO-1) on RPE flat mounts of the same eyes showing RPE cell boundaries. (b) IL-18 secretion by WT and Gsdmd^−/− mouse RPE cells mock transfected or transfected with Alu RNA (n = 3 independent experiments; Data presented are mean ± SD, *P = 0.01, two-tailed t test) (c) Immunoblots of pro-caspase-1 (pro-Casp1) and the p10 cleavage product of caspase-1 (Casp1 p10) in WT and Gsdmd^−/− mouse RPE cells mock transfected or transfected with Alu RNA. (d) Immunoblots of pro-caspase-11 (pro-Casp11) and the p30 cleavage product of caspase-11 (Casp11 p30) in WT and Gsdmd^−/− mouse RPE cells mock transfected or transfected with Alu RNA. (e) Immunoblots of Gasdermin D and cleavage product of Gasdermin D (Gsdmd p30) in mock transfected or Alu RNA transfected human primary RPE cells, WT mouse primary RPE cells, and WT BMDM as well as in RPE tissue from WT mice subretinally injected with vehicle or Alu RNA. (f, g) Fundus photographs and immunofluorescence staining of zonula occludens-1 (ZO-1) on RPE flat mounts, (f) Gsdmd^−/− mice reconstituted via in vivo subretinal transfection of empty vector plasmid (pNull; n = 4 eyes), plasmids expressing wild type gasdermin D (pGSDMD-WT; n = 4 eyes) or mutant gasdermin D incapable of undergoing p30 cleavage (pGSDMD-D276A; n = 5 eyes) subretinally injected with Alu RNA. (g) Gsdmd^−/− mice subretinally administered vehicle control (Vehicle; n = 4 eyes), recombinant mature IL-18 (recIL-18; n = 4 eyes), mature IL-18 expression plasmid (pIL-18ss; n = 5 eyes) or empty vector control (pNull; n = 4 eyes) were subretinally injected with Alu RNA. (h) RT-qPCR examination of GSDMD mRNA abundance in the RPE tissue of human AMD eyes (n = 7 eyes) and in healthy age-matched control eyes (n = 6 eyes). *P = 0.045, two-tailed t test; error bars denote geometric means with 95% confidence intervals. (i) Immunolocalization of gasdermin D in the RPE of human geographic atrophy eyes and age-matched healthy controls. For all immunoblots, cropped gel image of bands of interest of representative immunoblots of three independent experiments and densitometric analysis (mean (SEM)) are shown. In a, f and g, binary (Healthy %) and morphometric (PM, polymegethism (mean (SEM)) quantification of RPE degeneration are shown (Fisher’s exact test for binary; two-tailed t test for morphometry; *P < 0.05; **P < 0.01; ***P < 0.001). Loss of regular hexagonal cellular boundaries in ZO-1 stained flat mounts is indicative of degenerated RPE. The degenerated retinal area is outlined by blue arrowheads in the fundus images.

(a) Top, fundus photographs of eyes of WT (n = 7 eyes) and Ifnar^−/− (n = 14 eyes) mice subretinally injected with Alu expression plasmid (pAlu) or empty vector (pNull). Bottom, immunofluorescence staining of zonula occludens-1 (ZO-1) on RPE flat mounts of the above eyes showing RPE cell boundaries. (b) Immunoblot of pro-caspase-11 (pro-Casp1) and the p30 cleavage product of caspase-11 (Casp11 p30) in WT and Ifnar^−/− mouse RPE cells mock transfected or transfected with Alu RNA. (c) Immunoblot of pro-caspase-4 in IFN-β-treated human RPE cells. (d) IFN-β secretion by human RPE cells transfected with Alu expression plasmid (pAlu) or empty vector (pNull). Data presented are mean ± SD; n = 3 independent experiments; *P = 0.0012, two-tailed t test. (e) Immunoblot of phosphorylated STAT2 (pSTAT2) and total STAT2 in human RPE cells transfected with Alu expression plasmid (pAlu) or empty vector (pNull) or DICER1 or control (Ctr) anti-sense oligonucleotides (AS). (f) Fundus photographs and immunofluorescence staining of zonula occludens-1 (ZO-1) on RPE flat mounts of WT mice subject to Alu RNA co-administration with IFN-β neutralizing antibody (n = 6 eyes) or Isotype IgG (n = 4 eyes). (g) Fundus photographs and immunofluorescence staining of zonula occludens-1 (ZO-1) on RPE flat mounts of WT (n = 6 eyes) or Irf3^−/− (n = 6 eyes) or Stat2^−/− (n = 7 eyes) mice subretinally injected with Alu expression plasmid (pAlu) or empty vector (pNull). (h) Immunolocalization of IFN-β in the RPE of human geographic atrophy eyes and age-matched unaffected controls. Representative image from control and Dry AMD eyes are presented, n = 4 eyes. (i) Abundance of IFN-β mRNA in the RPE of the human geographic atrophy eyes compared to age-matched healthy controls, (Data presented are mean ± SEM; n = 4 eyes; *P = 0.018, two-tailed t test). For all immunoblots, cropped gel image of bands of interest of representative immunoblots of three independent experiments and densitometric analysis (mean (SEM)) are shown. In a, f and g, binary (Healthy %) and morphometric (PM, polymegethism (mean (SEM)) quantification of RPE degeneration are shown (Fisher’s exact test for binary; two-tailed t test for morphometry; *P < 0.05; **P < 0.01; ***P < 0.001). Loss of regular hexagonal cellular boundaries in ZO-1 stained flat mounts is indicative of degenerated RPE. The degenerated retinal area is outlined by blue arrowheads in the fundus images.

(a) Relative abundance of Ifnb mRNA in WT and Mb21d1^−/− mouse RPE cells mock-transfected or transfected with Alu RNA. Data presented are mean ± SEM; n = 4 cell culture replicates; *P = 0.0001, two-tailed t test. (b) Immunoblots of pro-caspase-1 (pro-Casp1) and the p20 cleavage product of caspase-1 (Casp1 p20) in WT and Mb21d1^−/− mouse RPE cells transfected with Alu expression plasmid (pAlu) or empty vector control (pNull). (c) Immunoblots of pro-caspase-11 (pro-Casp1) and the p30 cleavage product of caspase-1 (Casp11 p30) in WT and Mb21d1^−/− mouse RPE cells transfected with Alu expression plasmid (pAlu) or empty vector control (pNull). (d) IL-18 secretion by WT and Mb21d1^−/− mouse RPE cells mock transfected or transfected with Alu RNA. Data presented are mean ± SD; n = 3 independent experiments; *P = 0.032, two-tailed t test. (e) Relative abundance of IFNB mRNA in control (sh Scramble) or cGAS shRNA knockdown human RPE cells transfected with or DICER1 or control (Ctr) anti-sense oligonucleotides (AS). Data presented are mean ± SEM; n = 3 cell culture replicates; *P = 0.0002, two-tailed t test. (f) Immunoblot of phosphorylated STAT2 (pSTAT2); pro-caspase-4 and casp4 p30; pro-caspase-1 and p20 cleavage casp1 p20 in control (sh Scramble) or cGAS shRNA knockdown human RPE cells mock-transfected or transfected with Alu RNA. Knockdown efficiency of cGAS is shown by cGAS immunoblot and tubulin was used as a loading control. (g) Fundus photographs and immunofluorescence staining of zonula occludens-1 (ZO-1) on RPE flat mounts of WT (n = 6 eyes) and Mb21d1^−/− (n = 8 eyes) mice subretinally injected with vehicle or Alu RNA. (h) Fundus photographs and immunofluorescence staining of zonula occludens-1 (ZO-1) on RPE flat mounts of Mb21d1^−/− (n = 7 eyes) mice reconstituted by in vivo transfection of cGAS expression plasmid (pFlag-cGAS) or control GFP expression plasmid (pFlag-GFP), subretinally injected with Alu RNA. (i) Fundus photographs and immunofluorescence staining of zonula occludens-1 (ZO-1) on RPE flat mounts of Mb21d1^−/− mice subretinally co-administered Alu RNA with recombinant IFN-β (n = 6 eyes), vehicle control (n = 6 eyes), IFN-β expression plasmid (pIFNB; n = 5 eyes) or empty vector control (pNull; n = 5 eyes). For all immunoblots, cropped gel image of bands of interest of representative immunoblots of three independent experiments and densitometric analysis (mean (SEM)) are shown. In g, h and i, binary (Healthy %) and morphometric (PM, polymegethism (mean (SEM)) quantification of RPE degeneration are shown (Fisher’s exact test for binary; two-tailed t test for morphometry; *P < 0.05; **P < 0.01; ***P < 0.001). Loss of regular hexagonal cellular boundaries in ZO-1 stained flat mounts is indicative of degenerated RPE. The degenerated retinal area is outlined by blue arrowheads in the fundus images.

(a) Immunolocalization of cGAS in the RPE of human geographic atrophy eyes and age-matched unaffected controls. Representative image from control and Dry AMD eyes are presented, n = 4 eyes. (b) Immunoblots of pro-caspase-1 and p20 cleavage product of caspase-1 (Casp1 p20) in WT and Tmem173^−/− mouse RPE cells transfected with Alu expression plasmid (pAlu) or empty vector control plasmid (pNull). (c) Immunoblots of pro-caspase-11 and p30 cleavage product of caspase-11 (Casp11 p30) in WT and Tmem173^−/− mouse RPE cells mock transfected or transfected with Alu RNA. (d) Fundus photographs and immunofluorescence staining of zonula occludens-1 (ZO-1) on RPE flat mounts of WT (n = 6 eyes) and Tmem173^−/− (n = 10 eyes) mice subretinally injected with vehicle or Alu RNA. (e) Fundus photographs and immunofluorescence staining of zonula occludens-1 (ZO-1) on RPE flat mounts of Tmem173^−/− mice subretinally co-administered Alu RNA with recombinant IFN-β (n = 4 eyes) or vehicle control (n = 4 eyes); or IFN-β expression plasmid (pIFNB; n = 4 eyes) or empty vector control (pNull; n = 4 eyes). For all immunoblots, cropped gel image of bands of interest of representative immunoblots of three independent experiments and densitometric analysis (mean (SEM)) are shown. In d and e, binary (Healthy %) and morphometric (PM, polymegethism (mean (SEM)) quantification of RPE degeneration are shown (Fisher’s exact test for binary; two-tailed t test for morphometry; *P < 0.05; **P < 0.01; ***P < 0.001). Loss of regular hexagonal cellular boundaries in ZO-1 stained flat mounts is indicative of degenerated RPE. The degenerated retinal area is outlined by blue arrowheads in the fundus images.

(a) Relative abundance of cytosolic mitochondrial DNA (mtDNA) in human RPE cells mock-transfected or transfected with Alu RNA (Data presented are mean ± SEM; n = 3 independent experiments; *P = 0.0018, two-tailed t test). (b) Relative enrichment of mtDNA in cGAS immunoprecipitate in ChIP like pull-down assay. Mock or Alu RNA transfected, indicated mouse embryonic fibroblast (MEF) were analyzed upon HA-cGAS immunoprecipitation with anti-HA antibody or isotype conrol. Data presented are mean ± SEM; n = 3; *P = 0.008, two-tailed t test. (c) Relative abundance of cytosolic mitochondrial DNA (mtDNA) in WT and Ppif^−/− mouse RPE cells mock transfected (n = 4 cell culture replicates) or transfected with Alu RNA (Data presented are mean ± SEM; n = 6 cell culture replicates; *P = 0.004, two-tailed t test). (d) Fundus photographs and immunofluorescence staining of zonula occludens-1 (ZO-1) on RPE flat mounts of WT (n = 6 eyes) and Ppif^−/− (n = 12 eyes) mice subretinally injected with vehicle or Alu RNA. (e) Immunoblot for procaspase-1 (procasp-1) and p20 cleavage product of caspase-1 in WT and Ppif^−/− mouse RPE cells mock transfected or transfected with Alu RNA. (f) Immunoblot for procaspase-11 (procasp-11) and p30 cleavage product of caspase-11 (Casp11 p30) in WT and Ppif^−/− mouse RPE cells mock transfected or transfected with Alu RNA. (g) Immunoblot for procaspase-4 (procasp-4) and p30 cleavage product of caspase-4 (Casp4 p30) in WT and mitochondrial DNA deficient Rho⁰ ARPE19 human RPE cells mock-transfected or transfected with Alu RNA. (h, i) WT and mitochondrial DNA deficient Rho⁰ ARPE19 human RPE cells mock-transfected or transfected with Alu RNA. (h) IL-18 secretion; data presented are mean ± SD; n = 4 independent experiments; *P = 0.0001, two-tailed t test. (i) IFN- β secretion; data presented are mean ± SD; n = 4 independent experiments; *P = 0.004, two-tailed t test. (j) Fundus photographs and immunofluorescence staining of zonula occludens-1 (ZO-1) on RPE flat mounts of the Ppif^−/− mice subretinally co-administered Alu RNA with recombinant IFN-β (n = 5 eyes) or vehicle control (n = 5 eyes); or IFN-β expression plasmid (pIFNB; n = 6 eyes) or empty vector control (pNull; n = 5 eyes). For all immunoblots, cropped gel image of bands of interest of representative immunoblots of three independent experiments and densitometric analysis (mean (SEM)) are shown. In d and j, binary (Healthy %) and morphometric (PM, polymegethism (mean (SEM)) quantification of RPE degeneration are shown (Fisher’s exact test for binary; two-tailed t test for morphometry; *P < 0.05; **P < 0.01; ***P < 0.001). Loss of regular hexagonal cellular boundaries in ZO-1 stained flat mounts is indicative of degenerated RPE. The degenerated retinal area is outlined by blue arrowheads in the fundus images.