|Home | About | Journals | Submit | Contact Us | Français|
The E2F transcription factor integrates cellular signals and coordinates cell cycle progression. Our prior studies demonstrated selective induction and stabilization of E2F1 through ATM-dependent phosphorylation in response to DNA damage. Here we report that DNA topoisomerase IIβ binding protein 1 (TopBP1) regulates E2F1 during DNA damage. TopBP1 contains eight BRCT (BRCA1 carboxyl-terminal) motifs and upon DNA damage is recruited to stalled replication forks, where it participates in a DNA damage checkpoint. Here we demonstrated an interaction between TopBP1 and E2F1. The interaction depended on the amino terminus of E2F1 and the sixth BRCT domain of TopBP1. It was specific to E2F1 and was not observed in E2F2, E2F3, or E2F4. This interaction was induced by DNA damage and phosphorylation of E2F1 by ATM. Through this interaction, TopBP1 repressed multiple activities of E2F1, including transcriptional activity, induction of S-phase entry, and apoptosis. Furthermore, TopBP1 relocalized E2F1 from diffuse nuclear distribution to discrete punctate nuclear foci, where E2F1 colocalized with TopBP1 and BRCA1. Thus, the specific interaction between TopBP1 and E2F1 during DNA damage inhibits the known E2F1 activities but recruits E2F1 to a BRCA1-containing repair complex, suggesting a direct role of E2F1 in DNA damage checkpoint/repair at stalled replication forks.
The E2F transcription factors E2F1 to E2F6 bind to E2F sites in promoters and regulate the expression of a large array of genes that encode proteins important for DNA replication and cell cycle progression. In response to growth signals, activated G1 cyclin-dependent kinases phosphorylate retinoblastoma protein (pRb) and release E2F from pRb binding. This event is critical in controlling G1/S transition (9, 23). Among the E2F family members, E2F1, E2F2, and E2F3 are transcriptional activators and are induced in response to growth stimulation, with peak accumulation at G1/S. Together, they are essential for cellular proliferation since a combined mutation of E2F1, -2, and -3 completely blocks cellular proliferation (31), whereas E2F4 and E2F5 function mainly as transcriptional repressors (9, 23).
Among the same subclass of E2F family, each individual E2F member has its unique biological properties. The unique feature for E2F1 is its activity in apoptosis induction (8, 14, 21) and its function as a tumor suppressor. E2F1−/− mice are viable but develop a broad spectrum of tumors (10, 37). This unique tumor suppressor function could be partly attributed to the apoptotic activity of E2F1. In comparison, overexpression of E2F2 or E2F3 was shown to induce apoptosis as well, but to a lesser extent than with E2F1 (21, 28).
Recently, we showed that DNA damage specifically induces E2F1 but not the other E2F family proteins (17). In response to DNA damage, two key regulators of DNA damage checkpoints, ATM and ATR (ATM-Rad3-related protein) kinases, phosphorylate E2F1 but not E2F2, E2F3, or E2F4. The specific phosphorylation of E2F1 at serine 31 by ATM/ATR leads to stabilization of E2F1. The induction of E2F1 is required for DNA damage-induced apoptosis in thymocytes, establishing a role for E2F1 in response to DNA damage. E2F1 was also shown to be associated with Nbs1 and Mre11 recombination-repair complexes (19). This association was suggested to target the Mre11 complex near origins of replication to suppress the firing of these origins upon DNA damage (19). Taken together, these data suggest a role for E2F1 in the DNA damage checkpoint. Loss of the E2F1-mediated checkpoint leads to tumor development, as observed in E2F1−/− animals but not in other E2F knockout mice. Interestingly, Ren and colleagues have identified direct E2F target genes that encode components of the DNA damage checkpoint and repair pathways by chromatin immunoprecipitation and DNA microarray analysis, further supporting the role of E2F in genomic surveillance (26).
While E2F1 is induced in an ATM-dependent manner during DNA damage, how it is involved in the DNA damage checkpoint response remains unclear. We speculate that another protein(s) may interact with E2F1 and regulate E2F1 function or collaborate with E2F1 for the DNA damage response. Among the E2F family, we anticipate that these interactions will be found to be specific to E2F1, since only E2F1 is induced in response to DNA damage (17). The N terminus of E2F1 contains an ATM phosphorylation site and is distinct from that of other E2F family members, rendering it a potential target for specific functional regulation. Thus, we used this fragment of E2F1 as a bait to perform yeast two-hybrid screen and identified DNA topoisomerase IIβ-binding protein I (TopBP1) as an E2F1-interacting partner. Here we report our characterization of the specific interaction and functional regulation of E2F1 by TopBP1.
TopBP1 was cloned initially in a yeast two-hybrid screen as a protein interacting with topoisomerase IIβ (34). It contains eight BRCA1 carboxyl-terminal (BRCT) motifs (Fig. (Fig.1),1), which are found in proteins involved in DNA repair (DNA ligases III and IV and XRCC1) and cell cycle checkpoints (Rad9, Cut5/Rad4, and Crb2). TopBP1 shares sequence homology with Cut5/Rad4 (Schizosaccharomyces pombe), DPB11 (Saccharomyces cerevisiae), and Mus101 (Drosophila melanogaster). Cut5/Rad4 is a checkpoint Rad protein. It contains four BRCT motifs and is required for DNA replication, DNA damage, and replication checkpoint (5, 7). Cut5/Rad4 associates with Crb2 and Chk1 and is required for activation of checkpoint protein kinases Chk1 and Cds1/Chk2 (reviewed in reference 39). Mus101 contains seven BRCT motifs and is most homologous to TopBP1. The C-terminal region of TopBP1 has 33% identity and 51% similarity with Mus101. Mutations of Mus101 result in hypersensitivity to DNA damage, defects in DNA synthesis, chromosomal instability, and failure to condense heterochromatic regions of chromosome (33). Moreover, Mus101 mutations also lead to a defect in eggshell formation due to defective amplification of clusters of chorion proteins genes, which is a form of DNA replication specific to follicle cells (25). Thus, TopBP1 may also be involved in DNA replication checkpoint control.
Upon gamma irradiation, TopBP1 colocalizes with Nbs1, BRCA1, and 53BP1 in the ionizing radiation-induced foci, where damaged DNA sites reside (18, 22, 36). The BRCT motif of TopBP1 can bind DNA breaks (35). Most of TopBP1 does not colocalize at sites of ongoing DNA replication in unirradiated cells but is relocalized to stalled replication forks upon DNA damage (18). Thus, it appears that the recruitment of TopBP1 may be important in the rescue of stalled replication forks. The involvement of TopBP1 in DNA replication is supported by the demonstration that incubation of an antibody against the sixth BRCT motif of TopBP1 inhibits replicative DNA synthesis in an in vitro HeLa cell nucleus replication assay (18).
TopBP1 contains a transactivation domain in its N terminus (Fig. (Fig.1).1). It also interacts with human papillomavirus type 16 transcription-replication factor E2 and enhances the ability of human papillomavirus type 16 E2 to activate transcription and replication (4). Thus, TopBP1 may be involved in transcription regulation as well. In addition to topoisomerase IIβ and human papillomavirus E2, TopBP1 can interact with human DNA polymerase , checkpoint protein human Rad9 (18), and Miz-1 (12). Interestingly, like E2F1, TopBP1 is phosphorylated by ATM upon gamma irradiation (36). The irradiation-induced phosphorylation inhibits the ubiquitination of TopBP1 and stabilizes TopBP1 (13).
We now provide evidence that TopBP1 interacts with E2F1 during DNA damage. This interaction leads to repression of known E2F1 activities, including transcriptional activation and apoptosis induction; instead, E2F1 is recruited to BRCA1-containing nuclear foci, suggesting a novel function involved in the replication checkpoint for E2F1 during DNA damage.
The N terminus of E2F1 (amino acids 1 to 109) was amplified by PCR, inserted into the GAL4 DNA-binding domain cloning vector pAS2-1 (Clontech), and then cotransformed with HeLa cDNA libraries constructed in pGADGH (a gift from Grame Bolger) into Saccharomyces cerevisiae strain Y190. Transformants were screened for growth on plates lacking tryptophan, leucine, and histidine and supplemented with 25 mM 3-amino-1,2,4-triazole at 30°C for 3 to 6 days. Histidine-positive colonies were further screened for positive interaction by β-galactosidase assay. Plasmids harboring cDNA were isolated from positive yeast colonies, transformed into Escherichia coli HB101 by electroporation, and further selected out by plating colonies on M9 plates. The nucleotide sequence of the cDNA was determined by sequencing.
pcDNA3-TopBP1 was constructed by cloning the KpnI/NotI fragment of pBluescript-TopBP1 (KIAA0259, a gift from Kazusa DNA Research Institute) to vector pcDNA3 digested with KpnI and NotI. To construct Flag-tagged TopBP1, the EcoRI fragment of KIAA0259 containing full-length TopBP1 was subcloned into pCMVTag2B (Stratagene). The BamHI/ApaI fragment of pCMVTag2B-TopBP1 was subcloned into pHCRed1-C1 (Clontech) for the HcRed1-TopBP1 fusion protein expression vector. Flag-Δ123TopBP1 and Flag-Δ678TopBP1 were obtained by cloning the KIAA0259 BglII/EcoRI fragment and HindIII fragment, respectively, into pCMVTag2. Flag-Δ8TopBP1 and Flag-Δ78TopBP1 were obtained by XhoI or SalI digestion of Tag2B-TopBP1 and religation. We subcloned the SalI/ApaI fragment of Tag2B-TopBP1 to XhoI- and ApaI-digested Flag-Δ678TopBP1 to obtain Flag-Δ6TopBP1. Flag-BRCT6 was constructed by digestion of Flag-TopBP1CT with SalI and XhoI, followed by religation. The mutant TopBP1 fragments were also subcloned to the pHcRed1C1 vector (Clontech). pEGFP-E2F1 was constructed by moving the BamHI/EcoRI fragment of pcDNA3HA-E2F1 (17) to the pEGFP-C1 vector. pEGFP-E2F1(1-83) was constructed by PCR amplification of amino acids 1 to 83 of E2F1, and the sequences were verified by sequencing. E2F1 mutant S31D was generated by using the GeneEditor in vitro site-directed mutagenesis system (Promega) with the primer 5′-GCGGCTGCTCGACTCTGATCAGATCGTCATCATC-3′.
The full-length cDNAs of TopBP1 and E2F1 were inserted into expression vector pGEX6P, encoding glutathione S-transferase (GST). GST, GST-E2F1, and GST-TopBP1 proteins were induced with 0.1 mM isopropylthiogalactopyranoside (IPTG) in E. coli strain BL21 and purified as described previously (17). The GST portion of GST-E2F1 was excised by PreScission protease (Pharmacia), and 1 μg of purified GST-TopBP1 or GST was incubated in NETN-A buffer (50 mM NaCl, 1 mM EDTA, 20 mM Tris, 0.5% NP-40) with 2 μg of purified E2F1 at 4°C on a rotating shaker for 3 h. GST-TopBP1 was pulled down with glutathione-Sepharose, and the beads were washed six times with NETN-B buffer (100 mM NaCl, 1 mM EDTA, 0.2 mM phenylmethylsulfonyl fluoride), subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and analyzed by Western blotting with anti-E2F1 antibody (C20; Santa Cruz).
293T, 293, HFF (human foreskin fibroblast), and REF 52 cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum, penicillin (50 IU/ml), and streptomycin (50 μg/ml) in a humidified incubator with 5% CO2 and 95% air at 37°C. Cells were transfected with appropriate vectors by a standard calcium phosphate method. The total amount of transfected DNA was made equivalent with pcDNA3 vector. After transfection, cells were incubated at 37°C for 2 days before analysis. Primary mouse embryo fibroblasts (MEFs) were isolated from 13.5-day-old embryos by standard methods. Primary MEFs were transfected with Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions.
The transfected cells were harvested 48 h later with TNN buffer (50 mM Tris, 0.25 M NaCl, 5 mM EDTA, 0.5% NP-40) supplemented with 1 mM dithiothreitol, 1 mM NaF, 1 mM sodium orthovanadate, 20 nM microcystin, 10 μg of leupeptin per ml, 10 μg of aprotinin per ml, 10 μg of pepstatin per ml, 1 mM phenylmethylsulfonyl fluoride, 2 μg of antipain per ml, and 1 μg of chymostatin per ml. An aliquot of the cell lysates was lysed with SDS lysis buffer, the rest of the cell lysates were incubated with appropriate antibodies or beads for 3 to 24 h at 4°C. Anti-Flag (M2) beads were purchased from Sigma, and protein G plus protein A-agarose beads were purchased from Oncogene. The beads were washed three times with TNN buffer. Immunoprecipitates were fractionated by SDS-PAGE and electrotransferred to an Immobilon-P membrane (Millipore). Equal protein loading was confirmed with Ponceau S staining. The specific signals were detected with appropriate antibodies. Antibodies to E2F1 (C-20 and KH-95), E2F2 (C-20), E2F3 (C-18), E2F4 (C-20), and hemagglutinin (HA) (Y11) were purchased from Santa Cruz. TopBP1 antibody was purchased from BD Transduction Laboratories. Flag antibody (F7425) was purchased from Sigma.
293 cells were plated on collagen-coated coverslips in six-well plates. To visualize enhanced green fluorescent protein (EGFP)-E2F1 and HcRed1-TopBP1, transfected cells were fixed in 3% paraformaldehyde for 20 min, and the nuclei were stained with Hoechst 33258. For immunostaining, 293 cells were fixed in 3% paraformaldehyde for 20 min, followed by permeabilization in 0.5% Triton X-100 for 10 min. Cells were then blocked in 2% bovine serum albumin-phosphate-buffered saline at room temperature for 30 min and incubated with primary antibody in blocking solution for 1 h, and fluorescein-conjugated goat anti-rabbit immunoglobulin G (IgG) or Texas Red-X goat anti-mouse IgG (both purchased from Molecular Probes, 1:400 dilution) for 1 h. For immunostaining, E2F1 antibody C20 (1:250) or KH95 (1:50), TopBP1 antibody (BD Transduction Laboratories, 1:250), and BRCA1 antibody (Ab-2, Oncogene, 1:100) were used. Images were captured on a Zeiss fluorescent microscope (Axioplan 2 imaging system). To reduce background signals in the cytoplasm and nucleoplasm, HFF were treated with cytoskeleton-stripping buffer before fixation (20).
The expression constructs (2 μg for E2F1 and its mutants, 10 μg for TopBP1 and its mutants), the promoter plasmids (1 μg for P14ARF-Luc, 1 μg for pKL12-Luc, and 1 μg for pKL12-E2FAB-Luc), and 1 μg of β-galactosidase plasmid were cotransfected into 293T cells as described above. Cells were harvested 2 days later with phosphate-buffered saline. One tenth of the sample was lysed with SDS lysis buffer for Western blotting, and the rest of the cell extract was lysed with reporter lysis buffer (Promega), and luciferase activity and β-galactosidase activity were measured following the manufacturer's procedures. The luciferase activity was normalized against the β-galactosidase activity. All transient expressions in this assay were carried out in triplicate.
The full-length cDNAs for HA-E2F1 and TopBP1 were constructed into adenovirus with the AdEasy system as described previously (11). Viruses were purified by CsCl banding. REF52 cells were used for flow cytometry and bromodeoxyuridine (BrdU) staining. The cells were starved in 0.25% fetal bovine serum for 2 days, followed by adenovirus infection. Serum-starved REF52 cells were infected with AdE2F1 and/or AdTopBP1, and BrdU labeling was performed as described previously between 19 and 40 h after infection (8). To assay E2F1-induced apoptosis, infected REF52 cells were cultured in Dulbecco's modified Eagle's medium with 0.25% fetal bovine serum for 4 more days before harvesting, and apoptosis was quantified by propidium iodide-flow cytometry (8).
The N-terminal portion of E2F1 (amino acids 1 to 109) contains a domain required for protein degradation and ATM phosphorylation, a cyclin A binding domain, and a nuclear localization sequence (NLS). In order to identify the E2F1-interacting partner(s), fusion proteins containing this N-terminal fragment of E2F1 and the GAL4 DNA binding domain were used to screen a HeLa cell cDNA library. A total of 12 million clones were screened, and five positive clones encoding a carboxyl portion of TopBP1 (amino acids 775 to 1435, named TopBP1CT) were isolated. TopBP1CT contains BRCT6, -7, and -8 and the NLS (Fig. (Fig.1A).1A). The interaction was verified by selective growth of yeast cells expressing both the N terminus of E2F1 and TopBP1CT on plates lacking tryptophan, leucine, and histidine (Fig. (Fig.2A).2A). Their interaction was also verified by β-galactosidase assay on colony-lift filters (data not shown).
To confirm the direct E2F1-TopBP1 interaction, we examined in vitro binding between these two proteins. GST-TopBP1 and E2F1 proteins were expressed in E. coli and purified. E2F1 protein was incubated with GST-TopBP1 or GST, and the GST-TopBP1 protein was then pulled down with glutathione-Sepharose. GST-TopBP1 but not GST pulled down E2F1, as detected with E2F1-specific antibody (Fig. (Fig.2B).2B). This result demonstrates a direct interaction between TopBP1 and E2F1 in vitro.
Next, we examined the binding between E2F1 and TopBP1 in HEK 293T cells. We transiently expressed Flag-tagged TopBP1 and its mutants with HA-E2F1 in 293T cells and immunoprecipitated TopBP1 from the lysates with anti-Flag-conjugated agarose beads, followed by immunoblotting. Full-length TopBP1 coimmunoprecipitated with E2F1 (Fig. (Fig.3A).3A). A TopBP1 mutant containing only the BRCT6, -7, and -8 motifs (TopBP1CT) also interacted with E2F1; however, a TopBP1 mutant lacking the BRCT6, -7, and -8 motifs (Δ678TopBP1) was unable to bind to E2F1. This result indicates a specific interaction between E2F1 and the C terminus of TopBP1. Deletion of the BRCT6 motif in TopBP1 or deletion of the N-terminal 88 amino acids of E2F1 significantly impaired the interaction, indicating the requirement for these domains for interaction (Fig. (Fig.3B).3B). Indeed, the BRCT6 motif alone was able to interact with E2F1 (Fig. (Fig.3C).3C). These results demonstrate that the BRCT6 motif of TopBP1 and the N terminus of E2F1 are required for their interaction.
Following ionizing radiation, TopBP1 is recruited to DNA breaks and colocalizes with Nbs1, BRCA1, and 53BP1 in discrete nuclear foci (18, 35, 36). Since E2F1 interacted with TopBP1, we tested whether E2F1 could localize with TopBP1 in these foci during DNA damage. We examined the localization of endogenous E2F1 and TopBP1 in human foreskin fibroblasts (HFF) by immunofluorescent staining. In growing HFF, both E2F1 and TopBP1 were distributed diffusely in nuclei (Fig. (Fig.4A).4A). Upon adriamycin treatment which caused DNA damage, E2F1 formed discrete nuclear foci and colocalized with TopBP1 in these foci (Fig. (Fig.4B4B).
To address whether binding to TopBP1 is responsible for E2F1 nuclear focus formation, we performed a series of cotransfection experiments. We constructed a plasmid encoding the EGFP-E2F1 fusion protein in which EGFP was fused to the N terminus of E2F1. This fusion protein was as functional as untagged E2F1 in transcriptional activity (data not shown). We also constructed a plasmid encoding a HcRed1-TopBP1 fusion protein. HcRed1 is a far-red fluorescent protein and was generated by mutagenesis of a nonfluorescent chromoprotein from the reef coral Heteractis crispa. After cotransfecting pEGFP-E2F1 and pHcRed1-TopBP1 into 293 cells, we determined the subcellular localization of both proteins by microscopy.
EGFP-E2F1 formed a diffuse nuclear pattern; in contrast, HcRed1-TopBP1 formed a nuclear punctate pattern (Fig. (Fig.5A).5A). The punctate nuclear pattern of HcRed1-TopBP1 was attributed to TopBP1 because HcRed1 protein was distributed diffusely throughout the cell (data not shown). A similar distribution was observed for Flag-TopBP1 by immunostaining (see Fig. Fig.9)9) and for endogenous TopBP1 upon DNA damage (Fig. (Fig.4B).4B). Coexpression of TopBP1 relocalized E2F1 from homogenous nuclear staining to discrete punctate nuclear foci, where E2F1 colocalized with TopBP1 (Fig. (Fig.5B).5B). The relocalization was specific to E2F1 because the distribution of EGFP was not affected by TopBP1 expression (Fig. (Fig.5D).5D). The fragment containing the E2F1 N-terminal 83 amino acids was sufficient to interact with TopBP1 and was recruited to the foci upon coexpression with TopBP1 (Fig. (Fig.5C).5C). In contrast, Δ6TopBP1, lacking the E2F1 interaction domain, was able to form foci but failed to induce E2F1 to the foci (Fig. (Fig.5E5E).
The observation that Δ6TopBP1 still formed foci is consistent with a prior report that only BRCT5 is required for TopBP1 focus formation (36). Δ78TopBP1 and Δ8TopBP1 lack the NLS in the carboxyl terminus (Fig. (Fig.1A)1A) and were localized mainly in the cytoplasm (Fig. (Fig.5F5F and data not shown). We also observed colocalization of EGFP-E2F1 and HcRed1-TopBP1 within nuclear foci in MCF7 cells, HFF, and primary mouse embryo fibroblasts (data not shown), indicating that the relocalization of E2F1 to nuclear foci induced by TopBP1 is not unique to one cell line. Taken together, the data show that TopBP1 is able to interact with E2F1 and recruit E2F1 to nuclear foci during DNA damage.
To test whether TopBP1 regulates E2F1 activity, we used a p14ARF promoter-luciferase construct (2) and pCMV-βgal for transfection control as a reporter assay for E2F1 transcriptional activity in 293T cells. The p14ARF promoter contains E2F binding sites, and deletion of these sites abolishes the responsiveness to E2F1 (2). TopBP1 suppressed E2F1 transcriptional activity without affecting E2F1 protein levels (Fig. (Fig.6A).6A). We examined a series of TopBP1 mutants (Fig. (Fig.1A)1A) for their ability to repress E2F1. TopBP1 mutants (TopBP1CT or Δ123-TopBP1, lacking the N-terminal BRCT1, -2, and -3 motifs), which were able to bind E2F1, also repressed E2F1 activity (Fig. (Fig.6A).6A). In contrast, Δ6-TopBP1, which lacks the E2F1 interaction domain, failed to repress E2F1 transcriptional activity. Δ8- and Δ78-TopBP1, lacking nuclear localization signals in the C terminus, were localized in the cytoplasm (Fig. (Fig.5F)5F) and thus failed to repress E2F1 transcriptional activity.
Several E2F1 deletion mutants (Fig. (Fig.1B)1B) were also examined for regulation by TopBP1. Two E2F1 mutants (Δ1-85 and Δ1-88) which lack the N terminus and did not interact with TopBP1 (Fig. (Fig.3B)3B) retained their transcriptional activity but were no longer repressed by TopBP1 (Fig. (Fig.6B).6B). In contrast, Marked box-deleted E2F1 (Δ283-358), which bound TopBP1 (data not shown), was repressed by TopBP1 (Fig. (Fig.6B).6B). The effect of TopBP1 on the repression of E2F1 activity was dose dependent, so that greater than 90% of E2F1 activity was inhibited when enough TopBP1 was expressed (Fig. (Fig.6C).6C). Similar dose-dependent repression of E2F1 transcriptional activity by TopBP1 was also observed in MCF7 and REF52 cells (data not shown).
The repression of E2F1 activity by TopBP1 is not an indirect consequence of relocating E2F1, since a mutant TopBP1 which failed to relocate E2F1 was still able to repress E2F1 transcriptional activity. TopBP1CT (containing BRCT6, -7, and -8) did not form foci because it lacked BRCT5, the domain required for TopBP1 focus formation (36) and did not relocalize E2F1 (data not shown). However, TopBP1CT was able to repress E2F1 transcriptional activity (Fig. (Fig.6A),6A), indicating a specific regulation of E2F1 activity by TopBP1 rather than an indirect consequence of protein relocalization.
The role of E2F in transcriptional activation of the p68 subunit of DNA polymerase α has been established (24). The reporter plasmid pKL12(−164) contains the wild type promoter of p68, while pKL12 E2FAB carries mutations of both E2F sites and no longer responds to E2F (24). We used this E2F-responsive promoter construct as an additional assay to test the regulation of E2F1 activity by TopBP1. As observed for the p14ARF promoter, TopBP1 repressed E2F1-mediated transcriptional activation of the p68 promoter but not the promoter carrying mutations of the E2F sites (Fig. (Fig.6D).6D). Taken together, the data show that the transcriptional activity of E2F1 is regulated by TopBP1 through the interaction between the N terminus of E2F1 and BRCT6 of TopBP1.
DNA damage induces colocalization of E2F1 and TopBP1 in nuclear foci (Fig. (Fig.4),4), suggesting that their physical interaction is stimulated by DNA damage. To test that, endogenous E2F1 was immunoprecipitated from 293T cell lysates, and the E2F1-containing complexes were immunoblotted with a specific antibody against TopBP1. Their interaction was not detectable in untreated cells but was readily detectable upon neocarzinostatin (NCS) treatment (Fig. (Fig.7A).7A). NCS, a radiomimetic chemical, leads to double-strand DNA breaks and induction of ATM kinase activity (1). This result demonstrates the interaction between endogenous TopBP1 and E2F1 during DNA damage.
Although the increased protein levels of E2F1 and TopBP1 upon NCS treatment contributed to the induction of their interaction, the amplitude of induction in interaction (greater than 10-fold) compared to the induction in protein levels (1.5- to 2-fold) suggests that DNA damage induces their interaction. The fact that colocalization of endogenous E2F1 and TopBP1 in discrete nuclear foci was observed only upon DNA damage further supports the induction of their interaction by DNA damage (Fig. (Fig.4B).4B). Since ATM phosphorylates E2F1 at serine 31 (17), within the TopBP1 binding domain, we postulated that ATM phosphorylation of E2F1 at serine 31 might regulate its binding to TopBP1. To address this hypothesis, we examined the ability of two E2F1-S31 mutants, S31A and S31D (mimicking unphosphorylated and phosphorylated forms, respectively), to interact with TopBP1.
As shown in Fig. Fig.7B,7B, the S31A mutation significantly decreased the interaction between E2F1 and Flag-TopBP1, while the S31D mutation preserved the binding. Similar results were observed between Myc-tagged TopBP1 and E2F1 mutants (data not shown). Mutations at serine 31 also affected functional regulation. TopBP1 did not repress the transcriptional activity of S31A-E2F1 but repressed that of S31D-E2F1 (Fig. (Fig.7C).7C). These results indicate that modification of the E2F1 S31 residue regulates the physical and functional interaction between E2F1 and TopBP1. In conjunction with the requirement for ATM for E2F1 S31 phosphorylation and the induction of endogenous E2F1-TopBP1 interaction by the ATM activator, these results suggest that ATM phosphorylation may likewise regulate the interaction between E2F1 and TopBP1.
Next, we examined whether the binding and regulation by TopBP1 were specific to E2F1. While the interaction between TopBP1 and E2F1 was readily detected, there was no detectable interaction between TopBP1 and E2F2, -3, or -4 (Fig. (Fig.8A).8A). Moreover, TopBP1 did not repress the transcriptional activity of E2F2, -3, or -4 (Fig. (Fig.8B).8B). Thus, regulation by TopBP1 is unique to E2F1, as the specificity observed in ATM phosphorylation and DNA damage-induced induction.
TopBP1 contains eight BRCT motifs and is able to bind multiple proteins. We hypothesized that TopBP1 could recruit E2F1 to a complex and mediate the interaction between E2F1 and other proteins. TopBP1 colocalizes with BRCA1 in S-phase nuclei (18), and ionizing irradiation causes TopBP1 to form foci and colocalize with BRCA1 (18, 36). BRCA1 is associated with a large complex, named BASC (BRCA1-associated genome surveillance complex), which contains DNA repair proteins, Mre11 complex, DNA replication factor C, and others (30). BRCA1 is also associated with histone deacetylase I (HDAC1) (32, 38), CtBP-interacting protein (CtIP) (16), and the SWI/SNF-related chromatin-remodeling complex (3). These interactions may potentially mediate the repression of E2F1 transcriptional activity by TopBP1. Indeed, purified E2F1 and BRCA1 interacted in vitro, but the interaction was weak, and it was suggested that E2F1 might need to associate with other proteins in order to bind efficiently to BRCA1 (29).
To test whether TopBP1 could facilitate the association between E2F1 and BRCA1, we transiently expressed HA-E2F1 alone or HA-E2F1 and Flag-TopBP1 in 293 cells and performed immunofluorescent staining with antibodies specific for E2F1, BRCA1, and TopBP1. HA-E2F1 was distributed in a homogenous nuclear pattern (Fig. (Fig.9A);9A); however, upon coexpression with Flag-TopBP1, HA-E2F1 colocalized with endogenous BRCA1 (Fig. (Fig.9B)9B) as well as Flag-TopBP1 (Fig. (Fig.9C)9C) in the nuclear foci. Similarly, Flag-TopBP1 colocalized with endogenous E2F1 (Fig. (Fig.9D)9D) and endogenous BRCA1 (Fig. (Fig.9E).9E). Thus, expression of TopBP1 is able to recolocalize both E2F1 and BRCA1 to the same foci.
Localization of TopBP1 in nuclear foci could theoretically be mediated by E2F1. However, it is not very likely because the domain in TopBP1 that is required for punctate nuclear localization (BRCT5) is distinct from its E2F1-binding domain (BRCT6). To definitively rule out this possibility, we examined the subnuclear localization of HcRed1-TopBP1 in primary mouse embryo fibroblasts (MEFs) prepared from an E2F1−/− embryo and its E2F1+/+ sibling embryo. HcRed1-TopBP1 formed nuclear foci in both E2F1+/+ and E2F1−/− MEFs (Fig. (Fig.9F),9F), and endogenous TopBP1 formed foci upon adriamycin treatment in both cell types as well (data not shown), indicating that localization of TopBP1 to foci does not require E2F1. Taken together, these data demonstrate that TopBP1 recruits E2F1 to the BRCA1-containing complex in these nuclear foci. The physical association between E2F1 and BASC suggests a novel role for E2F1 in DNA damage-induced foci. The proximity between E2F1 and chromatin-remodeling proteins within the BRCA1-containing complex could be responsible for the repression of E2F1 transcriptional activity.
E2F1 induces S-phase entry and apoptosis in serum-starved fibroblasts. Infection of recombinant adenoviruses expressing E2F1 in serum-starved REF52 cells has been established to quantitate E2F1-induced DNA synthesis and apoptosis (8). To investigate whether TopBP1 regulates E2F1 function in the induction of S-phase entry and apoptosis, we infected serum-starved REF52 cells with recombinant adenoviruses expressing TopBP1 and E2F1 and assessed the effect of TopBP1 on E2F1-induced BrdU incorporation and E2F1-induced apoptosis. As shown in Fig. 10A, expression of TopBP1 by AdTopBP1 infection significantly inhibited E2F1-stimulated BrdU incorporation in serum-starved REF52 cells without affecting E2F1 protein levels. This inhibition was not a result of a direct effect of TopBP1 on DNA replication because AdTopBP1 infection did not affect serum-stimulated BrdU incorporation. The fact that AdTopBP1 inhibits E2F1-induced but not serum-induced DNA synthesis indicates an effect specifically targeted at E2F1 function by TopBP1.
AdTopBP1 infection also inhibited E2F1-induced apoptosis from 32% to 19% (Fig. 10B), as assayed by propidium iodide-flow cytometry. In other independent experiments, TopBP1 inhibited E2F1-induced apoptosis from 64.5% to 41% and from 75.2% to 58.3%, respectively. Thus, TopBP1 suppresses E2F1 activity in S phase and apoptosis induction.
In response to oncogenic mutations that lead to a loss of Rb function, deregulated E2F activity can trigger apoptosis. During DNA damage, E2F1 is specifically induced and is required for etoposide-induced apoptosis in thymocytes (17). The induction of E2F1 is mediated by ATM phosphorylation-dependent protein stabilization. The ability of E2F1 to induce apoptosis provides a safeguard mechanism for oncogenic as well as genotoxic stress. Supporting this concept is the fact that E2F1−/− mice are viable but have a propensity to develop tumors (10, 37). On the other hand, the apoptosis-inducing activity of E2F1 must be under control to allow normal cellular growth to proceed. Even in the presence of DNA damage, cells need to control cell cycle progression as well as apoptosis to allow DNA repair to complete unless the damage is beyond repair.
The current study uncovers an E2F1-specific regulator that by itself is involved in DNA replication and damage response. The interaction is induced by DNA damage and requires ATM phosphorylation, suggesting a specific functional regulation during genotoxic stress, at the time when both TopBP1 and E2F1 are induced. TopBP1 interacts with E2F1, represses its known activities, and recruits it to the BRCA1-containing repair complex, suggesting a novel role for E2F1 in this complex during DNA damage. These results provide evidence coupling the DNA repair-damage checkpoint with the cell cycle machinery.
In this paper, we provide evidence that purified E2F1 and TopBP1 are able to interact in vitro, indicating a direct interaction between these proteins. The interaction was also demonstrated in vivo and was dependent on the N terminus of E2F1 and the BRCT6 motif of TopBP1. This interaction is unique to E2F1, as TopBP1 specifically recruits E2F1 but not other E2F members to stalled replication forks for checkpoint function.
Prior studies had identified serine 31 as an ATM phosphorylation site (17). While phosphorylation of serine 31 is required for E2F1 stabilization, it does not appear to be required for E2F1 transcriptional activity. Now we demonstrate that the interaction between TopBP1 and E2F1 is induced upon DNA damage and is regulated by the modification on serine 31 of E2F1. A mutation imitating dephosphorylated E2F1 (S31A) significantly diminished the interaction. and a mutation mimicking phosphorylation (S31D) preserved the interaction. Taken together, these observations suggest that phosphorylation of E2F1 by ATM may stimulate its interaction with TopBP1. In this case, ATM would mediate the stabilization of both proteins and also induce their interaction. It is worth noting that we cannot exclude that another kinase(s) may also phosphorylate E2F1 at serine 31.
TopBP1 is phosphorylated by ATM at serine 405 (36). The TopBP1 mutants with N termini deleted, including TopBP1CT and Δ123-TopBP1, interacted with E2F1 and regulated E2F1 activity to the same degree as wild-type TopBP1, indicating that the N terminus of TopBP1 (containing the serine 405 residue) is not required for E2F1 binding. However, whether phosphorylation at serine 405 of TopBP1 could further stimulate E2F1 binding warrants further investigation.
The interaction between TopBP1 and E2F1 leads to repression of E2F1 transcriptional activity but not the E2F1 protein level. The inhibition of E2F1 activity by TopBP1 is rather specific and is not due to a general transcription repression, since TopBP1 did not inhibit the activity of E2F2, E2F3, E2F4, or several E2F1 mutants. In fact, TopBP1 functions as a transcription coactivator with human papillomavirus type 16 E2 protein (4). Unlike the interaction with E2, in which the N-terminal transactivation domain of TopBP1 is required for synergistic activation with E2 (4), TopBP1 does not require its N-terminal domain to repress E2F1 function. The data presented in this paper were obtained with six TopBP1 mutants and five E2F1 mutants (including point mutations). Among these mutants, binding correlated very well with functional regulation, strongly suggesting that TopBP1 represses E2F1 activity through a direct physical interaction of the two proteins. The repression of E2F1 activity is not an indirect consequence of protein seclusion, since a TopBP1 mutant (TopBP1CT) can repress E2F1 activity without relocating E2F1 to foci. TopBP1 may mediate the interaction between E2F1 and other proteins involved in chromatin remodeling and inhibit transcription. Further investigation is needed to explore the mechanism of how TopBP1 regulates E2F1 activity. In the current study, we used overexpression of TopBP1 and E2F1 to demonstrate and characterize this novel interaction. Future studies will examine the physiological role of endogenous protein interactions.
The fact that TopBP1 inhibits E2F1 function in S-phase entry and apoptosis induction underscores the extent of the regulation. The repression of E2F1 activity by TopBP1 may be important for cell survival in coping with DNA damage. Since TopBP1 can bind DNA breaks and interact with multiple proteins through its unusual number of BRCT motifs, it may be involved in the coordination of these proteins for repairing stalled replication forks. Through the interaction with E2F1, TopBP1 may control cell cycle progression and inhibit E2F1-induced apoptosis to allow completion of DNA repair. On the other hand, TopBP1 expression peaks in S phase, when high levels of E2F1 protein are present (18). TopBP1 may be responsible for controlling E2F1-induced apoptosis during normal cellular proliferation. In this regard, while both E2F1 and E2F3 DNA binding activities accumulate during the initial G1 following a growth stimulus, only E2F3 activity, not E2F1 activity, reaccumulates in subsequent G1/S transitions despite reaccumulation of both E2F1 and E2F3 proteins (15). Interestingly, antisense oligomers that inhibited expression of TopBP1 induced apoptosis (36). Whether this apoptosis is mediated by E2F1 remains to be determined. It would also be very interesting to determine whether TopBP1 regulates E2F1 during cell cycle.
Our observation on the induction of E2F1 foci by DNA damage provides a new insight on E2F1 function. Upon expression of TopBP1, both E2F1 and BRCA1 are recruited to TopBP1 foci. Previously TopBP1 foci were shown to contain 53BP1, Nbs1, and BRCA1 (18, 36). The nuclear foci contained stalled replication forks, as revealed by PCNA immunostaining or detection of BrdU incorporation sites (18). The BRCT motifs of TopBP1 were shown to be able to bind damaged DNA directly (35). The E2F1 binding domain (BRCT6) was required for TopBP1 to induce E2F1 foci (Fig. (Fig.5E).5E). Thus, it appears that TopBP1 is first recruited to stalled replication forks and then recruits other proteins (including E2F1) to these sites. The role of TopBP1 in these foci remains speculative. It was suggested that TopBP1 might be involved in rescue of the stalled replication forks since it is required for DNA replication in a cell-free replication system (18). More intriguing is that a neutralizing TopBP1 antibody against the BRCT6 motif but not other regions of TopBP1 can inhibit DNA replication (18). BRCT6 is also responsible for critical E2F1 binding and regulation. This result raises the possibility that E2F1 may mediate TopBP1 function on DNA replication forks.
Recently, the notion of direct involvement of E2F1 in DNA replication has gained some support from two series of experiments. E2F1 was reported to interact with the Mre11 complex (containing Mre11, Rad50, and Nbs1) at the sites of DNA replication (19). The Mre11-Nbs1 complex is important for recombinational DNA repair, replication, and activation of the S-phase checkpoint induced by DNA damage. E2F1 can target the complex to E2F sites proximal to replication origins (19). Moreover, Orr-Weaver and colleagues have reported that a Drosophila E2F1 mutation with reduced DNA binding activity but retaining transactivation and Rb binding functions impaired ORC (origin recognition complex) localization to replication origins within the chorion gene clusters and led to abnormal eggshell formation (27). Furthermore, dE2F and ORC are bound to the chorion origins of replication in vivo, and mutations of Rbf (a retinoblastoma homologue in the fly) fail to limit DNA replication (6). These results suggest a direct involvement of E2F in control of replication origin firing in D. melanogaster. Our data that TopBP1 recruits E2F1 to nuclear foci further suggest a direct involvement of E2F1 in DNA replication control in mammalian cells.
The interaction between TopBP1 and E2F1 is also supported by studies with their Drosophila homologues. Mus101 mutations lead to a defect in eggshell formation due to defective amplification of clusters of chorion proteins genes (25). This defect is reminiscent of that observed in dE2F mutants (27), suggesting genetic interaction between Mus101 and dE2F. Thus, the interaction described in this report may represent an evolutionarily conserved mechanism in controlling DNA replication.
We thank Graeme Bolger for the HeLa cDNA library, Kazusa DNA Research Institute for the KIAA0259 construct, Joe Nevins for the Δ1-88 E2F1 and Δ283-358 E2F1 constructs, Karen Vousden for the p14ARF promoter-Luc (E1β-Luc) construct, and Fumio Hanaoka for pKL12 and pKL12 E2FAB. We also appreciate critical reading of the manuscript by Joe Nevins and Graeme Bolger. We thank Marion Spell at the UAB Flow Cytometry core facility for the flow cytometry analyses and Kun-Sang Chang for suggestions on immunostaining.
The work was supported by grants from a General Motors Cancer Research Scholar Award (W.-C.L.), American Cancer Society UAB Research Grant Program (IRG-6000141, P. I. Albert F. LoBuglio), NIH/NCI K12 CA 7693705 (W.-C.L.), and a UAB Avon/Breast Cancer SPORE Career Development Award (W.-C.L. and F.-T.L.). W.-C.L. and F.-T.L. are recipients of an HHMI/UAB Faculty Development Award.