A subset of ATM- and ATR-dependent phosphorylation events requires the BRCA1 protein

Abstract

BRCA1 is a central component of the DNA damage response mechanism and defects in BRCA1 confer sensitivity to a broad range of DNA damaging agents. BRCA1 is required for homologous recombination and DNA damage-induced S and G(2)/M phase arrest. We show here that BRCA1 is required for ATM- and ATR-dependent phosphorylation of p53, c-Jun, Nbs1 and Chk2 following exposure to ionizing or ultraviolet radiation, respectively, and is also required for ATM phosphorylation of CtIP. In contrast, DNA damage-induced phosphorylation of the histone variant H2AX is independent of BRCA1. We also show that the presence of BRCA1 is dispensable for DNA damage-induced phosphorylation of Rad9, Hus1 and Rad17, and for the relocalization of Rad9 and Hus1. We propose that BRCA1 facilitates the ability of ATM and ATR to phosphorylate downstream substrates that directly influence cell cycle checkpoint arrest and apoptosis, but that BRCA1 is dispensable for the phosphorylation of DNA-associated ATM and ATR substrates.

Fig. 1. Adenovirus-infected HCC1937 cells express BRCA1 and are complemented for their radiosensitive phenotype. (A) Expression of BRCA1 protein in adenovirus-infected HCC1937 cells. Nuclear extracts from the indicated cell lines were subjected to immunoblotting using anti-BRCA1 antibody. The same extracts were also examined using anti-ATM antibody as a loading control and to verify ATM expression in BRCA1-defective cells. 1BRneo and AT5BIVA are transformed fibroblasts derived from a normal and an A-T patient, respectively. 293T is a tumour cell line. HCCAdco and HCCAdB1 are HCC1937 cells infected with empty adenovirus vector or adenovirus expressing full-length BRCA1 protein, respectively. (B) Clonogenic survival after exposure of the indicated cell lines to 2 Gy irradiation. The results shown were obtained from a single survival analysis carried out on the corrected cell lines. (C) Radiosensitivity monitored by the formation of micronuclei in the indicated cell lines. Either unirradiated cells (empty boxes) or cells irradiated with 6 Gy γ-rays (filled boxes) were scored for micronuclei. The results represent the mean of three experiments and the error bars represent the standard deviation of the mean between these experiments. Each experiment was carried out after a separate infection with adenovirus.

Fig. 2. Phosphorylation of p53, c-Jun>, Nbs1, CtIP and Chk2 is impaired in HCC1937 cells following exposure to IR. (A) Phosphorylation of the indicated substrates was examined in the absence of irradiation and at 0.5, 1 and 4 h after exposure to 20 Gy γ-rays. Phosphospecific antibodies against p53, c-Jun, Nbs1 and Chk2 were employed. Below these samples are the non-phosphospecific antibodies used as expression controls. CtIP was examined by immunoblotting. Phosphorylation was also examined in HCC1937 cells expressing BRCA1 following adenovirus infection (HCCAdB1) and HCC1937 cells infected with empty adenovirus (HCCAdco). Phosphorylation was compared with 293T cells which express similar levels of BRCA1 to HCCAdB1 cells. The majority of immunoblots have been analysed from three independently prepared cellular extracts. (B) Phosphorylation of p53 at 8 h post-irradiation. At this later time point, marked phosphorylation was observed in AT5BIVA but not in HCC1937 cells. Similar results were also obtained 12 h post-irradiation (data not shown).

Fig. 3. Phosphorylation of p53, c-Jun, Nbs1 and Chk2 is impaired in HCC1937 cells following exposure to UV. Phosphorylation of the indicated substrates was examined in the absence of irradiation and at 0.5, 1 and 4 h after exposure to UV (20 J/m²). Phosphorylation was also examined in HCC1937AdB1 and HCCAdco cells, and compared with that observed in 293T cells. Details of the antibodies used are given in Figure 2. CtIP is not phosphorylated after exposure to UV irradiation.

Fig. 4. HCC1937 cells activate ATM normally and phosphorylate H2AX, Rad17, Rad9 and Hus1 after exposure to IR and UV irradiation. (A) ATM was immunoprecipitated from the indicated cell lines without exposure to IR and 4 h after exposure to 20 Gy. α-ATM indicates the level of immunoprecipitated ATM in each sample. The immunoprecipitated material was examined for ATM kinase activity using a p53 peptide. Phosphorylation of p53 was determined using the anti-p53^Ser15 antibody. ATM kinase activity is activated normally independently of BRCA1 expression. (B) The phosphorylation of H2AX phosphospecific anti-p-H2AX^Ser139 antibody was examined 1 h after exposure of HCC1937 cells to IR (20 Gy) or UV (20 J/m²). Results are only shown for HCC1937 cells, but similar results were observed with 1BRneo and 293T cells. Results are shown at low-power magnification to show the response of multiple cells and also in individual cells. (C) The kinetics of p-H2AX foci formation at varying times after exposure of 1BR3, 293T and HCC1937 to IR (20 Gy) or UV (20 J/m²). (D) The mobility shift observed after immunoblotting of Rad9 and Hus1 was examined 4 h after exposure of 1BRneo cells to 20 J/m² with and without treatment with λ-phosphatase (λ-PPase). (E) The phosphorylation of Rad17, Rad9 and Hus1 was examined in 1BRneo, AT5BIVA and HCC1937 cells following exposure to IR (20 Gy) and UV (20 J/m²). Phosphorylation of Rad17 was examined using anti-p-Rad17 antibody and anti-Rad17 antibodies were used as a control for Rad17 expression. Phosphorylation of Rad9 and Hus1 was examined by mobility shift after immunoblotting. Phosphorylation was observed to a similar extent in 1BRneo, AT5BIVA and HCC1937 cells. (F) Rad17 was phosphorylated efficiently in HCCAdB1 and HCCAdco cells after UV and IR. Phosphorylation of Rad17 was examined in 293T cells and cells infected with empty adenovirus or adenovirus expressing BRCA1. (G) Formation of Rad9, Hus1 and Rad17 foci occurs normally in HCC1937 cells following exposure to UV. Thirty minutes after exposure to UV (20 J/m²), cells were examined by immunofluorescence for foci using anti-Rad9, anti-Hus1 and anti-Rad17 antibodies. Foci formed to a similar level in 1BRneo and HCC1937 cells. Additionally, the Hus1 and Rad17 foci were shown to overlap in merged images. Only a single cell has been shown to enhance visualization of the foci, but foci were observed in ∼40% of the cells 30 min after UV treatment (20 J/m²). A slightly higher level of foci was observed in untreated HCC1937 cells relative to untreated 1BRneo cells, which may reflect the elevated spontaneous instability reported in these cells (data not shown).

Fig. 4. HCC1937 cells activate ATM normally and phosphorylate H2AX, Rad17, Rad9 and Hus1 after exposure to IR and UV irradiation. (A) ATM was immunoprecipitated from the indicated cell lines without exposure to IR and 4 h after exposure to 20 Gy. α-ATM indicates the level of immunoprecipitated ATM in each sample. The immunoprecipitated material was examined for ATM kinase activity using a p53 peptide. Phosphorylation of p53 was determined using the anti-p53^Ser15 antibody. ATM kinase activity is activated normally independently of BRCA1 expression. (B) The phosphorylation of H2AX phosphospecific anti-p-H2AX^Ser139 antibody was examined 1 h after exposure of HCC1937 cells to IR (20 Gy) or UV (20 J/m²). Results are only shown for HCC1937 cells, but similar results were observed with 1BRneo and 293T cells. Results are shown at low-power magnification to show the response of multiple cells and also in individual cells. (C) The kinetics of p-H2AX foci formation at varying times after exposure of 1BR3, 293T and HCC1937 to IR (20 Gy) or UV (20 J/m²). (D) The mobility shift observed after immunoblotting of Rad9 and Hus1 was examined 4 h after exposure of 1BRneo cells to 20 J/m² with and without treatment with λ-phosphatase (λ-PPase). (E) The phosphorylation of Rad17, Rad9 and Hus1 was examined in 1BRneo, AT5BIVA and HCC1937 cells following exposure to IR (20 Gy) and UV (20 J/m²). Phosphorylation of Rad17 was examined using anti-p-Rad17 antibody and anti-Rad17 antibodies were used as a control for Rad17 expression. Phosphorylation of Rad9 and Hus1 was examined by mobility shift after immunoblotting. Phosphorylation was observed to a similar extent in 1BRneo, AT5BIVA and HCC1937 cells. (F) Rad17 was phosphorylated efficiently in HCCAdB1 and HCCAdco cells after UV and IR. Phosphorylation of Rad17 was examined in 293T cells and cells infected with empty adenovirus or adenovirus expressing BRCA1. (G) Formation of Rad9, Hus1 and Rad17 foci occurs normally in HCC1937 cells following exposure to UV. Thirty minutes after exposure to UV (20 J/m²), cells were examined by immunofluorescence for foci using anti-Rad9, anti-Hus1 and anti-Rad17 antibodies. Foci formed to a similar level in 1BRneo and HCC1937 cells. Additionally, the Hus1 and Rad17 foci were shown to overlap in merged images. Only a single cell has been shown to enhance visualization of the foci, but foci were observed in ∼40% of the cells 30 min after UV treatment (20 J/m²). A slightly higher level of foci was observed in untreated HCC1937 cells relative to untreated 1BRneo cells, which may reflect the elevated spontaneous instability reported in these cells (data not shown).

Fig. 5. Brca1^–/– ES cells are impaired in phosphorylation of c-Jun, p-53 and Chk2. (A) Wild-type and Brca1^–/– ES cells were transfected with a Cre-expressing plasmid. The result shows western blot analysis using a pool of cells selected in puromycin. (B) Cells from the same batch of cells used in (A) were examined for the expression of p-c-jun, p-p53, p-Chk2 and p-Rad17 by immunofluorescence. Filled symbols represent Brca1^+/+ positive cells and closed symbols are Brca1^–/– cells. The percentage of positive cells represents the percentage of cells with one or more foci. Whilst the phosphorylation of the first three substrates decreased in Brca1^–/– cells, Rad17 phosphorylation was seen to occur independently of Brca1 expression. (C) Cell extracts from the pool of cells used in (A) were also examined by immunoblotting using anti-p-c-jun^ser63. Cells were exposed to IR (20 Gy) or UV (20 J/m^–2) and were examined 1 and 4 h after irradiation. The Brca1^–/– cells showed no detectable c-Jun phosphorylation.

Fig. 6. Model of the role of BRCA1 in damage response signalling. ATM and ATR respond to DNA damage induced by IR or UV, respectively. ATR is associated with ATRIP at the damage sites (not shown for simplicity). Rad17 is recruited to the damage sites independently of ATR, and recruits Rad9, Rad1 and Hus1. BRCA1 is also recruited to the damage sites, in part by an interaction with Rad9 and Hus1 that occurs in undamaged cells. A Rad17–Rad1–Rad9–Hus1 complex also forms at the damage sites after IR treatment but the requirements for its assembly are not well understood. ATR phosphorylates H2AX, Rad17, Rad9 and Hus1 in the absence of BRCA1, possibly because these substrates are directly attached to the DNA. ATM or ATR also phosphorylates a range of additional substrates including Chk1, Chk2, p53, c-Jun, Nbs1 and CtIP, which are required to effect damage-response processes including checkpoint delay and apoptosis. Although the figure shows symmetrical pathways for ATM and ATR phosphorylation events, there are likely to be differences in their requirements. For ATR, the downstream phosphorylation events require BRCA1 and, most likely, the assembly of the Rad17–Rad9–Rad1–Hus1 complex. For ATM, phosphorylation of these substrates also requires the presence of BRCA1, but any requirement for the Rad17 complex is currently unknown. ATM may also require Nbs1 for phosphorylating these substrates (Girard et al., 2002). A requirement of BRCA1 for the phosphorylation of Smc1 and Chk1 by ATM has also recently been demonstrated (Kim et al., 2002; Yarden et al., 2002). We propose that BRCA1 acts as a scaffold to facilitate the downstream phosphorylation of ATM and ATR substrates. BRCA1-dependent phosphorylation events are shown as solid lines, and BRCA1-independent phosphorylation events are represented by broken lines.