DNA-damage-induced type I interferon promotes senescence and inhibits stem cell function

Abstract

Expression of type I interferons (IFNs) can be induced by DNA-damaging agents, but the mechanisms and significance of this regulation are not completely understood. We found that the transcription factor IRF3, activated in an ATM-IKKα/β-dependent manner, stimulates cell-autonomous IFN-β expression in response to double-stranded DNA breaks. Cells and tissues with accumulating DNA damage produce endogenous IFN-β and stimulate IFN signaling in vitro and in vivo. In turn, IFN acts to amplify DNA-damage responses, activate the p53 pathway, promote senescence, and inhibit stem cell function in response to telomere shortening. Inactivation of the IFN pathway abrogates the development of diverse progeric phenotypes and extends the lifespan of Terc knockout mice. These data identify DNA-damage-response-induced IFN signaling as a critical mechanism that links accumulating DNA damage with senescence and premature aging.

Figure 1

Induction of double strand breaks leads to production of functional IFNβ protein A. IFNβ protein was detected in FLAG-tagged TRF1-FokI (wild type or nuclease-inactive D450A mutant)-transfected mouse embryo fibroblasts. Immunofluorescence using indicated antibodies is shown. Magnification bar for all panels: 10 μm. B. Levels of TRF-FokI proteins detected by immunoblotting (upper panel) and quantification of percent of cells single or double positive (red bars) for FLAG and IFNβ proteins in 22-25 fields randomly chosen from 3 independent experiments performed as described in A (lower panel). Here and thereafter: data are shown as average± S.E.M.; * p<0.05; ** p<0.01; ***p<0.001. C. Expression of Irf7 promoter-driven IRF7-mCherry fusion protein in TRF1-FokI-transfected NIH3T3 cells treated as indicated with RNAi (control or against Ifnb) or antibodies (control IgG or neutralizing antibody against mouse IFNβ, 10 μg/ml for 72h after transfection). D. Immunofluorescent detection of IFNβ-positive cells after induction of DSB with 4-OHT (2.5μM) and Shield (1μM) compound for indicated times in the U2OS-DSB reporter cells (U2OSr) cells. E. Quantification of data shown in D in 9-11 fields randomly chosen from 3 independent experiments. F. Immunoblot analysis of lysates from the U2OSr cells transfected with indicated RNAi and treated or not with 4-OHT (2.5μM) and Shield (1μM) compound for 6h.

Figure 2

Effect of DSB induction on IRF3 localization A. Immunofluorescent detection of IFNβ-positive cells before (“off”) or after (“on”) induction of DSB with 4-OHT (2.5μM) and Shield (1μM) compound for 6h in the U2OSr cells that were prior either transfected with indicated siRNA oligos for 66h or pretreated with ATM inhibitor (Ku55933,10μM for 1h). Magnification bar for all panels: 10 μm. B. Recruitment of IRF3 to the DNA damage foci after inducing DSB with 4-OHT/Shield compounds for indicated times detected by immunofluorescence in the U2OSr cells. C. Quantification of cells containing IRF3-positive foci in the nucleus in 8-10 fields randomly chosen from 3 independent experiments performed as described in panels B and D. D. Recruitment of IRF3 to the DNA damage foci in the U2OSr cells pre-treated with vehicle or ATM inhibitor (Ku55933, 10μM for 1h) or DNA-PK inhibitor (Nu7441, 4μM for 1h) as indicated. E. IRF3 localization in U2OSr cells that received indicated siRNA oligos 66h before inducing DSB with 4-OHT (2.5μM) and Shield (1μM) compound as in panel D.

Figure 3

DSB induces expression of mRNA of IFNβ A. Relative expression of Ifnb mRNA in response to combined treatment with 4-OHT/Shield in parental U2OS cells and U2OSr cells. B. Relative expression of Ifnb mRNA normalized per GFP mRNA levels in NIH3T3 cells co-transfected with mouse genomic DNA (1μg, sonicated for indicated times) and GFP plasmid (100ng) per 12-well plate well. Average data from 3 experiments (each in triplicate) are shown. The right panel depicts a representative ethidium bromide-stained agarose gel demonstrating the length of DNA fragments used for transfection. M (Markers): GeneRuler 100bp Plus DNA Ladder. C. Relative levels of Ifnb mRNA in human fibroblasts (IMR90 levels taken as 1.0). Asterisks indicates significant differences between MRC5 fibroblasts cultured for 66 passages over those cultured for 37 passages (each passage being a two-fold split) and between fibroblasts from patients with Werner syndrome (WS, passage 23) or Hutchinson-Gilford Progeria Syndrome (HGPS, passage 23) and normal IMR90 (passage 25). D. Relative levels of Ifnb mRNA in fibroblasts from wild type (WT, taken as 1.0), late generation Terc^−/−or Werner-Bloom-Terc deficient animals (WBT). E. Relative levels of Ifnb and Ifna4 mRNA in gut, spleen and lymph nodes (LNs) tissues from late generation Terc^−/− mice compared to that in wild type mice (taken as 1.0). F. Relative levels of indicated IFN-stimulated genes mRNA in spleen tissues from late generation Terc^−/− mice compared to that in wild type mice (taken as 1.0).

Figure 4

Role of IFNβ in DNA-damage-induced senescence of mouse and human fibroblasts A. Cell cycle distribution of fibroblasts from patients with Hutchinson Gilford Progeria (HGPS) and Werner (WS) syndrome cultured in the presence of neutralizing antibody against human IFNβ or control antibody (IgG, both at 10μg/mL for 48 hours). B. Expression of SA-βGal by HGPS and WS human fibroblasts cultured in the presence of neutralizing antibody against human IFNβ or control antibody (IgG, both at 10μg/mL for 3-4 days). Magnification bar: 100 μm. Quantitated data reflect at least 200 total cells counted in 3 independent experiments. C. Expression of protein markers of senescence (mH2A, Hp1β, p21, p16, p53) in MRC5 fibroblasts (passage 37 or 66) and in WS fibroblasts was analyzed by immunoblotting using the indicated antibodies. IRF7 protein levels serve as an IFN-induced gene product control. Levels of β-actin were used as a loading control. D. Expression of SA-βGal by Terc^−/− and WBT fibroblasts cultured in the presence of neutralizing antibody against murine IFNβ or control antibody (IgG, both at 10μg/mL for 10 days). Magnification bar: 100 μm. Quantitated data reflect at least 200 total cells counted in 3 independent experiments. E. Expression of protein markers of senescence (Hp1β, p21, p16, p53 and p19^ARF) in wild type, Terc^−/− and WBT fibroblasts was analyzed by immunoblotting using the indicated antibodies. IRF7 protein levels serve as an IFN-induced gene product control. Levels of β-actin were used as a loading control.

Figure 5

IFN-dependent proliferative failure, senescence and apoptosis in the gut tissues from Terc-deficient mice is rescued by Ifnar1 ablation A. Hematoxylin and eosin (H&E) staining of the intestinal tissues from eight month old male mice of indicated genotypes. Magnification bar here and thereafter: 100μm. B. Immunofluorescent analysis of IRF7 expression in the intestinal tissues of 22-25 day old mice of indicated genotypes. Right panel depicts the number of IRF7-positive cells per field in the small intestines of 3 mice (at least 10 random fields observed for each). C. Analysis of SA-βgal positive cells in the intestinal tissues (counterstained with Nuclear Fast Red) of 22-25 day old mice of indicated genotypes. D. Quantitation of the number of SA-βGal positive cells per basal crypt in the small intestines of 3 mice (30-100 crypts were analyzed for each). E. Immunofluorescent analysis of Ki67 levels in the intestinal tissues of 22-25 day old mice of indicated genotypes. F. Quantitation of the number of Ki67 positive cells per basal crypt in the small intestines of 3 mice (30-100 crypts were analyzed for each). G. Immunofluorescent analysis of cleaved caspase-3 levels in the intestinal tissues of 22-25 day old mice of indicated genotypes. H. Quantitation of the number of cleaved caspase-3 -positive cells per field in the small intestines (at least 10 random fields observed for each of 3 mice).

Figure 6

IFN signaling amplifies DDR and ensuing expression of senescence mediators in the intestines from Terc-deficient mice A. Immunofluorescent analysis of telomere dysfunction-induced foci (arrows) using PNA telomere probe and anti-53BP1 antibodies. Two representative crypts are shown for each Ifnar1 background (top for SK, and bottom for DK). The number of PNA-associated, non-PNA-associated and total 53BP1 foci per intestinal crypt was scored in at least 50 crypts. Magnification bar: 20 μm. B. Immunofluorescent analysis of p53 expression in the intestinal tissues from 22-25 day old mice of indicated genotypes. Right panel depicts the percentage of p53-positive crypts in the small intestines (30-100 crypts were analyzed for each of 3 mice). Magnification bar: 50 μm. C. Immunofluorescent analysis of p16^INK4A expression in the intestinal tissues from 22-25 day old mice of indicated genotypes. Right panel depicts the number of p16^INK4A-positive cells per field (at least 10 random fields observed for each of 3 mice). Magnification bar: 100 μm. D. Relative mRNA levels of indicated genes in the indicated mouse intestines assessed by qPCR (levels in wild type mice taken as 1.0). Expression of Ifit3m mRNA was used as a control for a bona fide IFN-stimulated gene.

Figure 7

DDR-induced IFN signals play a key role in failure of stem and germ cells and development of phenotypes associated with premature aging in the Terc-deficient mice A. Number of hematopoietic precursor cells (LSK) in the bone marrow from 22-25 day old mice of indicated genotypes calculated from at least 4 individual animals. B. Number of hematopoietic stem cells (LSK CD150⁺CD48⁻) in the bone marrow from 22-25 day old mice of indicated genotypes calculated from at least 4 individual animals. C. Colony forming efficiency of bone marrow cells from age matched mice of indicated genotype. D. H&E staining of the testis tissues from 22-25 day old male mice of indicated genotypes. E. Number of litters per a mating pair (per first 4 months of mating) in mice of indicated genotypes (n=8 pairs for each genotype). F. Number of pups per litter in in mice of indicated genotypes (n=7 for litters per each genotype) G. Characteristic physical appearance and average body weight of 22-25 days old mice of indicated genotypes (n=10). H. Kaplan-Meier analysis of survival of G4 mice of indicated genotypes.