If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Divisions of Developmental Oncology Research andDivisions of Molecular Pharmacology and Experimental Therapeutics, Mayo Graduate School, and the Divisions ofDivisions of Developmental Oncology ResearchRadiation Oncology, Mayo Clinic and Foundation, Rochester, Minnesota 55905
* This work was supported by National Institutes of Health Grant R01-CA84321 and grants from the Mayo Clinic Foundation (to L. M. K.) and the Magnus Ehrnrooth and Oskar Öflund Foundations (to P. R. M.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Rad17, Rad1, Hus1, and Rad9 are key participants in checkpoint signaling pathways that block cell cycle progression in response to genotoxins. Biochemical and molecular modeling data predict that Rad9, Hus1, and Rad1 form a heterotrimeric complex, dubbed 9-1-1, which is loaded onto chromatin by a complex of Rad17 and the four small replication factor C (RFC) subunits (Rad17-RFC) in response to DNA damage. It is unclear what checkpoint proteins or checkpoint signaling events regulate the association of the 9-1-1 complex with DNA. Here we show that genotoxin-induced chromatin binding of 9-1-1 does not require the Rad9-inducible phosphorylation site (Ser-272). Although we found that Rad9 undergoes an additional phosphatidylinositol 3-kinase-related kinase (PIKK)-dependent posttranslational modification, we also show that genotoxin-triggered 9-1-1 chromatin binding does not depend on the catalytic activity of the PIKKs ataxia telangiectasia-mutated (ATM), ataxia telangiectasia and Rad3-related (ATR), or DNA-PK. Additionally, 9-1-1 chromatin binding does not require DNA replication, suggesting that the complex can be loaded onto DNA in response to DNA structures other than stalled DNA replication forks. Collectively, these studies demonstrate that 9-1-1 chromatin binding is a proximal event in the checkpoint signaling cascade.
DNA damage and replication stress activate checkpoint signaling pathways that block cell cycle progression, activate programmed cell death, and influence DNA repair. Biochemical and genetic studies in yeasts and mammals have identified many of the key components and arranged them into checkpoint signaling cascades (reviewed Refs.
). Activated ATM, ATR, Chk1, and Chk2 then phosphorylate additional checkpoint proteins, including Rad9, BRCA1, p53, Cdc25A, Cdc25C, Nbs1, and other proteins that mediate checkpoint activation and DNA repair.
In addition to the PIKKs, the checkpoint proteins, Rad9, Hus1, Rad1, and Rad17 (using Schizosaccharomyces pombe nomenclature) are key elements of checkpoint signaling pathways that are conserved from the yeasts to humans. Disruption of these genes in S. pombeblocks genotoxin-induced Chk1 activation (
). Such studies suggest that these proteins function early in the checkpoint signaling pathway. Recent biochemical studies further support this notion. Rad9, Hus1, and Rad1 form a stable heterotrimeric complex (the 9-1-1 complex) (
). PCNA subunits assemble into a toroidal clamp complex that is loaded around DNA by the clamp loader, replication factor C (p140-RFC), a protein complex composed of one large subunit (p140) and four small subunits (reviewed in Ref.
). Consistent with the proposed biochemical functions of the 9-1-1 and Rad17-RFC complexes as a clamp-clamp loader pair, Rad9, Hus1, and Rad1 bind chromatin in response to DNA damage in a Rad17-dependent manner (
), suggesting that many different lesions might be recognized by Rad17-RFC. Recent work demonstrated that stalled replication forks induce Hus1 chromatin binding in Xenopus egg extracts in a DNA-polymerase α-dependent manner (
). It is unclear, therefore, whether DNA lesions generate chromatin-bound 9-1-1 complex because they block or slow DNA replication fork progression or whether other DNA structures can trigger 9-1-1 chromatin binding.
Although the 9-1-1 complex and the PIKKs are required for Chk1 activation, it is unclear how these proteins affect one another's function. In humans, ATR and the 9-1-1 complex are recruited to DNA lesions independently of one another (
), raising the possibility that other PIKKs may regulate genotoxin-induced Rad9 chromatin binding.
Collectively, these studies raise the following questions. First, does Ser-272 phosphorylation affect 9-1-1 chromatin binding? Second, do other PIKK-dependent posttranslational modifications occur on Rad9, and do they regulate Rad9 chromatin binding? Third, is DNA replication fork stalling the sole signal that induces 9-1-1 chromatin binding? In this study, we use Rad9 mutants, pharmacologic PIKK inhibitors, and G1-synchronized cells to address the roles of DNA replication and other checkpoint proteins in the regulation of the 9-1-1 complex. Collectively, our results demonstrate that the 9-1-1 complex can be loaded onto DNA independently of all known potential regulators, suggesting that loading of the 9-1-1 complex onto chromatin is a proximal event in the checkpoint signaling pathway.
Antibodies recognizing human Rad9, Hus1, Rad1, and Rad17 have been described previously (
). Anti-phospho-Chk1 (Ser-345) and anti-Chk1 (G4) antibodies were purchased from Cell Signaling Technology (Beverly, MA) and Santa Cruz Biotechnology (Santa Cruz, CA), respectively, and were used according to the manufacturer's instructions. Anti-phospho-Rad9 (Ser-272) antibodies (described in Ref.
) were a generous gift from Dr. Eva Lee (The University of Texas Health Science Center, San Antonio, TX). Anti-phospho-Chk2 (Thr-68) and anti-Chk2 were generous gifts from Dr. Junjie Chen (Mayo Clinic, Rochester, MN) (
) was a gift from F. Quelle (University of Iowa) and was cultured in RPMI 1640 containing 10% fetal bovine serum and 70 pg/ml interleukin-3. K562 (1 × 107) cells were transiently transfected by electroporation (345-volt, 10-ms pulse) using a T820 electroporator (BTX Inc., San Diego, CA). A total of 40 μg of plasmid DNA was used per transfection.
G1-enriched populations of 32D/Bcl-xL cells were prepared by culturing the cells in RPMI 1640 containing 10% fetal bovine serum but lacking interleukin-3 for 30 h. Cell cycle distribution profiles were determined by flow cytometric analysis as described previously (
4-Nitroquinoline oxide (4-NQO) was purchased from Sigma, dissolved in Me2SO, and stored at −20 °C. Hydroxyurea (HU) and caffeine were obtained from Sigma. Wortmannin was obtained from R. Schultz, Developmental Therapeutics Program, National Cancer Institute. HU was prepared fresh in phosphate-buffered saline. Wortmannin was dissolved in Me2SO, stored at −70 °C, and diluted in phosphate-buffered saline prior to addition to the cells. Caffeine was dissolved fresh in 10 mm PIPES. Cells were irradiated with a 137Cs source at a dose rate of 10 Gy/min.
Preparation of Cell Lysates and Chromatin-binding Assay
To analyze protein expression, cells were lysed in 50 mm HEPES, 1% Triton X-100, 10 mm NaF, 30 mm Na4P207, 150 mm NaCl, and 1 mm EDTA containing freshly added 10 mm β-glycerophosphate, 1 mmNa3VO4, 10 μg/ml pepstatin A, 5 μg/ml aprotinin, 10 μg/ml leupeptin and 20 μm microcystin-LR. All lysates were either immunoprecipitated as indicated or boiled with SDS-PAGE sample buffer. S-protein-agarose (Novagen, Inc., Madison, WI) was used to precipitate S-tagged proteins. Proteins were separated on 10% SDS-PAGE gels, transferred to Immobilon P membranes, and immunoblotted as described. Fractionation of unbound and chromatin-bound Rad9 was performed as described previously (
S tag-Rad9-pIRES2-EGFP was generated by adding an in-frame S tagTM (Novagen Inc.) to the amino terminus of Rad9 using a PCR strategy. S tag-Rad9 was cloned into pIRES2-EGFP (BD Sciences Clontech, Palo Alto, CA) using XhoI and BamHI. Mutation of Ser-272 to Ala was performed using the sequence-overlap extension technique (
Rad9 Ser-272 Is Phosphorylated in Response to Multiple Genotoxins
ATM and other PIKKs phosphorylate SQ consensus sites in a variety of checkpoint and DNA repair proteins. Rad9 contains one SQ site (Ser-272), which is phosphorylated by ATM in response to double-strand DNA breaks (
). Because we wanted to examine the role of inducible Rad9 phosphorylation in response to a variety of DNA lesions, we first examined whether the replication inhibitor HU and the UV mimetic 4-NQO also induced Ser-272 phosphorylation (Fig. 1). We treated the human erythroleukemia cell line K562 with IR, HU, and 4-NQO. Immunoblotting of Rad9 immunoprecipitates with anti-phospho-Rad9 (Ser-272) revealed that, like IR, both HU and 4-NQO also induced Rad9 phosphorylation.
Rad9 Ser-272 Phosphorylation Is Not Required for Genotoxin-induced Chromatin Binding
Earlier studies showed that genotoxin-induced Rad9 chromatin binding is independent of ATR (
). However, Rad9 is also phosphorylated by ATM on Ser-272, suggesting that ATM or other unidentified PIKKs might affect the interaction of the 9-1-1 complex with DNA in response to various genotoxic stresses. To explore this possibility, we mutated Ser-272 to Ala (S tag-Rad9-S272A) in epitope-tagged Rad9 (S tag-Rad9) and expressed wild-type and mutant S tag-Rad9 in K562 cells. Precipitation of S tag-Rad9 and immunoblotting for Hus1, Rad1, and Rad17 revealed that Ser-272 phosphorylation did not affect the interaction of Rad9 with Hus1 or Rad1, the other members of the heteromeric 9-1-1 clamp, or with Rad17, the putative clamp loader for the 9-1-1 complex (Fig.2A). To assess whether Ser-272 phosphorylation regulated genotoxin-induced chromatin binding, we expressed wild-type S tag-Rad9 and S tag-Rad9-S272A and assessed genotoxin-induced Rad9 chromatin binding by examining the chromatin-bound Rad9 fraction after genotoxin treatment (Fig.2B). Wild-type S tag-Rad9 and S tag-Rad9-S272A were inducibly bound to DNA following treatment with IR, HU, and 4-NQO. However, in some experiments, S tag-Rad9-S272A exhibited slightly higher levels of chromatin binding in the absence of exogenous genotoxins, suggesting that expression of this mutant might generate low levels of genotoxic stress. Collectively, these results demonstrate that Ser-272 phosphorylation does not influence the interaction of Rad9 with its binding partners or conversion of Rad9 to a chromatin-bound form, suggesting that Ser-272 phosphorylation may regulate signaling events that lie downstream of Rad9 chromatin binding.
PIKK Catalytic Activity Is Not Required for 9-1-1 Chromatin Binding
We have previously observed that chromatin-bound Rad9 is enriched in the inducibly hyperphosphorylated form (
). Interestingly, comparison of the chromatin-bound forms of Rad9-S272A revealed that the Rad9-S272A mutant undergoes genotoxin-induced mobility shifts (see Fig.2B), suggesting that additional inducible posttranslational modifications occur on Rad9. We reasoned that this inducible posttranslational modification would be dependent upon PIKK activity. Thus, we examined whether the PIKK inhibitors wortmannin and caffeine would block the additional genotoxin-inducible modifications. Wortmannin preferentially inhibits DNA-PK and ATM, although it also inhibits ATR less potently (
). Each drug alone was unable to completely block Rad9 phosphorylation induced by HU, IR, and 4-NQO (data not shown), possibly because Rad9 can be phosphorylated by a number of different PIKKs. To simultaneously inhibit ATR, ATM, and DNA-PK, K562 cells were preincubated with a combination of wortmannin and caffeine. Cells were then treated with IR, HU, or 4-NQO. To assess the effectiveness of the drug mixture in blocking checkpoint signaling, we examined Chk1 and Chk2 activation by immunoblotting with anti-phospho-specific antibodies that recognize activated Chk1 and Chk2 (Fig.3A). All three genotoxins triggered Chk1 activation, whereas only IR and 4-NQO induced Chk2 activation. Wortmannin and caffeine pretreatment blocked Chk1 and Chk2 activation, demonstrating that PIKK activity was fully inhibited. We then examined whether PIKK inhibition affected Rad9 chromatin binding in response to IR, HU, and UV. Rad9 was extracted sequentially with low and high salt buffers to separate unbound from chromatin-bound Rad9. The chromatin-bound fractions were then analyzed by immunoblotting (Fig. 3B). In control cells, there was a small amount of Rad9 bound to chromatin in the absence of genotoxins (the level of background binding varies among experiments). As expected, all three genotoxins triggered additional Rad9 chromatin binding, with the most modified form heavily induced. Pretreatment with the PIKK inhibitor drug combination completely blocked the genotoxin-induced Rad9 mobility shifts, demonstrating that the Rad9 modifications required PIKK activity. Despite complete inhibition of the inducible Rad9 mobility shifts by caffeine and wortmannin pretreatment, genotoxin-induced Rad9 chromatin binding was induced by all three genotoxins.
Rad1 is also phosphorylated in response to genotoxins, and this phosphorylation is detected as reduced mobility when analyzed by SDS-PAGE (
). Thus, we also asked whether Rad1-inducible phosphorylation was required for Rad1 chromatin binding (Fig.3C). As was observed with Rad9, wortmannin and caffeine pretreatment blocked Rad1 phosphorylation. However, Rad1 chromatin binding was not affected by PIKK inhibition. Collectively, the results presented in Figs. 2 and 3 indicate that genotoxin-induced phosphorylation of 9-1-1 complex subunits is not required for 9-1-1 chromatin binding.
9-1-1 Chromatin Binding Does Not Require DNA Replication
The results presented above indicate that 9-1-1 chromatin binding is independent of PIKK activity. However, a recent study demonstrated that inhibition of DNA replication induced Hus1 binding in a DNA polymerase α-dependent manner, demonstrating that the Rad17-RFC clamp-loading complex, which loads the 9-1-1 clamp, might be recruited to stalled replication forks (
). Collectively, these results suggest that the upstream sensor for DNA damage is a stalled replication fork. Once stalled, the DNA replication machinery is postulated to recruit Rad17-RFC to load the 9-1-1 complex, which, coupled with ATR recruitment, activates the checkpoint signaling cascade. Because DNA lesions physically block replication fork progression, these results raise the possibility that DNA lesions are first sensed by the moving replication fork. To address whether DNA lesion-induced 9-1-1 chromatin binding was dependent upon DNA replication, we determined whether genotoxins triggered Rad9 chromatin binding in cell populations deficient in S-phase cells. We generated a G1-enriched cell population by depriving the interleukin-3-dependent myeloid cell line 32D/Bcl-XL of interleukin-3. Comparison of the exponentially growing and factor-deprived cell populations revealed that the exponentially growing population contained 27% S-phase cells and that the factor-deprived population contained 2% S-phase cells. Both cell populations were treated with IR, HU, or 4-NQO and analyzed for Rad9 chromatin binding (Fig. 4). The replication inhibitor HU triggered Rad9 chromatin binding in the exponentially growing cells. However, because the factor-deprived cells were in G1, HU did not induce Rad9 chromatin binding or Rad9 modification, confirming that very few cells were in S-phase in the factor-deprived population. In contrast, both IR and 4-NQO converted Rad9 to the chromatin-bound form in the factor-deprived population, although Rad9 chromatin binding was slightly less efficient. Thus, these results demonstrate that 9-1-1 chromatin binding does not require stalled replication forks in the face of other types of DNA damage, although they contribute in the case of HU-induced replicative stress.
Rad9, Hus1, Rad1, and Rad17 are key participants in the complex checkpoint signaling pathways activated by genotoxins. Although an elegant biochemical model has been presented describing the functions of these proteins, it remains unclear how these protein complexes are regulated (
). In the present studies we examined the roles of other checkpoint signaling proteins and the DNA replication apparatus in the binding of the 9-1-1 complex to DNA in response to a panel of genotoxic agents. Our studies demonstrated that 9-1-1 chromatin binding does not require PIKK catalytic activity, nor does it require DNA replication. Taken together, these results suggest that the loading of the 9-1-1 complex onto DNA is a proximal event in the activation of the checkpoint signaling pathways.
Previous work showed that IR induced Rad9 Ser-272 phosphorylation in an ATM-dependent manner, and this phosphorylation has been implicated in the G1-S phase checkpoint (
). Thus, we asked whether this phosphorylation event might regulate Rad9 by affecting its interaction with its binding partners or with DNA. Modification of Ser-272 did not alter the interaction of Rad9 with Hus1, Rad1, or Rad17. Moreover, Ser-272 phosphorylation had no effect on 9-1-1 chromatin binding in response to IR or other genotoxins. Taken in context with the studies demonstrating a role for Ser-272 phosphorylation in checkpoint activation, our studies suggest that this phosphorylation event may regulate a downstream, Rad9-dependent signaling event.
We also identified additional genotoxin-induced post-translational modifications on Rad9. Although these modifications are most likely due to phosphorylation, we did not attempt to verify this possibility because Rad9 is heavily constitutively phosphorylated, which induces a dramatic mobility shift on SDS-PAGE (
), loss of ATR may not disrupt all PIKK-dependent signals impinging upon the 9-1-1 complex. Thus, we used a complementary pharmacological approach to address the roles of the PIKKs in 9-1-1 chromatin binding in mammalian cells. These results demonstrated that the PIKK inhibitor mixture, which completely blocked genotoxin-induced Chk1, Chk2, Rad9, and Rad1 phosphorylation, did not block 9-1-1 chromatin binding, demonstrating that the clamp-loading reaction does not require PIKK activity.
A recent report demonstrated that Hus1 chromatin binding in response to replication inhibition required DNA polymerase α, suggesting that stalled replication forks signal the recruitment of the 9-1-1 checkpoint complex (
). Because many DNA lesions induce 9-1-1 chromatin binding, and these same lesions also slow or block replication, it was possible that stalled replication forks are the sole signal for DNA lesion-induced 9-1-1 chromatin binding. In such a scenario, the DNA replication complex may be the sensor that triggers 9-1-1 chromatin binding. However, the fact that 9-1-1 chromatin binding occurred in cell populations that were nearly devoid of S-phase cells demonstrates that Rad17-RFC responds to other DNA structures, possibly recognizing DNA lesions directly.
The results presented here demonstrate that genotoxin-induced 9-1-1 chromatin binding does not require PIKK catalytic activity or DNA replication forks, suggesting that, in the absence of replication, Rad17 may directly recognize DNA lesions produced by genotoxins. Rad17-RFC then loads the 9-1-1 complex onto the DNA where Rad9 is further phosphorylated by PIKK family members that are independently recruited to the DNA lesions. PIKK-dependent phosphorylation of chromatin-bound Rad9 and Rad1 then potentially orchestrates the activation of Chk1 and other downstream checkpoint signaling events.
We thank Drs. Jann Sarkaria, Junjie Chen, and Scott Kaufmann for thoughtful advice throughout the course of these studies. We thank Edgardo Parrilla and Scott Kaufmann for critical reading of the manuscript, and Wanda Rhodes for manuscript preparation. We also thank Dr. J. Chen for the anti-phospho-Chk2 (Thr-68) antiserum, Dr. F. Quelle for the 32D cell line, Dr. E. Lee for the anti-phospho-Rad9 antibody, and Dr. R Schultz for the wortmannin.