Effect of Ku80 deficiency on mutation frequencies and spectra at a LacZ reporter locus in mouse tissues and cells

Abstract

Non-homologous end joining (NHEJ) is thought to be an important mechanism for preventing the adverse effects of DNA double strand breaks (DSBs) and its absence has been associated with premature aging. To investigate the effect of inactivated NHEJ on spontaneous mutation frequencies and spectra in vivo and in cultured cells, we crossed a Ku80-deficient mouse with mice harboring a lacZ-plasmid-based mutation reporter. We analyzed various organs and tissues, as well as cultured embryonic fibroblasts, for mutations at the lacZ locus. When comparing mutant with wild-type mice, we observed a significantly higher number of genome rearrangements in liver and spleen and a significantly lower number of point mutations in liver and brain. The reduced point mutation frequency was not due to a decrease in small deletion mutations thought to be a hallmark of NHEJ, but could be a consequence of increased cellular responses to unrepaired DSBs. Indeed, we found a substantial increase in persistent 53BP1 and gammaH2AX DNA damage foci in Ku80-/- as compared to wild-type liver. Treatment of cultured Ku80-deficient or wild-type embryonic fibroblasts, either proliferating or quiescent, with hydrogen peroxide or bleomycin showed no differences in the number or type of induced genome rearrangements. However, after such treatment, Ku80-deficient cells did show an increased number of persistent DNA damage foci. These results indicate that Ku80-dependent repair of DNA damage is predominantly error-free with the effect of alternative more error-prone pathways creating genome rearrangements only detectable after extended periods of time, i.e., in young adult animals. The observed premature aging likely results from a combination of increased cellular senescence and an increased load of stable, genome rearrangements.

Figure 1. Spontaneous mutant frequencies in different organs from Ku80 ^−/− and wt mice.

Error bars indicate standard deviations (n = 5–8). Statistical tests were applied as described in the materials and methods to compare mutant frequencies between Ku80 ^−/− and wt mice.

Figure 2. Frequencies of point mutations (A) and genome rearrangements (B) in liver, spleen and brain from Ku80 ^−/− and wt mice, as determined from the mutants collected from the results shown in Fig. 1.

Figure 3. Increased apoptosis in liver of Sod ^−/− mice but not in Ku80 ^−/− liver tissue.

Representative images are shown from livers of Ku80 ^−/− and wt mice at 12 months and Sod1 ^−/− and wt mice at 4 months. Cells undergoing apoptosis were detected by in situ labeling of nuclear DNA fragmentation (TUNEL) as described in the text.

Figure 4. Ku80 ^−/− animals accumulate markers of DNA double-strand breaks in liver.

Livers from Ku80 ^−/− or wt animals were analyzed using immunohistochemistry on tissue arrays (γH2AX) or immunofluorescence on fresh frozen sections (53BP1). (A) Representative γH2AX staining in livers of 10-week and 12-month old wt and Ku80 ^−/− animals. (B) Representative 53BP1 staining in livers of 6-month old wt and Ku80 ^−/− animals. Frozen sections were deposited on glass slides, fixed and analyzed by immunofluorescence for the presence of 53BP1 DNA damage foci. The left panels show the nuclei stained in blue (DAPI) and 53BP1 DNA damage foci (red). The right panels display the 53BP1 staining only (Grayscale). Top panels show a section from a wt liver and the lower panels show a section from a Ku80 ^−/− liver. (C) The number of cells with at least one 53BP1 focus was quantified in 3 independent images from the livers of one representative pair of animals. The data are plotted as mean percentage of cells +/−S.D. from 3 independent measurements (n = total number of nuclei counted in all 3 measurements). Unpaired Student T-Test, p<0.0001 (n = 505 nuclei for wt and 381 nuclei for Ku80 mutants).

Figure 5. LacZ mutant frequencies in H₂O₂-treated Ku80 ^−/− and wt MEFs.

Frequencies of genome rearrangements (white) and point mutations (black) in (A) proliferating or (B) quiescent Ku80 ^−/− or wt MEFs. Error bars indicate standard deviations. No statistically significant differences were observed between wt and Ku80 ^−/− MEFs whether or not they were subjected to H₂O₂ exposure.

Figure 6. Persistent DNA damage foci in Ku80 ^−/− MEFs.

(A) Persistent DNA damage foci were visualized by immunofluorescence. Wt and Ku80 ^−/− MEFs were treated with bleomycin or H₂O₂ and stained 3 d later for the presence of 53BP1 foci. Top panels–53BP1 in red and nuclei counterstained with DAPI (blue). Bottom panels–53BP1 in grayscale. (B) Quantification of DNA damage foci (53BP1) shown in A, in untreated cells and 3 d after treatment of wt and Ku80 ^−/− MEFs.