Spatial organization of the mammalian genome surveillance machinery in response to DNA strand breaks

Abstract

We show that DNA double-strand breaks (DSBs) induce complex subcompartmentalization of genome surveillance regulators. Chromatin marked by gamma-H2AX is occupied by ataxia telangiectasia-mutated (ATM) kinase, Mdc1, and 53BP1. In contrast, repair factors (Rad51, Rad52, BRCA2, and FANCD2), ATM and Rad-3-related (ATR) cascade (ATR, ATR interacting protein, and replication protein A), and the DNA clamp (Rad17 and -9) accumulate in subchromatin microcompartments delineated by single-stranded DNA (ssDNA). BRCA1 and the Mre11-Rad50-Nbs1 complex interact with both of these compartments. Importantly, some core DSB regulators do not form cytologically discernible foci. These are further subclassified to proteins that connect DSBs with the rest of the nucleus (Chk1 and -2), that assemble at unprocessed DSBs (DNA-PK/Ku70), and that exist on chromatin as preassembled complexes but become locally modified after DNA damage (Smc1/Smc3). Finally, checkpoint effectors such as p53 and Cdc25A do not accumulate at DSBs at all. We propose that subclassification of DSB regulators according to their residence sites provides a useful framework for understanding their involvement in diverse processes of genome surveillance.

Figure 1.

The local impact of laser microirradiation. (A) U2OS cells were either sensitized with BrdU followed by laser microirradiation or cultured without any presensitization followed by exposure to IR. 1 h later, the cells were fixed, immunostained with an antibody to the p32 subunit of RPA, and subject to z stack recording. (left) Representative 3D projections of cells exposed to the microlaser and 3 Gy of IR, respectively, are shown. (right) Quantification of the RPA foci in the microirradiated tracks or in the whole cell nuclei exposed to IR was obtained from 10 independent cells for each treatment. (B) U2OS cells were treated by the microlaser or exposed to increasing doses of IR. 1 h later, the cells were fixed and processed for RPA immunodetection as in A. A region spanning the entire microirradiated nuclear track containing the RPA foci (left) was placed over the maximum nuclear diameter of the IR-treated cells (right). The graph summarizes quantification of the RPA foci in these regions from 10 independent cells for each treatment. All images in this section are 3D projections as in A. (C) U2OS cells were microirradiated as in A. 1 h later, the cells were fixed and coimmunostained with antibodies to γ-H2AX and phospho-serine 15 of p53 (S15-P). The total nuclear fluorescence associated with S15-P was determined and compared with that measured in cells exposed for 1 h to the indicated doses of IR. The blue line marks the nucleus of an unirradiated cell to illustrate the background fluorescence associated with the S15-P antibody. The graph represents quantification of the S15-P fluorescence intensities from at least 50 cells for each treatment. Error bars indicate standard deviation. Bars, 10 μm.

Figure 2.

Protein interactions with the DSB-flanking chromatin. (A) Exponentially growing U2OS cells were sensitized with BrdU and microirradiated. 1 h later, the cells were fixed and stained with the indicated antibodies. Insets show higher magnifications of the microirradiated fields. (B) Accumulation of checkpoint mediators at the sites of DNA damage can occur throughout interphase. U2OS cells stably expressing GFP-tagged Mdc1 were treated as in A. After fixation, the cells were immunostained for cyclin A and the p32 subunit of RPA to indicate the cell cycle position (schematically illustrated in the right panel). Bars, 10 μm.

Figure 3.

Protein assemblies restricted to subchromatin microfoci. U2OS cells were microirradiated as in Fig. 2. Immunostaining with target specific antibodies (FANCD2, Rad51, RPA, and ATRIP) and a direct imaging of GFP-ATR revealed that accumulation of these proteins is restricted to nuclear subdomains that are distinct from the DSB-flanking chromatin compartments (the latter marked by γ-H2AX). Bars, 10 μm.

Figure 4.

Protein assembly in the ssDNA microcompartments is restricted to the S and G2 phases of the cell cycle. (A) U2OS cells were microirradiated as in Fig. 2. 1 h later, the ssDNA was revealed by immunodetection of BrdU without previous denaturation or nuclease treatment. Cells were coimmunostained with an antibody to γ-H2AX to demonstrate the location of the ssDNA microfoci within the larger regions of the DSB-modified chromatin and to show that generation of ssDNA occurred only in a subset of the microirradiated cells. (B) Cells were treated, and the ssDNA was detected as in A. Coimmunostaining of FANCD2 (shown here as an example) and other proteins from the spatial category (unpublished data) revealed close colocalization with ssDNA. (C) U2OS cells stably expressing GFP-Mdc1 were microirradiated and 1 h later subject to ssDNA detection as in A. In parallel, the cells were immunostained for cyclin B1 to indicate the cell cycle position. Cyclin B1 and ssDNA do not overlap and are therefore displayed in the same channel (red). (D) U2OS cells were treated as in A and coimmunostained with antibodies to Rad51, γ-H2AX (to detect the microirradiated tracks), and cyclin B1 (to reveal the cells in S/G2). The latter two proteins are simultaneously displayed in the same channel (red). Bars, 10 μm.

Figure 5.

Nbs1 interacts with both chromatin and the ssDNA subcompartments. (A) U2OS cells were microirradiated as in Fig. 2 and coimmunostained with the indicated antibodies. Under these standard conditions, Nbs1 occupies broad areas of γ-H2AX–decorated chromatin. (B) U2OS cells were transfected with control or H2AX-targeting siRNA oligonucleotides as indicated. 4 d later, the cells were microirradiated, incubated for 1 h, preextracted (see Materials and methods), and immunostained with an antibody to Nbs1. Note that in H2AX-depleted cells, Nbs1 assembles at the DSB sites in a form of subchromatin microfoci. (C) U2OS cells with stably down-regulated Mdc1 by short hairpin RNA were treated and immunostained for endogenous Nbs1 as in B. ssDNA compartments were detected by antibodies to RPA (top) or BrdU (bottom). Note that the fraction of Nbs1 that remains assembled at the DSB sites under these conditions is restricted to ssDNA (insets) and could be readily detected only in cells that are able to form these compartments (S/G2 phase). Arrows indicate the direction of the laser line during microirradiation. Bars, 10 μm.

Figure 6.

Spatial pattern of BRCA1 assembly at the DSB sites. (A) U2OS cells were transfected with control or Mdc1-targeting siRNA oligonucleotides for 4 d. The cells were then microirradiated and 1 h later fixed and coimmunostained with antibodies to γ-H2AX and BRCA1. (B) U2OS cells were treated with the siRNA oligonucleotides as in A. 1 h after microirradiation, the cells were fixed and immunostained with the indicated antibodies. Note that in the absence of Mdc1 (A and B, bottom), BRCA1 is lost from the DSB-flanking chromatin but remains assembled in the microfoci along the microirradiated tracks. The complete loss of 53BP1 from the DSB sites (B, bottom left) serves as a control of efficient Mdc1 down-regulation. (C) U2OS cells were microirradiated and 1 h later fixed and immunostained with antibodies to BRCA1 and cyclin A (the latter to reveal the cells in S/G2 phases). BRCA1 assembly could be detected also in the G1 cell (marked by the green arrow), although the overall abundance of BRCA1 in the nucleus and at in the DSB tracks is reduced compared with the S/G2 cells.

Figure 7.

DSB responses without cytologically discernible protein retention. (A) U2OS cells were microirradiated as in Fig. 2 and coimmunostained with antibodies to γ-H2AX, total Chk1, or Chk1 phosphorylated on serine 317 (S317-P), as indicated. Although the cells contained DSBs only within the nuclear tracks exposed to the laser (see the γ-H2AX pattern), the activated form of Chk1 was disseminated throughout the nucleus. The dotted yellow line marks the boundary between microirradiated and control cells. (B) U2OS cells were treated as in A and immunostained with antibodies to DNA-PK, Ku70, total Smc1, or Smc1 phosphorylated on serine 957 (S957-P), as indicated. Although the total Smc1 protein did not massively relocate to the DSB sites, it became locally phosphorylated within the microirradiated regions. Bars, 10 μm.

Figure 8.

Local accumulation of Ku70 and Smc1 in cells exposed to high doses of laser irradiation. (A) U2OS cells cultured without presensitization with halogenated thymidine analogues were locally irradiated with a high laser dose (73% energy output). Although such treatment induced local accumulation of Ku70 (left) and Smc1 (right) in the irradiated tracks, it was also accompanied by a pronounced destruction of the laser-exposed nuclear regions manifest by the decreased DNA staining (see the inset for magnification). (B) U2OS cells were either sensitized by BrdU and microirradiated with moderate laser dose (55% energy output; top) or cultured without presensitization and exposed to high laser dose (73% energy output; bottom). 1 h later, the cells were fixed and coimmunostained with antibodies to RPA and 53BP1 (three independent cells for each treatment are shown). Although exposure to the low laser energy was compatible with local DSB processing, formation of RPA foci and a robust assembly of 53BP1, the high laser dose generated extreme density of the RPA (without a clear resolution into individual repair foci) and impaired assembly of 53BP1 at the DSB-flanking chromatin. Bars, 10 μm.

Figure 9.

The major spatial patterns of DSB-induced protein redistribution visualized in living cells. (A) U2OS cell lines stably expressing the indicated DSB regulators tagged with GFP and/or the GFP spectral variants were sensitized with BrdU and microirradiated with moderate laser doses, as in Fig. 2. 1 h later, the microirradiated regions were retrieved and the microlaser-induced protein redistribution was recorded in living cells. Insets show higher magnification of the microirradiated fields; arrows indicate the laser direction through the respective nuclei. (B) The same set of cell lines as in A was exposed to 4 Gy of IR, and the living cells were recorded 1 h later. All localization patterns in A and B were maintained from the first signs of their appearance for up to several hours after microirradiation and IR exposure, respectively. Bars, 10 μm.