Site-specific DICER and DROSHA RNA products control the DNA-damage response

Abstract

Non-coding RNAs (ncRNAs) are involved in an increasingly recognized number of cellular events. Some ncRNAs are processed by DICER and DROSHA RNases to give rise to small double-stranded RNAs involved in RNA interference (RNAi). The DNA-damage response (DDR) is a signalling pathway that originates from a DNA lesion and arrests cell proliferation3. So far, DICER and DROSHA RNA products have not been reported to control DDR activation. Here we show, in human, mouse and zebrafish, that DICER and DROSHA, but not downstream elements of the RNAi pathway, are necessary to activate the DDR upon exogenous DNA damage and oncogene-induced genotoxic stress, as studied by DDR foci formation and by checkpoint assays. DDR foci are sensitive to RNase A treatment, and DICER- and DROSHA-dependent RNA products are required to restore DDR foci in RNase-A-treated cells. Through RNA deep sequencing and the study of DDR activation at a single inducible DNA double-strand break, we demonstrate that DDR foci formation requires site-specific DICER- and DROSHA-dependent small RNAs, named DDRNAs, which act in a MRE11–RAD50–NBS1-complex-dependent manner (MRE11 also known as MRE11A; NBS1 also known as NBN). DDRNAs, either chemically synthesized or in vitro generated by DICER cleavage, are sufficient to restore the DDR in RNase-A-treated cells, also in the absence of other cellular RNAs. Our results describe an unanticipated direct role of a novel class of ncRNAs in the control of DDR activation at sites of DNA damage.

Figure 1. DICER or DROSHA inactivation impairs DDR foci formation in irradiated cells

a. DICER or DROSHA knock down WI38 cells were irradiated (10Gy) and fixed 7 h later. b. Histogram shows the percentage of cells positive for pATM, pS/TQ, MDC1 and γH2AX foci. c. WT and DICER^exon5 cells were irradiated (2 Gy) and fixed 2 h later. Histogram shows the percentage of cells positive for pATM, pS/TQ and MDC1 and γH2AX, foci. Error bars indicate s.e.m. (n ≥3). Differences are statistically significant (*p-value <0.01).

Figure 2. Irradiation-induced DDR foci are sensitive to RNase A treatment and are restored by small and DICER dependent RNAs

a. Irradiated HeLa cells (2 Gy) were treated with PBS (-) or RNase A (+) and probed for 53BP1, pATM, pS/TQ, MDC1 and γH2AX foci. Histogram shows the percentage of cells positive for DDR foci. b. 100, 50 or 20 ng of gel-extracted total RNA and 50 ng of RNA extracted from each gel fraction (>100, 35-100 and 20-35 nt) were used for DDR foci reconstitution after RNase treatment. c. 53BP1, pS/TQ and pATM foci are restored in RNase-treated cells when incubated with RNA of wild-type cells but not with RNA of DICER^exon5 cells or tRNA. e. Histogram shows the percentage of cells positive for DDR foci. Error bars indicate s.e.m. (n ≥3). Differences are statistically significant (*p-value <0.01).

Figure 3. Site-specific DDR focus formation is RNase A-sensitive and can be restored by site-specific RNA in a MRN-dependent manner

a. Cut NIH2/4 cells display a 53BP1 and γH2AX focus colocalizing with Cherry-Lac focus. 53BP1, but not γH2AX, focus is sensitive to RNase A and is restored by incubation with total RNA. b. Histogram shows the percentage of cells in which 53BP1 and Cherry-Lac foci co-localize. Addition of 50, 200, or 800 ng of RNA purified from cut NIH2/4 rescues 53BP1 foci formation in a dose-dependent manner. c. RNA purified from cut NIH2/4 restores 53BP1 focus while RNA from parental cells expressing I-SceI does not. d, e. RNase A-treated cut NIH2/4 cells were incubated with RNA from cut NIH2/4 cells, or parental ones, to test 53BP1 or pATM focus reformation in the presence of the MRN inhibitor mirin (100 μM). Histogram shows the percentage of cells positive for DDR focus. Error bars indicate s.e.m. (n ≥3). Differences are statistically significant (*p-value <0.05).

Figure 4. Chemically synthesized small RNAs and in vitro generated DICER RNA products are sufficient to restore DDR focus formation in RNase A-treated cells in a sequence-specific manner

a. Chemically synthesized oligonucleotides were annealed and were tested to restore DDR focus formation in RNase A-treated cut NIH2/4 cells. Mixed with a constant amount (800 ng) of parental cells RNA, a concentrations range (1 ng/μl – 1 fg/μl, ten-fold dilution steps) of locus-specific or GFP RNAs was used. Locus-specific synthetic RNAs (down to 100 fg/μl) allow site-specific DDR activation. b. Short double-stranded RNAs generated by recombinant DICER were tested to restore DDR focus formation in RNase A-treated cut NIH2/4 cells. 1 ng/μl RNAs were tested mixed with 800 ng of parental cells RNA. Locus-specific DICER RNAs, but not control RNAs, allow site-specific DDR activation. Histograms show the percentage of cells positive for DDR focus. Error bars indicate s.e.m. (n ≥3). Differences are statistically significant (*p-value <0.05).