Neuroimmune signaling mediates astrocytic nucleocytoplasmic disruptions and stress granule formation associated with TDP-43 pathology

Abstract

Alterations in transactivating response region DNA-binding protein 43 (TDP-43) are prevalent in amyotrophic lateral sclerosis (ALS), frontotemporal dementia (FTD), and other neurological disorders. TDP-43 influences neuronal functions and might also affect glial cells. However, specific intracellular effects of TDP-43 alterations on glial cells and underlying mechanisms are not clear. We report that TDP-43 dysregulation in mouse and human cortical astrocytes causes nucleoporin mislocalization, nuclear envelope remodeling, and changes in nucleocytoplasmic protein transport. These effects are dependent on interleukin-1 (IL-1) receptor activity and nuclear factor kappa-light-chain-enhancer of activated B cells (NF-κB) signaling and are associated with the formation of cytoplasmic stress granules. Stimulation of IL-1 receptors and NF-κB signaling are necessary and sufficient to induce astrocytic stress granules and rapid nucleocytoplasmic changes, which are broadly alleviated by inhibition of the integrated stress response. These findings establish that TDP-43 alterations and neuroimmune factors can induce nucleocytoplasmic changes through NF-κB signaling, revealing mechanistic convergence of proteinopathy and neuroimmune pathways onto glial nucleocytoplasmic disruptions that may occur in diverse neurological conditions.

Keywords: Astrocytes; Integrated stress response; Neurodegeneration; Neuroimmune signaling; Nuclear pore; Stress granules; TDP-43.

Fig. 1.

TDP-43 dysregulation causes nucleocytoplasmic changes in astrocytes. (A, B) Immunolabeling (A) and quantification (B) of NUP-98 (green) in WT and ΔNLS primary mouse astrocytes. Data show mean extranuclear NUP-98 per well. Astrocyte marker GFAP (yellow), pan-TDP-43 (red), nuclear marker DAPI (blue). Unpaired t-test with Welch's correction: ***p < 0.0001; n = 18 wells per group; five independent experiments. (C, D) Immunolabeling of lamin B (green) in WT and ΔNLS primary mouse astrocytes (C) and quantification of nuclear circularity (D). Data show mean nuclear circularity. Unpaired t-test with Welch's correction: *p < 0.05; n = 4 wells per group; four independent experiments. (E, F) Immunolabeling (E) and quantification (F) of transfected reporter tdT-NES-NLS (red) in WT and ΔNLS primary mouse astrocytes. Data show extranuclear-to-nuclear ratios of mean signal. Unpaired t-test: ***p < 0.0001; n = 8 wells per group; six independent experiments. (G) Schematic representing differentiation of mature cortical astrocytes from human induced pluripotent stem cells (hiPSCs). (H, I) Immunolabeling (H) and quantification (I) of astrocyte marker GFAP (green). Data show the percentage of total cells expressing GFAP. Unpaired t-test with Welch's correction: n = 7 wells per group; two independent differentiations. (J) Quantification of NUP-98 in human isogenic control astrocytes (Con) and TDP-43 mutant (Q331K) iPSC-astrocytes. Unpaired t-test with Welch's correction: ***p < 0.0001; n = 11 wells per group; two independent differentiations. (K) Quantification of nuclear circularity in Con and Q331K iPSC-astrocytes. Unpaired t-test with Welch's correction: ***p < 0.001; n = 4 wells per group; two independent differentiations. (L) Quantification of transfected reporter tdT-NES-NLS in Con and Q331K iPSC-astrocytes. Unpaired t-test: **p < 0.01; n = 3 wells per group; two independent differentiations. Scale bars: 20 μm (A, C), 10 μm (E, H). Data are represented as mean ± SEM.

Fig. 2.

Astrocyte nucleocytoplasmic disruptions are mediated by IL-1 receptors and NF-κB. (A, B) Immunolabeling (A) and quantification (B) of NUP-98 (green) in WT and ΔNLS primary astrocytes treated with vehicle (Veh), IL-1 receptor antagonist (IL-1ra, 1 μg/ml), or NF-κB inhibitor (NFκB-i, TPCA-1, 500 nM). Astrocyte marker GFAP (yellow) indicates extranuclear regions. DAPI (blue) shows cell nuclei. One-way ANOVA: F(3,11) = 15.64, p = 0.0003; Dunnett's test: **p < 0.01, ***p < 0.001 vs. ΔNLS/Veh; n = 3–4 wells per group from three independent experiments. N: nucleus. (C, D) Immunolabeling for lamin B (C) and quantification of nuclear circularity (D). One-way ANOVA: F(3, 9) = 23.08, p = 0.0001; Dunnett's test: ***p < 0.001 vs. ΔNLS/Veh; n = 3–4 wells per group; three independent experiments. (E, F) Immunolabeling (E) and quantification (F) of transfected reporter tdT-NES-NLS (red). One-way ANOVA: F(3, 9) = 15.69, p = 0.0006; Dunnett's test: **p < 0.01, ***p < 0.001 vs. ΔNLS/Veh; n = 3–4 wells per group; three independent experiments. Scale bars: 20 μm (A, C), 10 μm (E). Data are represented as mean ± SEM.

Fig. 3.

Stimulation of astrocytic IL-1 receptors is sufficient to induce nucleocytoplasmic disruptions. (A, B) Immunolabeling (A) and quantification (B) of NUP-98 (green) in WT primary astrocytes treated with vehicle or IL-1α (3 ng/ml). Astrocyte marker GFAP (yellow) indicates extranuclear regions. DAPI (blue) shows cell nuclei. N: nucleus. Unpaired t-test with Welch's correction: **p < 0.01; n = 5–6 wells per group; four independent experiments. (C, D) Immunolabeling of lamin B (green) (C) and quantification of nuclear circularity (D). Unpaired t-test: *p < 0.05; n = 4 wells per group; four independent experiments. (E, F) Immunolabeling (E) and quantification (F) of transfected reporter tdT-NES-NLS (red). Unpaired t-test: ***p < 0.0001; n = 6 wells per group; three independent experiments. (G) Quantification of NUP-98 in human iPSC-astrocytes treated with vehicle or IL-1α. Unpaired t-test: **p < 0.01; n = 3–4 wells per group; two independent differentiations. (H) Quantification of nuclear circularity in human iPSC-astrocytes treated with vehicle or IL-1α and immunolabeled with lamin B. Unpaired t-test with Welch's correction: **p < 0.01; n = 4 wells per group; two independent differentiations. (I, J) Immunolabeling (I) and quantification (J) of transfected reporter tdT-NES-NLS (red) in human iPSC-astrocytes treated with vehicle or IL-1α. Unpaired t-test: **p < 0.01; n = 3 wells per group; two independent differentiations. Scale bars: 20 μm (A, C), 10 μm (E, I). Data are represented as mean ± SEM.

Fig. 4.

TDP-43 pathology and neuroinflammation induce astrocyte nucleocytoplasmic protein transport alterations in vivo. (A) Immunolabeling for tdT-NES-NLS (red) and astrocytic marker S100β (green) in neocortex of double transgenic ΔNLS and control NTG mice after retro-orbital injection with an AAV vector (PHP.eB-GfaABC1D-tdT-NES-NLS). DAPI shows cell nuclei (blue). (B, C) Immunolabeling (B) and quantification (C) of extranuclear-to-nuclear ratios of tdT-NES-NLS (black). One-way ANOVA: F(2,30) = 78.31, p < 0.0001; Dunnett's test: ***p < 0.0001 vs. NTG; n = 197–282 cells from 11 ROIs of two 8–10-month-old mice per genotype. (D) Immunolabeling for tdT-NES-NLS (red) and astrocytic marker S100β (green) in neocortex of transgenic Aldh1l1-Cre mice after retro-orbital injection with an AAV vector (PHP.eB-GfaABC1D-DIO-tdT-NES-NLS). DAPI shows cell nuclei (blue). Mice received intraperitoneal injection of vehicle (24 h) or LPS (5 mg/kg, 6 h or 24 h). (E, F) Immunolabeling (E) and quantification (F) of extranuclear-to-nuclear ratios of tdT-NES-NLS (black). One-way ANOVA: F(2, 69) = 16.73, p < 0.0001; Dunnett's test: ***p < 0.0001 vs. Veh; 165–203 cells from 24 ROIs of two 9–10-month-old mice per treatment group. Scale bars: 20 μm (A, D); 50 μm (top row) and 20 μm (bottom row) (B, E). Data are represented as median and quartiles.

Fig. 5.

Astrocyte nucleocytoplasmic changes are induced rapidly but resolve slowly. (A–C) Immunolabeling (A, B) and quantification (C) of NUP-98 (green) in WT primary astrocytes treated with vehicle or IL-1α (3 ng/ml) for indicated durations. Astrocyte marker GFAP (yellow) indicates extranuclear regions. DAPI shows cell nuclei (blue). N: nucleus. One-way ANOVA: F(6, 123) = 15.42, p < 0.0001; Dunnett's test: ***p < 0.001 vs. Veh; n = 18–19 wells per group; three independent experiments. (D—F) Immunolabeling of lamin B (green) (D, E) and quantification of nuclear circularity (F). One-way ANOVA: F(6, 16) = 9.95, p = 0.0001; Dunnett's test vs. Veh: **p < 0.01, ***p < 0.001 vs. Veh; n = 3 wells per group; three independent experiments. (G, H) Immunolabeling (G) and quantification (H) of transfected reporter tdT-NES-NLS (red). One-way ANOVA: F(6, 14) = 19.00, p < 0.0001; Dunnett's test: **p < 0.01, ***p < 0.001 vs. Veh; n = 3 wells per group; three independent experiments. (I, J) Immunolabeling (I) and quantification (J) of NUP-98 (green) in WT primary astrocytes treated with vehicle or IL-1α (3 ng/ml) for 24 h, then washed and maintained in media without IL-1α for indicated durations. One-way ANOVA: F(5, 55) = 13.15, p < 0.0001; Dunnett's test: **p < 0.01, ***p < 0.001 vs. Veh; n = 8–15 wells per group; two independent experiments. (K,L) Immunolabeling of lamin B (green) (K) and quantification of nuclear circularity (L). One-way ANOVA: F(5, 18) = 13.27, p < 0.0001; Dunnett's test: *p < 0.05, ***p < 0.001 vs. Veh; n = 3–6 wells per group; two independent experiments. (M, N) Immunolabeling (M) and quantification (N) of transfected reporter tdT-NES-NLS (red). One way-ANOVA: F(5, 15) = 17.67, p < 0.0001; Dunnett's test: **p < 0.01, ***p < 0.001 vs. Veh; n = 3–5 wells per group; two independent experiments. Scale bars: 20 μm, except 10 μm (G, M). Data are represented as mean ± SEM.

Fig. 6.

Nucleocytoplasmic changes are associated with stress granules and mediated by integrated stress responses. (A–C) Immunolabeling (A) for stress granules using colocalization of oligo(dT) (red) and HuR (green) and co-immunolabeling for NUP-98 (white) in WT primary astrocytes pre-incubated in vehicle or ISRIB (2 μM) for 4 h and treated with IL-1α (3 ng/ml) for 30 min. N: nucleus. (B) Quantification of stress granules/cell averaged per well. Two-way ANOVA: F(1, 19) = 6.67 for interaction effect, p = 0.0183; Bonferroni's test: **p < 0.01, ***p < 0.001 for indicated comparisons; n = 5–10 wells per group; two independent experiments. (C) Quantification of extranuclear NUP-98 levels. Two-way ANOVA: F(1, 36) = 55.61 for interaction effect, p < 0.0001; Bonferroni's test: ***p < 0.0001 for indicated comparisons; n = 10 wells per group; two independent experiments. (D, E) Immunolabeling of lamin B (green) (D) and quantification of nuclear circularity (E) in WT primary astrocytes pre-incubated in vehicle or ISRIB (2 μM) for 4 h and treated with vehicle or IL-1^α (3 ng/ml) for 2 h. Two-way ANOVA: F(1, 8) = 52.57 for interaction effect, p < 0.0001; Bonferroni's test: ***p < 0.001 for indicated comparisons; n = 3 wells per group; two independent experiments. (F, G) Immunolabeling (F) and quantification (G) of transfected reporter tdT-NES-NLS (red) in WT primary astrocytes pre-incubated in vehicle or ISRIB (2 μM) for 4 h and treated with vehicle or IL-1α (3 ng/ml) for 2 h. Two-way ANOVA: F(1,12) = 44.44, p < 0.0001; Bonferroni's test: ***p < 0.0001 for indicated comparisons; n = 4 wells per group; three independent experiments. (H–J) Immunolabeling (H) of stress granules using colocalization of oligo(dT) (red) and HuR (green) and co-immunolabeling for NUP-98 (white) in WT or ΔNLS primary astrocytes treated with vehicle or ISRIB (2 μM) for 24 h. (I) Quantification of stress granules. One-way ANOVA: F(2,23) = 40.70, p < 0.0001; Dunnett's test: ***p < 0.0001 vs. ΔNLS/Veh; 8–10 wells per group; two independent experiments. (J) Quantification of extranuclear NUP-98 levels. One-way ANOVA: F(2, 27) = 58.31, p < 0.0001; Dunnett's test: ***p < 0.0001 vs. ΔNLS/Veh; n = 10 wells per group; two independent experiments. (K, L) Immunolabeling (K) of lamin B (green) and quantification of nuclear circularity (L). One-way ANOVA: F(2, 7) = 37.97, p = 0.0002; Dunnett's test: ***p < 0.001 vs. ΔNLS/Veh; n = 3–4 wells per group; three independent experiments. (M, N) Immunolabeling (M) and quantification (N) of transfected reporter tdT-NES-NLS (red). One-way ANOVA: F(2, 7) = 17.80, p = 0.0018; Dunnett's test: **p < 0.01 vs. ΔNLS/Veh; n = 3–4 wells per group, three independent experiments. (O) ΔNLS primary astrocytes were treated with vehicle or NF-κB inhibitor (NFκB-i, TPCA-1, 500 nM, 72 h). Quantification of stress granules/cell averaged per well. Unpaired t-test: ***p < 0.0001; n = 9–10 wells per group. Scale bars: 10 μm, except 20 μm (D, K). Data are represented as mean ± SEM.