DNA damage–induced ATR is inhibited by estradiol (E₂). (a) Time course of IR and UV activation of ATR, inhibited by 10 nM E₂. Random cycling MCF7 cells were briefly exposed to DNA-damaging agents in the presence or absence of 10 nM E₂. Endogenous Chk1 phosphorylation at Serine 345 was determined at various times after UV or IR exposure using a phospho-specific antibody for IB, after IP of total Chk1 protein and separation by SDS-PAGE. Total Chk1 serves as loading controls and was determined similarly by IB, reflecting an additional gel separation of the same cell lysates. IP with nonspecific, IgG antibody produced no kinase activity against Phas-1, the peptide used as the substrate for in vitro ATR activity assay described in Materials and Methods (Figure 1d). (b) Physiological E₂ inhibits HU-induced ATR activity via ER in S-phase cells. Phas-1 peptide is the substrate for in vitro kinase activity, using ATR IP from MCF7 cells exposed for 2 h to 1 mM HU. Total ATR protein is the loading control. Endogenous p53 phosphorylation at Serine15 using phospho-specific antibody was also determined by IB, after total p53 IP. ICI182780 (ICI, 1 μM) is an ER antagonist. The bar graph represents three Phas-1 experiments combined, single determinations per condition in each experiment. * p < 0.05 by ANOVA plus Schefe's test for control vs HU, ⁺ p < 0.05 for HU vs HU+E₂, ⁺⁺ p < 0.05 for HU+10 nM E₂ vs 10 nM E₂+ICI. (c) Left, ATR in vitro assay in random cycling MCF7 cells using two different antibodies to the C- or N-terminus of ATR, respectively. Right, immunoblots of ATR or ATM, after IP of protein from MCF7 cells with ATR or ATM antibodies to show specificity of the ATR ab-1 used. (d) IgG antibody was used for IP produces and produced no kinase activation or protein–protein interactions under any conditions. This was shown by subsequent immunoblot or in vitro kinase activity assay. Left, IP with IgG, followed by IB for IgG (positive control), IB for phospho-Chk1, TopBP1, claspin, or after IP with the IgG antibody. Similarly, no ATR kinase activity (Phas-1 kinase) resulted from IP with IgG. Right, positive detection with appropriate antibodies for all studies and modulation by UV. (e) S-phase MCF7 cells were briefly exposed to IR±E₂, and kinase activity was determined just before (time 0) and 15 min after radiation.

E₂ modulates TopBP1 protein association with ATR. (a) siRNA knockdown of the TopBP1 proteins in MCF7 cells (top lanes). Specific immunoblots of TopBP1 protein was determined after IP, 48 h after MCF7 cells were transfected with siRNA. IP with IgG antibody yielded no proteins on subsequent specific immunoblot (Figure 1d). TopBP1 knockdown resulted in the loss of basal or IR-induced ATR activity, determined in S-phase–synchronized cells (middle lanes). Kinase assay was carried out as described in Materials and Methods. Chk1 protein serves as loading/protein normalization control (bottom lane), and the study was repeated. (b) ATR:TopBP1 associations are inhibited by TopBP1 siRNA after UV exposure. Top: WB reveals less protein–protein interaction (lane 3 vs lane 7) upon TopBP1 knockdown or E₂ addition (lanes 4) in the setting of UV exposure. ATR immunoblots serve as loading/normalization controls. Bottom, experiments were done in random cycling MCF7 cells, as well as S-phase cells. Bar graph is from three experiments. * p < 0.05 for either control vs UV+control siRNA, ⁺ p < 0.05 for UV+control siRNA vs UV+control siRNA+E₂ or UV+siRNATopBP1. (c) E₂ inhibits ATR:TopBP1 associations. MCF7 cells were exposed to brief UV or IR±E₂, and the respective cell lysates were obtained at 5 and 15 min after radiation treatment. Lysates were immunoprecipitated with antibody for ATR, and precipitates were subjected to gel electrophoresis then IB with antibodies for TopBP1. In some conditions, the PI3K inhibitor, LY294002, was added. Total ATR IB was the loading control, and IgG antibody used for IP showed no associations of ATR with TopBP1 (Figure 1d). (d) Time course of TopBP1:ATR associations after DNA damage. S-phase MCF7cells were exposed to brief UV, and IP of ATR was followed by IB for TopBP1 at various time points after UV. IP with IgG antibody resulted in no bands (Figure 1d). The study was repeated. (e) Top, E₂ rapidly stimulates activation of AKT in MCF7 cells exposed to UV, determined by IB with specific antibody to phospho-Serine 473. Total AKT IB serves as normalization control. Bottom, UV- or HU-induced ATR:TopBP1 associations are inhibited by E₂, the steroid effect being dependent on AKT. Cells were transfected with control or specific siRNA to AKT or a dominant-negative AKT plasmid and then recovered and synchronized in S-phase. The cells were then exposed to radiation, without or with E₂, and specific IP:IB ensued from cell lysates obtained 5 min after UV. Bar graph is from three experiments. * p < 0.05 for control vs E₂. (f) E₂ inhibits ATR activation through PI3K. S-phase cells were exposed to UV±E₂, sometimes in the presence of LY294002 (PI3K inhibitor). ATR activity as Chk1 phosphorylation was determined in this representative study of two experiments. Bar graph is from three experiments. * p < 0.05 for control vs UV or UV+E₂+LY, ⁺ p < 0.05 for UV vs UV+E₂.

Chk1 activity is inhibited by E₂. (a) S-phase MCF7 cells were exposed for 2 h to HU±E₂ and HU+E₂+ICI at which time the cells were lysed. Endogenous Chk1 protein was immunoprecipitated from cell lysates for in vitro kinase activity assays using exogenous GST-Cdc25C as substrate. Total Chk1 protein serves as the loading control for each condition. Bar graph is from three experiments. * p < 0.05 for control vs HU, ⁺ p < 0.05 for HU vs HU+E₂, ⁺⁺ p < 0.05 for HU+E₂ 10 nM vs HU+E₂ 10 nM+ICI. (b) Cdc25C is retained in cytoplasm after phosphorylation induced by UV and inhibited by E₂. MCF7 were cultured on glass-bottom dishes, synchronized with 16-h exposure to nocodazole, and then released from nocodazole for 6 h. The cells were then exposed to brief UV or IR±E₂, and immunostaining with primary antibodies to total (nonphosphorylated) Cdc25C protein, phospho-Serine 216 Cdc25C, and phospho-Serine 10 of Histone H3 ensued. The second antibody was FITC-conjugated, goat-anti mouse IgG. Fluorescent microscopy was then performed. E₂ alone–treated cells show Cdc25C localized mainly in the nucleus. (c) Chk1:Claspin association is inhibited by E₂. Cell lysates underwent IP with a Claspin antibody, followed by IB of the precipitates using a Chk1 antibody. Lysates were from MCF7 cells exposed to IR or UV±E₂ for 5 or 15 min, respectively. Some cells were also exposed to LY294002. The study was repeated, and reverse order for IP:IB was also done (data not shown). Total Claspin protein immunoblots serve as loading controls, whereas IP with nonspecific IgG antibody showed no association of Chk1 and Claspin (Figure 1d). IP with Chk-1 antibody followed by IB with Chk-1 antibody shows the Chk-1 protein, whereas IP with a Chk-2 antibody did not show IB with Chk-1 ab. Specificity of Chk-2 antibody is also shown for comparison. (d) Chk1 Serine345 phosphorylation is inhibited by E₂ in PI3K/AKT-dependent manner. Some MCF7 cells were transfected to express DN-AKT or siRNA to AKT, recovered, and investigated during random cycling. Most cells were exposed to UV or HU±E2 in the absence or presence of LY294002. The cell lysates were used for phospho-Chk1 (S345) immunoblot, and the study was repeated. (e) Chk1 Serine280 phosphorylation is stimulated by E₂ in PI3K/AKT-dependent manner. Experiments were carried out as described in panel d, and IB was performed with phospho-Serine 280 Chk1 antibody. (f) S280AChk1 mutation prevents E₂-induced Claspin:Chk1 protein dissociation upon DNA damage. MCF7 cells were transiently transfected to express Flag-wt or S280AChk1 and then subjected to UV in the absence or presence of E₂ and LY. Chk1:Claspin association was determined by IB of claspin after IP with Flag antibody. Western blot of total Flag-tagged wt and S280AChk1 proteins are shown. Bar graph is from three experiments; * p < 0.05 for control vs UV (left) or UV+E₂, UV+E₂+LY (right), ⁺ p < 0.05 for UV vs UV+E₂ (left), ⁺⁺ p < 0.05 for UV+E₂ vs UV+E₂+LY. (g) Cdc2 (Cdk1) inhibitory phosphorylation is blocked by E₂. MCF7 cells were synchronized with nocodazole for 16 h, released for 6 h, and then exposed to UV or HU±E₂. IB of cell lysates with specific antibody determined Tyr15 phosphorylation of Cdc2. Total Cdc2 protein was immunoblotted as control. (h) E₂ stimulates Cdc2 kinase activity in the setting of DNA damage. Cdc2 was immunoprecipitated after brief UV±E2 exposure from nocodazole-synchronized and released MCF7 cells. Cells were also exposed to E₂ in the absence of UV. In vitro kinase activity was determined using p70S6 kinase protein as exogenous substrate.

Outcomes of DNA damage signaling. (a) Serine 15 phosphorylation of p53 is required for up-regulated p21 gene expression, inhibited by E₂. HCC1937 cells were transfected to express wt or S15Ap53, and cells were recovered and synchronized in S-phase and then exposed to brief UV±E2. p21 was determined by RT-PCR 6 h later, and GAPDH expression served as control. (b) γH2AX complex formation in response to UV and E₂. Left, MCF7 cells were assessed at various times after UV±E2 as shown by immunofluorescent microscopy. Right, immunoblot for γH2AX. Assay is described in Materials and Methods. The bar graph is the mean ± SEM of two immunoblot experiments combined. (c) Rad 51 complex formation is delayed by E₂. Immunoblots of Rad51 protein in the nucleus from MCF7 cells exposed to UV or UV±E₂ in a representative experiment is shown. The bar graph is the mean ± SEM of immunoblot density from three experiments. * p < 0.05 for control vs UV at 2 and 4 h; ⁺ p < 0.05 for UV vs UV+E₂ at 2 and 8 h. (d) Single-cell DNA repair. Experiments were carried out as described in Materials and Methods. The trailing tail of the DNA reflects unrepaired DNA and was quantified over 8 h in two experiments using Komet imaging software, version 5.5 (Andor Tech). Data are mean ± SEM from 200 cells assessed per condition, in each of two experiments.

Schematic of estrogen and membrane ERα-inhibiting ATR pathway signaling after DNA damage. E₂ acting at membrane-localized ERα rapidly activates PI3K/AKT to phosphorylate and dissociate TopBP1 from ATR in the setting of DNA damage. This leads to inhibition of ATR activity and downstream signaling to the G2/M checkpoint and p53-induced p21 transcription. Signaling through PI3K and AKT also inhibits Chk1 activity by phosphorylating claspin.

E₂ inhibits ATR activity through plasma membrane ERα. (a) E₂, PPT (ERα-specific agonist), but not DPN (ERβ-specific agonist) inhibited endogenous p53 and Chk1 phosphorylation at ATR sites as stimulated by IR and HU. ER agonists in the absence of DNA damage had no effect (data not shown). (b) ER null HCC-1569 cells were transiently transfected to express the E domain of ERα targeted to the cell membrane or nucleus. After overnight cell recovery, ATR activity was determined in random cycling cells 15 or 5 min after IR or UV. Bar graph is from three experiments, * p < 0.05 for control vs IR or IR+E₂ (nuclear), ⁺ p < 0.05 for IR vs IR+E₂. (c) S522AERα expression blocks E₂ inhibition of ATR. The mutant ER or pcDNA3 plasmids were transfected into MCF7, the cells recovered overnight, and S-phase cells were subjected to UV. Total Chk1 protein serves as loading normalization. * p < 0.05 for control vs various conditions, ⁺ p < 0.05 for UV vs UV+E₂ or pcDNA3+UV vs pcDNA3+UV+E₂. (d) Random cycling MCF7, ZR-75-1, and T47D breast cancer cells were exposed to E₂, progesterone, or testosterone, and ATR activity after UV or IR was determined. E₂ had no effect on HCC-1569 cells.

Mechanisms of E₂ inhibition of ATR activity. (a) Mutant TopBP1 does not dissociate from ATR. Flag-tagged wtTopBP1 or S1159ATopBP1 was expressed in MCF7 cells. After recover, random cycling cells were used for ATR:TopBP1 association determined by IP:IB under UV±E2 and LY conditions. Western blot of total Flag wt and S1159A (mutant) TopBP1 proteins are shown. Experiments were also repeated in S-phase cells. (b) E₂ stimulates the phosphorylation of TopBP1 at Serine 1159 through PI3K/AKT signaling. MCF7 cells were transfected to express Flag-wt or S1159ATopBP1, and the cells recovered, labeled with ³²P, and subjected to E₂, UV, LY, E₂+UV, or E₂+LY. IP using Flag antibody followed by PAGE, and autoradiography ensued. In parallel studies, total Flag-wt or S1159ATopBP1 protein was IB as expression/loading controls. Results are representative of two experiments. (c) S1159DTopBP1 does not associate with ATR after DNA damage. Flag-wt TopBP1 or AKT site, phospho-mimetic mutant S1159DTopBP1 were expressed in MCF7 cells, and the cells were recovered and then exposed to UV±E₂. IP with Flag antibody was followed by IB with ATR antibody. No association of wtTopBP1 and ATR was seen when nonspecific IgG was substituted (data not shown). (d) ER requires EGFR to inhibit ATR activation. MCF7 were first transfected with dominant-negative EGFR(DN) (pRK5-HERCD533; lane 13) or empty vector (pRK5), and the cells were recovered and then synchronized in S-phase. The cells were then exposed to IR or UV (±E₂) or UV+EGF, and in some cells, UV+tyrphostin 1478 (EGFR tyrosine kinase inhibitor; lane 4) or UV+EGF+LY. p53 and Chk1 phosphorylation at the ATR sites were determined from cell lysates obtained at 15 and 5 min after IR or UV, respectively, by specific IB. Total Chk1 protein serves as loading control.

E₂ overcomes the G2/M checkpoint and promotes cell viability after DNA damage. (a) G2/M checkpoint was determined by cells showing nuclear phospho-histone H3 (Serine 10) staining. Randomly cycling MCF7 cells were exposed to brief IR±E2 or E₂ alone, and at 4 h after IR the cells were fixed and permeabilized for antibody staining. A representative study of two is shown, and nuclei are identified by DAPI stain. Bar graph is from 200 cells counted per condition in each of two experiments for phospho-H3 staining. (b) Estrogen promotes bypass of the G2/M checkpoint. MCF7 cells were synchronized with nocodazole for 16 h to induce a G2/M block. Nocodazole-containing media was then removed and replaced with fresh DMEM-F12 media alone (control or IR alone conditions) or with E₂ 10 nM in DMEM-F12, and then cells were exposed to IR. The cells were stained with antibody to detect phospho-H3 over 6 h, and the bar graph represents results from 200 cells per condition from each of two experiments. (c) Cell viability of MCF7 exposed briefly to DNA-damaging agents in the absence or presence of E₂ was determined by the MTT assay, as described in Materials and Methods. Bar graph is from the mean ± SEM of three experiments. * p < 0.05 by ANOVA+Schefe's test for control vs DNA-damaging agents; ⁺ p < 0.05 for UV-damaging agents vs UV-damaging agents+E₂.

E₂ modulates DNA damage complexes in breast cancer and normal breast epithelial cells. (a) Cyclobutane pyrimidine dimers form after exposure to UV. Spectrophotometry quantified CPD immunofluorescence of S-phase MCF7cells exposed to brief radiation ± E₂, and cells were fixed with 4% paraformaldehyde at the times indicated and incubated with CPD-specific antibody. PI fluorescence from the nuclei of the cells was used to normalize all readings. Data are mean ± SEM from three experiments; * p < 0.05 for control vs UV or UV+E₂ at 2 h; ⁺ p < 0.05 for UV vs UV+E₂ at 4–8 h. (b) E₂ stimulates increased chromosomal damage in the setting of IR. MCF7 were exposed to 2 Gy IR and then incubated with or without E₂ and ICI for 48 h. Metaphase spreads were conducted as described in Materials and Methods. Chromosome analysis was conducted from 200 metaphase spreads per condition in each of two experiments. * p < 0.05 for control (sham irradiation) vs IR or IR+E₂; ⁺ p < 0.05 for IR+E₂ vs IR+E₂+ICI. (c) E₂ modulates ATR activation after IR exposure of S-phase mouse mammary epithelial cells (top), and ATR activation in ERα-transfected human breast epithelial cells (MCF10A; bottom). Primary cultures of mouse epithelial cells were from 8-wk-old female mice. ERα or pcDNA3 (control) was transfected into the ER-negative MCF10A cells for experiments, the cells recovered and synchronized in S-phase. (d) E₂ delays γH2AX foci formation and resolution after IR in S-phase normal breast epithelial cells. Studies were carried out as described in Materials and Methods.