DNA damage response inhibition at dysfunctional telomeres by modulation of telomeric DNA damage response RNAs

Abstract

The DNA damage response (DDR) is a set of cellular events that follows the generation of DNA damage. Recently, site-specific small non-coding RNAs, also termed DNA damage response RNAs (DDRNAs), have been shown to play a role in DDR signalling and DNA repair. Dysfunctional telomeres activate DDR in ageing, cancer and an increasing number of identified pathological conditions. Here we show that, in mammals, telomere dysfunction induces the transcription of telomeric DDRNAs (tDDRNAs) and their longer precursors from both DNA strands. DDR activation and maintenance at telomeres depend on the biogenesis and functions of tDDRNAs. Their functional inhibition by sequence-specific antisense oligonucleotides allows the unprecedented telomere-specific DDR inactivation in cultured cells and in vivo in mouse tissues. In summary, these results demonstrate that tDDRNAs are induced at dysfunctional telomeres and are necessary for DDR activation and they validate the viability of locus-specific DDR inhibition by targeting DDRNAs.

Figure 1. Deprotection of telomeres leads to enhanced transcription of both telomere DNA strands.

(a) Total cell RNA was isolated from MEFs of the indicated genotype and treated with 4-hydroxytamoxifen (4OHT). Gel-extracted small RNA fraction (<40 nucleotides) was used for miScript PCR amplification to specifically detect DDRNAs. Error bars represent the s.e.m. n=3 independent experiments. *P<0.05, **P<0.01, Student’s t-test. (b) Small RNA (<200 nucleotides) fractions were isolated from 4OHT-treated MEFs of the indicated genotype, enriched for species with telomeric sequences using a telomeric bait, and sequenced. Histograms show for each reported read length the number of telomeric reads, either G rich or C rich, normalized on mir29b reads. For both teloG and teloC, MEFs Trf2^F/F have a significantly higher proportion of small RNA with respect to mir29b reads than MEFs Trf2^F/+. P<0.001, Fisher’s exact test. (c) Total cell RNA was isolated from MEFs of the indicated genotype and used for strand-specific RT–qPCR to detect telomeric dilncRNAs. Error bars represent the s.e.m. n=3 independent experiments. *P<0.05; ***P<0.001, Student’s t-test. (d,e) MEFs Trf2^F/F-expressing GFP-TRF1 were treated with vehicle or 4OHT and analysed 48 h later. (d) Representative images of teloG or teloC dilncRNA transcripts by smFISH. Zoomed in view of the boxed regions (12.5 × 12.5 μm²), with GFP-TRF1 and FISH signals, is shown on the right. The indicated numbers show the percentage of RNA signals co-localizing with GFP-TRF1±s.e.m. Scale bar, 5 μm (e) Quantification of data presented in d, showing the number of spots per cell and the relative intensity (a.u.=arbitrary units). Lines depict the mean±s.e.m. n=2 independent experiments; at least 50 cells per sample have been analysed; ***P<0.001.

Figure 2. DICER and DROSHA are involved in dilncRNA processing and DDRNA generation and necessary for full DDR activation at deprotected telomeres.

(a–b) Total cell RNA was isolated from MEFs of the indicated genotype that were previously treated with 4-hydroxytamoxifen (4OHT) and transfected with the indicated siRNA. (a) Gel-extracted small RNA fraction (<40 nucleotides) was used for miScript PCR amplification to specifically detect DDRNAs. Error bars represent the s.e.m. n=3 independent experiments. *P<0.05, Student’s t-test. (b) Total RNA was used for strand-specific RT–qPCR. Error bars represent the s.e.m. n=3 independent experiments. *P<0.05, Student’s t-test. (c) MEFs Trf2^F/F were treated with 4OHT, transfected with the indicated siRNA and stained for the indicated DDR markers. Scale bar, 10 μm. (d) Quantification of data presented in c. Bar graphs show the percentage of DDR-positive cells±95% confidence interval n=3 independent experiments; at least 150 cells per sample have been analysed; ***P<0.001, χ²-test.

Figure 3. DDR foci formation at deprotected telomeres is DDRNA dependent.

(a) MEFs Trf2^F/F were treated with 4-hydroxytamoxifen (4OHT), permeabilised, treated with RNase A, or BSA as control, and stained for the indicated DDR markers. Scale bar, 20 μm. (b) Quantification of data presented in a. Bar graphs show the percentage of DDR-positive cells±95% confidence interval (CI); n=3 independent experiments; at least 100 cells per sample have been analysed; ***P<0.001, χ²-test. (c–f) After RNase A treatment, cells were incubated with: (c) cell RNA from MEFs Trf2^F/F with normal (uninduced) or deprotected (induced) telomeres, or tRNA as a control; (d) cell RNA from MEFs Trf2^F/F treated with 4OHT and transfected with the indicated siRNA; (e) small double-stranded RNAs generated by recombinant DICER; (f) synthetic double-stranded RNAs. Bar graphs show the percentage of 53BP1-positive cells normalized on γH2AX-positive cells±95% CI n=3 independent experiments; at least 100 cells per sample have been analysed; *P<0.05, **P<0.01, ***P<0.001, χ²-test. (g) MEFs Trf2^F/F-expressing GFP-TRF1 were permeabilized and incubated with the indicated Alexa Fluor 647-conjugated double-stranded RNAs. The graph shows the mean number of dots per cell±s.e.m., detected by proximity ligation assay (PLA), using an anti-GFP antibody, recognizing the telomeric protein TRF1 and an anti-Cy5 antibody, recognizing the Alexa Fluor 647 antigen; n=2 independent experiments; at least 80 cells per sample have been analysed; ***P<0.001, Student’s t-test.

Figure 4. Antisense oligonucleotides against tDDRNAs inhibit DDR activation at deprotected telomeres.

(a) MEFs Trf2^F/F were treated, or not, with 4-hydroxytamoxifen (4OHT), transfected with the indicated ASO, and stained for the indicated DDR markers. Scale bar, 10 μm (b) Quantification of data presented in a. Dot plots show the number or intensity of DDR foci per cell (a.u.=arbitrary units). Lines depict the mean±s.e.m. n=3 independent experiments; at least 50 cells per sample have been analysed; **P<0.01, ***P<0.001, one-way analysis of variance with Sidak corrections for multiple comparisons.

Figure 5. Antisense oligonucleotides against tDDRNAs inhibit DDR activation in vivo in mice with deprotected telomeres.

(a–d) Mice were injected with vehicle or tamoxifen (tam) to induce telomere uncapping, and 24 h later, with the indicated ASO or PSB. (a) liver and (c) kidney sections were stained for the indicated DDR markers. Scale bars, 10 μm. (b,d) Quantification of the data presented in a and c. Dot plots show the intensity of DDR foci per cell in (b) liver and (d) kidney (a.u.=arbitrary units). Lines depict the mean±s.e.m. n=4 mice per group; at least 200 cells per sample have been analysed; ***P<0.001, one-way analysis of variance with Sidak corrections for multiple comparisons.