NF-κB inhibition delays DNA damage-induced senescence and aging in mice

Abstract

The accumulation of cellular damage, including DNA damage, is thought to contribute to aging-related degenerative changes, but how damage drives aging is unknown. XFE progeroid syndrome is a disease of accelerated aging caused by a defect in DNA repair. NF-κB, a transcription factor activated by cellular damage and stress, has increased activity with aging and aging-related chronic diseases. To determine whether NF-κB drives aging in response to the accumulation of spontaneous, endogenous DNA damage, we measured the activation of NF-κB in WT and progeroid model mice. As both WT and progeroid mice aged, NF-κB was activated stochastically in a variety of cell types. Genetic depletion of one allele of the p65 subunit of NF-κB or treatment with a pharmacological inhibitor of the NF-κB-activating kinase, IKK, delayed the age-related symptoms and pathologies of progeroid mice. Additionally, inhibition of NF-κB reduced oxidative DNA damage and stress and delayed cellular senescence. These results indicate that the mechanism by which DNA damage drives aging is due in part to NF-κB activation. IKK/NF-κB inhibitors are sufficient to attenuate this damage and could provide clinical benefit for degenerative changes associated with accelerated aging disorders and normal aging.

Figure 1. NF-κB activation is increased in tissues of old WT and progeroid, DNA repair–deficient mice.

Kidney sections from NF-κB^EGFP mice were imaged using fluorescent microscopy to detect EGFP expression (green). Nuclei were counterstained with Hoechst dye (blue; original magnification, ×20). (A) Young adult (3-month-old) and old WT NF-κB^EGFP (2-year-old) mice. (B) Ercc1^–/ΔNF-κB^EGFP and WT NF-κB^EGFP mice at 3 months of age. (C) Ercc1^–/–NF-κB^EGFP and WT NF-κB^EGFP mice at 21 days of age. (D) Quantification of EGFP expression. The number of EGFP⁺ cells was counted in 5 random fields of tissue per mouse (n = 6 mice per group). The fold difference in the number of EGFP⁺ cells relative to the mean value (black bar) of the group is reported. Diamond symbols represent individual mice (controls in green and Ercc1^–/–NF-κB^EGFP mice in yellow). P values were calculated using a Student’s t test. (E) Ercc1^–/– and WT primary MEFs were passaged 5 times at 20% O₂ to promote the onset of senescence (58). The levels of p-p65, IκBα, and p-IκBα in nuclear and cytoplasmic extracts were measured by immunoblot. (F) NF-κB EMSA was performed with a radiolabeled oligonucleotide containing an NF-κB binding site using nuclear extracts from Ercc1^–/– and WT primary MEFs.

Figure 2. Genetic depletion of the p65 subunit of NF-κB delays aging symptoms and chronic diseases in progeroid Ercc1^–/Δ mice.

(A) EMSA on nuclear extracts from passage 5 WT, Ercc1^–/–, Ercc1^–/–p65^+/–, and Ercc1^–/–p65^–/– MEFs grown at 20% O₂ to measure NF-κB activity after depletion of p65. (B) Ercc1^–/Δ and Ercc1^–/Δp65^+/– mice were evaluated biweekly for the onset of spontaneous symptoms associated with aging. The aging score, which represents the fraction of aging symptoms delayed in a particular mouse compared with its sibling, for littermate pairs of Ercc1^–/Δ (red) and Ercc1^–/Δp65^+/– (orange) mice is a measure of healthspan (11). The mean aging score for each genotype is represented by a black bar. (C) Representative images of Ercc1^–/Δ and Ercc1^–/Δp65^+/– sex-matched littermates at 15 weeks of age. (D) Histopathologic changes in Ercc1^–/Δp65^+/– and Ercc1^–/Δ mice. Liver sections from 10-week-old mice were stained with oil red O to detect neutral lipids (hepatic steatosis; original magnification, ×100). Kidney specimens from 15-week-old mice were stained with H&E to detect proteinaceous renal tubular hyaline casts and glomerulosclerosis (original magnification, ×20). Cerebellar sections from 10-week-old mice were immunostained for GFAP (red), a marker of neurodegeneration. Nuclei were counterstained with DAPI (blue; original magnification, ×40). μCT of the vertebrae to assess bone porosity (for quantification, see Supplemental Figure 3A).

Figure 3. Pharmacologic inhibition of IKK/NF-κB activation delays aging symptoms and chronic diseases in progeroid Ercc1^–/Δ mice.

(A) Immunodetection of p-p65 in nuclear extracts of WT or Ercc1^–/– primary MEFs treated with 200 μM NBD or untreated (UT). Lamin A/C was used as a loading control. The histogram indicates the level of p-p65 normalized to that of untreated WT cells and corrected for loading. Values denote mean ± SD from 3 experiments. (B) Sibling, sex-matched pairs of Ercc1^–/Δ mice were treated with 10 mg/kg 8K-NBD or 8K-mNBD i.p., 3 times per week, beginning at 5 weeks of age. The aging score was calculated between Ercc1^–/Δ littermate pairs treated with 8K-mNBD (blue) or 8K-NBD (green). The mean aging score is represented by a black bar (P = 0.003, Student’s t test). (C) Representative images of Ercc1^–/Δ mice treated with 8K-NBD or 8K-mNBD peptide at 15 and 19 weeks of age. (D) Histopathologic changes analyzed in tissue sections from 18-week-old Ercc1^–/Δ mice treated with 8K-NBD or untreated. Liver sections were stained with oil red O to detect neutral lipids (hepatic steatosis; original magnification, ×100). Kidney specimens were stained with H&E to detect hyaline casts and glomerulosclerosis (original magnification, ×20). Cerebellar sections were immunostained for GFAP (red), a marker of neurodegeneration. Nuclei were counterstained with DAPI (blue; original magnification, ×40). μCT of vertebrae to measure bone porosity (for quantification, see Supplemental Figure 3B).

Figure 4. 8K-NBD inhibits NF-κB in vivo and corrects gene expression changes associated with aging.

RNA was isolated from livers of 18- to 19-week-old Ercc1^–/Δ mice chronically treated with 8K-NBD or 8K-mNBD (n = 4 per group). Differences in gene expression were analyzed using total genome Affymetrix arrays. (A) GO enrichment analysis of the networks and pathways regulated by 8K-NBD. The biological processes most significantly affected by chronic inhibition of IKK/NF-κB are ranked by their relative enrichment score. GNRH, gonadotropin-releasing hormone. (B) qRT-PCR of genes identified by microarray analysis to be significantly differentially expressed in Ercc1^–/Δ mice chronically treated with 8K-NBD relative to sibling mutant animals treated with 8K-mNBD. The dashed red line indicates mean expression in mice treated with the mutant peptide. The bars indicate the mean expression in mice treated with the NF-κB inhibitor 8K-NBD (n = 4 per treatment group) ± SD. For the NF-κB–regulated genes, green coloring indicates genes implicated in inflammation. Blue coloring indicates genes implicated in cell survival. Expression of Cdkn2a (the gene encoding p16^INK4a), a marker of cellular senescence, is in black.

Figure 5. Inhibition of NF-κB reduces cellular senescence in vitro and in vivo.

(A) Proliferation of WT (black; n = 4), Ercc1^–/– (red; n = 3), Ercc1^–/–p65^+/– (purple; n = 1) and Ercc1^–/–p65^–/– (orange; n = 3) congenic primary MEFs grown at 20% O₂ for several passages. Ercc1^–/– MEFs grew slower than WT MEFs, while Ercc1^–/–p65^–/– MEFs showed better growth compared with that of Ercc1^–/– MEFs. (B) γH2AX staining (red) of passage 5 WT, Ercc1^–/–, and Ercc1^–/–p65^–/– primary MEFs grown at 20% O₂. Nuclei were counterstained with DAPI (blue; original magnification, ×20). The histogram indicates the percentage of cells positive for γH2AX foci. (C) SA β-gal staining of liver sections from 10-week-old control, Ercc1^–/Δ, and Ercc1^–/Δp65^+/– mice (original magnification, ×40). The histogram indicates the percentage of SA β-gal–positive cells from 5 images from at least 9 mice per genotype. Values denote the mean ± SEM. *P < 0.05, Tukey-Kramer test. (D) Immunodetection of p16 in liver extracts of 19-week-old untreated Ercc1^–/Δ and NBD-treated Ercc1^–/Δ mice.

Figure 6. Inhibition of NF-κB reduces oxidative stress and damage in vitro and in vivo.

(A) Ercc1^–/– and Ercc1^–/–p65^–/– passage 6 primary MEFs grown at 20% O₂ were stained with DiOC₆ (green) to mark mitochondria and MitoSOX (red) to detect mitochondrial superoxide anion (original magnification, ×40). (B) Liver sections from 10-week-old Ercc1^–/Δ and Ercc1^–/Δp65^+/– mice imaged for lipofuscin fluorescence (original magnification, ×20). The histogram indicates the total fluorescent area for 5 images from 3 different mice per genotype calculated using MetaMorph software. (C) The levels of the (5′R) and (5′S) diastereomers of cdG and cdA in nuclear DNA isolated from the livers of 10-week-old control, p65^+/–, Ercc1^–/Δ, and Ercc1^–/Δp65^+/– mice. (D) The levels of cdG and cdA in nuclear DNA isolated from the livers of 19-week-old control, untreated Ercc1^–/Δ, and 8K-NBD–treated Ercc1^–/Δ mice. *P < 0.05, Tukey-Kramer test. Values denote the mean ± SEM (n = 3 per group).