ATR suppresses telomere fragility and recombination but is dispensable for elongation of short telomeres by telomerase

Abstract

Telomere shortening caused by incomplete DNA replication is balanced by telomerase-mediated telomere extension, with evidence indicating that the shortest telomeres are preferred substrates in primary cells. Critically short telomeres are detected by the cellular DNA damage response (DDR) system. In budding yeast, the important DDR kinase Tel1 (homologue of ATM [ataxia telangiectasia mutated]) is vital for telomerase recruitment to short telomeres, but mammalian ATM is dispensable for this function. We asked whether closely related ATR (ATM and Rad3 related) kinase, which is important for preventing replicative stress and chromosomal breakage at common fragile sites, might instead fulfill this role. The newly created ATR-deficient Seckel mouse strain was used to examine the function of ATR in telomerase recruitment and telomere function. Telomeres were recently found to resemble fragile sites, and we show in this study that ATR has an important role in the suppression of telomere fragility and recombination. We also find that wild-type ATR levels are important to protect short telomeres from chromosomal fusions but do not appear essential for telomerase recruitment to short telomeres in primary mouse embryonic fibroblasts from the ATR-deficient Seckel mouse model. These results reveal a previously unnoticed role for mammalian ATR in telomere protection and stability.

Figure 1.

Generation and characterization of Seckel Terc compound mice. (A) Outline of the ATR^+/S/Terc^+/− × ATR^+/S/G2 Terc^−/− cross. G3 progeny inherit one set of short telomeres (from the G2 Terc^−/− parent) and one set of long telomeres (from the Terc^+/− parent), but only G3 Terc^+/−* progeny will be telomerase proficient. (B) The Seckel mutation does not affect telomerase activity. The TRAP assay shows successful telomerase reconstitution in G3 Terc^+/−* MEFs compared with their G3 Terc^−/−* littermates. Protein concentrations used are indicated above each lane. As a negative control, extracts were treated with RNase (+) or left untreated (−). IC, internal control. (C) Reduced ATR expression in ATR^S/S cells. Western blot showing ATR expression levels in ATR^+/+ and ATR^S/S MEFs. Actin loading control is indicated below.

Figure 2.

ATR proficiency is dispensable for elongation of short telomeres. (A) Telomere length histograms for each genotype (top, i–iv); mean lengths of p arm, q arm, and total telomeres (bottom left); and proportion of SFEs (bottom right) from Y chromosome as determined by Q-FISH are shown. Statistical comparison of proportion of chromosomes <1 kb using Fisher's exact test is shown. (B) Representative image of combined Q-FISH and SKY. (C–E) Telomere length distribution histograms for each genotype (top, i–iv); mean lengths of p arm, q arm, and total telomeres (bottom left); and proportion of SFEs (bottom right) from chromosome 1 (C), chromosome 2 (D), and chromosome 11 (E), as determined by Q-FISH, are shown. Statistical comparison of proportion of chromosomes <1 kb (C) and <4 kb (D and E) using Fisher's exact test is shown. (A and C–E) Error bars represent SEM. n, number of metaphases analyzed. Bar, 1 µm.

Figure 3.

Analysis of overall telomere length and telomeric recombination. (A) Histogram of telomere lengths from all chromosomes in G3 ATR^+/+ and ATR^S/S MEFs. Arrows indicate critically short telomeres in G3 Terc^−/−* MEFs that are rescued in G3 Terc^+/−* MEFs. Data from three independent MEF cultures per genotype are shown. (B) Schematic of CO-FISH telomere labeling resulting from leading-strand (red) and lagging-strand (green) DNA synthesis. T-SCE results in combined green and red fluorescence. (C) Representative CO-FISH images of full metaphases hybridized with probes against the leading (red) and lagging (green) telomeres. Arrowheads depict chromosomes where signals are interchanged at both sister telomeres. (D) Quantification of T-SCE. Only T-SCE events with exchanges between both leading- and lagging-strand probes simultaneously were considered positive. The number of T-SCE events scored out of total number of chromosomes is indicated. Data represent at least two independent MEF cultures per genotype. Statistical comparison using Fisher's exact test is shown. Error bars represent SEM. Bar, 1 µm.

Figure 4.

Determination of global recombination frequencies (SCE). (A) Quantification of global SCE frequencies. The number of SCE events scored out of the total number of chromosomes is indicated. Data represent at least two independent MEF cultures (n) per genotype. Statistical comparison using Student's t test is shown. Error bars represent SEM. (B) Representative global SCE images of full metaphases. Arrowheads depict chromosomes where there are SCE events. Bars, 5 µm.

Figure 5.

Chromosomal aberrations in ATR/Terc MEFs. (A–C) Quantification of chromosome fusions (A), chromatid breaks and fragments (B), and MTSs (C) in G3 ATR^+/+ and ATR^S/S MEFs before (nontreated) or after treatment with aphidicolin. n, number of metaphases from two to five independent MEFs per indicated genotype. The number of events out of the total number of telomeres or chromosomes scored is indicated above the bars. Statistical comparisons using Fisher's exact test are shown. Error bars represent SEM. Photos show representative images of a metaphase and of individual aberrations. Arrowheads show the chromatid-type fusion (A) and MTSs (C). The boxed area in the larger micrograph in C shows examples of MTSs in detail below. (D) Increased aneuploidy in ATR^S/S MEFs. Histograms showing the number of chromosomes per metaphase for each genotype are shown. Aneuploidy is defined as metaphases having other than 40 chromosomes. Statistical comparisons using Fisher's exact test are shown. Bars, 1 µm.

Figure 6.

Telomerase deficiency aggravates lifespan attenuation in Seckel mutant mice. (A) Kaplan-Meier survival curve of mice of the indicated genotype. n, number of mice per genotype. Statistical comparisons using log-rank test are shown. (B) Reduced body weight of Seckel mice at 7 wk of age is not worsened by telomerase deficiency. Statistical comparisons using Student's t test (unpaired, two sided) are shown. n, number of mice. (C) Severe osteoporosis associated with ATR Seckel mutation is not increased by telomerase deficiency (two to three mice per genotype). (B and C) Error bars represent SEM.