Retention of the human Rad9 checkpoint complex in extraction-resistant nuclear complexes after DNA damage.

Studies in yeasts and mammals have identified many genes important for DNA damage-induced checkpoint activation, including Rad9, Hus1, and Rad1; however, the functions of these gene products are unknown. In this study we show by immunolocalization that human Rad9 (hRad9) is localized exclusively in the nucleus. However, hRad9 was readily released from the nucleus into the soluble extract upon biochemical fractionation of un-irradiated cells. In contrast, DNA damage promptly converted hRad9 to an extraction-resistant form that was retained at discrete sites within the nucleus. Conversion of hRad9 to the extraction-resistant nuclear form occurred in response to diverse DNA-damaging agents and the replication inhibitor hydroxyurea but not other cytotoxic stimuli. Additionally, extraction-resistant hRad9 interacted with its binding partners, hHus1 and an inducibly phosphorylated form of hRad1. Thus, these studies demonstrate that hRad9 is a nuclear protein that becomes more firmly anchored to nuclear components after DNA damage, consistent with a proximal function in DNA damage-activated checkpoint signaling pathways.

In eukaryotes, DNA damage activates complex cellular responses that initiate DNA repair and slow or block progression through the cell cycle (reviewed in Refs. [1][2][3][4][5][6]. Activation of cell cycle arrest is mediated by the checkpoint signaling machinery. Many of the genes controlling checkpoint activation were identified by genetic studies in the yeasts Saccharomyces cerevisiae and Schizosaccharomyces pombe. Recently, homologs of the yeast checkpoint genes were identified in higher eukaryotes, suggesting that much of the checkpoint machinery is highly conserved (7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17)(18)(19)(20)(21)(22)(23). Detailed biochemical and genetic studies in both mammals and yeasts provide a working model for checkpoint activation. Central to this model are the phosphatidylinositol 3-kinase-related kinases (PIKKs) 1 (1,3,4,24,25). Human cells lacking the PIKK ataxia telangiectasia mutated, the product of the ATM gene, have global checkpoint defects and are extremely sensitive to ionizing radiation (reviewed in Ref. 26). Likewise, S. pombe lacking the PIKK spRad3 and S. cerevisiae lacking scMec1 have similar phenotypes; they cannot block cell cycle progression after DNA damage and are exquisitely sensitive to diverse genotoxic agents (27,28). The PIKKs function as signal transducers that relay activating signals to downstream effector protein kinases, including spChk1 and spCds1 in S. pombe (18) and hChk1 and hChk2 (hCds1) in humans (20,22,23). Although hChk1 and hChk2 have different functions, one common downstream target is the cell-cycle phosphatase Cdc25 (22,23,29,30). Phosphorylation of Cdc25 inhibits its activity (21,31), and phosphorylated Cdc25 is sequestered in the cytoplasm (32,33), preventing it from activating cyclin B/Cdc2. Thus, the PIKKs function as signal transducers that relay activating signals to effector protein kinases, which then block the G 2 /M transition.
The PIKKs are not the only components of the DNA damageactivated signaling pathway, however. In S. pombe, the checkpoint proteins spRad1, spHus1, and spRad9 are also essential for spChk1 activation (34) and cell cycle arrest (35,36) following DNA damage. Like the PIKKs, human homologs of all three proteins have also been identified (hRad1, hHus1, and hRad9), suggesting that the non-protein kinase components of the signaling pathway are also conserved (7)(8)(9)(10)(11)(12)(13)(14). Epistasis experiments in S. pombe and S. cerevisiae indicate that spRad1 (scRad17) and spRad9 (scDdc1) function in the same pathway (5,6). In agreement with the genetic data, biochemical analyses revealed that both human and S. pombe Rad1, Hus1, and Rad9 interact, further indicating that these proteins cooperate as a complex (10,(37)(38)(39). Sequence analyses of the mammalian and yeast proteins have provided a few clues regarding potential function. Rad1 has homology with Ustilago maydis Rec1 (40), which is a checkpoint protein and a 3Ј-5Ј exonuclease, suggesting that Rad1 may participate in DNA metabolic events that are required for checkpoint activation. However, it is unclear whether human or S. pombe Rad1 also possesses nuclease activity (12,14,17). In addition, recombinant hRad9 has recently been shown to possess 3Ј-5Ј exonuclease activity (41). In contrast, Rad1, hRad9, and hHus1 are all predicted to have structural homology with the sliding clamp protein proliferating cell nuclear antigen (PCNA) (39,42), which encircles DNA and tethers DNA polymerase ␦ to the DNA. These predicted structural similarities raise the intriguing possibility that Rad1, Rad9, and Hus1 may form clamp-like structures that localize to DNA in response to damage. Taken together, these observations have given rise to a tentative model in which Rad1, Rad9, and Hus1 are early participants in a DNA damage-signaling pathway, with the proteins possibly acting as sensors that scan the genome for damage (5,6) This hypothesis, however, has not been experimentally validated, and the biochemical functions of hRad9 and its binding partners, hRad1 and hHus1, are unknown.
In the present study we addressed the biochemical role of hRad9 by analyzing its cellular location, its interactions with DNA after damage, and its interactions with other checkpoint proteins. We show that in response to DNA damage the hRad9⅐hHus1⅐hRad1 checkpoint complex redistributes into a less extractable, chromatin-bound form, suggesting that these proteins have a proximal function in the DNA damage-signaling pathway.

EXPERIMENTAL PROCEDURES
Antibodies-Rabbit antisera that recognize hRad9, hHus1, hRad1, and hRad17 have been described previously (38). hRad9 antibodies were affinity-purified using hexahistidine-tagged hRad9 as an affinity matrix. Insoluble, bacterially expressed His 6 -hRad9 was purified from bacterial lysates using standard techniques. The purified protein was fractionated by SDS-PAGE and transferred to polyvinylidene difluoride membranes. The membranes were stained with Ponceau S, and the His 6 -hRad9 band was excised. The strips were blocked for 1 h with 2% bovine serum albumin in TBS (10 mM Tris, pH 7.4, 150 mM NaCl) containing 0.05% Tween 20. The strips were then incubated overnight with rabbit antiserum and washed with TBS containing 0.05% Tween 20 and 1 M NaCl. Anti-hRad9 antibodies were eluted by incubation in 100 mM glycine, pH 2.3, for 5 min. The eluate was then immediately neutralized with 1 M Tris-HCl, pH 8.0. Monoclonal antibodies against hRad9 were generated by the Mayo monoclonal antibody core facility by immunizing BALB/c mice with His 6 -hRad9.
Cell A, K562 cells were treated with diluent, 50 Gy of IR, or 10 mM hydroxyurea. Cells were permeabilized and stained overnight with affinity-purified anti-hRad9 followed by fluorescein-conjugated donkey anti-rabbit antibody. After washing to remove unbound antibodies, DNA was stained with Hoechst 33342, and the cells were imaged by confocal laser-scanning microscopy. B, K562 cells were lysed in high salt nuclear extraction buffer. Lysates were separated by SDS-PAGE (10% gel) and immunoblotted with affinity-purified hRad9.

FIG. 2. hRad9 is released from nuclei of permeabilized cells.
Control and irradiated (50 Gy) K562 cells were incubated for the indicated times. A portion of each cell sample was lysed to prepare total cell lysates. The remainder of the cells were lysed and serially centrifuged to generate pellets enriched in nuclei and mitochondria. The remaining particulate fractions were collected with a 105,000 ϫ g centrifugation (lysosomes and microsomes), and the S100 soluble supernatant was also analyzed. Cell-equivalent volumes were electrophoresed and immunoblotted for hRad9, the nuclear marker B23, and the mitochondrial marker cytochrome c oxidase.
FIG. 3. hRad9 is converted to a less extractable form after DNA damage. A, K562 cells were left untreated (control) or irradiated (50 Gy). One hour later the cells were permeabilized in digitonin containing low salt buffer and centrifuged to generate a soluble supernatant (low salt extract) and a pellet containing the nuclei. The nuclei were then washed in low salt extraction buffer (wash) and extracted with a high salt nuclear extraction buffer and centrifuged (HS). The nuclear remnant was then boiled in 2ϫ SDS-PAGE sample buffer to solubilize non-extracted proteins (Nuclear remnant). Cell-equivalent volumes of each extract were fractionated by SDS-PAGE and immunoblotted with affinity-purified anti-hRad9 rabbit antibodies. B, K562 and 293 cells were irradiated with the indicated doses of ␥-irradiation. One hour later low and high salt extracts were prepared. Lysates derived from cellequivalent volumes were immunoblotted for hRad9. C, A549 and lymphoblast cells were left untreated or irradiated (50 Gy). One hour later the cells were permeabilized and fractionated as described in panel B. The conversion of hRad9 to the extraction-resistant complex in shown.
were enriched in nuclei, mitochondria, lysosomes, and microsomes, exponentially growing K562 cells (1 ϫ 10 8 /assay point) were treated with nothing or 50 gray (Gy) of ionizing radiation (IR). Cells were washed in phosphate-buffer saline and resuspended in 3 ml of lysis buffer (25 mM HEPES, pH 7.5, 5 mM MgCl 2 , 1 mM EGTA, 10 g/ml pepstatin, 10 g/ml leupeptin, 1 mM Na 3 VO 4 , 10 mM ␤-glycerophosphate, 1 mM phenylmethylsulfonyl fluoride, 20 nM microcystin-LR) and incubated on ice for 20 min. All the following procedures were carried out at 4°C. Cells were then Dounce-homogenized, checked for complete lysis by trypan blue staining, and adjusted to 250 mM sucrose with 1 ml of lysis buffer containing 1 M sucrose. Cells were centrifuged at 800 ϫ g for 10 min to collect the nuclei-containing pellet. The supernatant was then serially centrifuged to collect the mitochondria-containing pellet (4000 ϫ g, 15 min) and the remaining particulate fractions (lysosome and microsome) (105,000 ϫ g, 2 h). Each pellet was solubilized in SDS-PAGE sample buffer. The 105,000 ϫ g supernatant (S100) was adjusted to 10% trichloroacetic acid and incubated on ice for 20 min, and the protein was collected by centrifugation. The acetone-washed pellet was solubilized in SDS-PAGE sample buffer. Cell-equivalent volumes from each cell fraction were separated by SDS-PAGE and immunoblotted for hRad9 and the mitochondrial marker cytochrome c oxidase (Molecular Probes) and the nuclear marker B23 (43).
Irradiation and Drug Treatments-Cells were irradiated with a 137 Cs source at a dose rate of 10.8 Gy/min. 4-Nitroquinoline oxide (4-NQO), etoposide, camptothecin, and paclitaxel were obtained from Sigma and dissolved in Me 2 SO and stored at Ϫ80°C. Hydroxyurea (Sigma) and flavopiridol were prepared fresh in PBS.
Immunoprecipitation and Phosphatase Analysis-For co-immunoprecipitation experiments, K562 cells were permeabilized and fractionated as described above; however, the nuclear extraction buffer was replaced with permeabilization buffer supplemented with 250 mM NaCl. Cell extracts were immunoprecipitated with mouse monoclonal anti-hRad9 antibodies and protein G-Sepharose or with the indicated rabbit polyclonal antisera and protein A-Sepharose (Sigma). Immunoprecipitates were washed three times in permeabilization buffer sup-plemented with 250 mM NaCl, boiled for 5 min in 2ϫ SDS-PAGE sample buffer, and fractionated by SDS-PAGE. For the -protein phosphatase experiments, hRad9 immunoprecipitates were incubated at 30°C for 30 min in 50 l of reaction buffer containing 125 units of -protein phosphatase (New England Biolabs) and 1 mM Na 3 VO 4 , as indicated. Reactions were stopped with an equal volume of 2ϫ SDS-PAGE sample buffer, and samples were separated by SDS-PAGE.
Micrococcal Nuclease Digestion-Washed nuclear pellets were resuspended in 100 l of nuclease reaction buffer (10 mM HEPES, pH 7.4, 10 mM KCl, 0.5 mM MgCl 2 , 2 mM CaCl 2 ). Varying amounts of micrococcal nuclease (Worthington Biochemical) were added to the nuclear suspensions and incubated for 30 min at 4°C. The reactions were stopped by the addition of 900 l of nuclease buffer supplemented with 5 mM EDTA to inhibit further digestion. The nuclei were recovered by centrifugation at 1000 ϫ g for 10 min, and the pellet was extracted with nuclear extraction buffer to release the remaining hRad9. The clarified supernatants and nuclear extracts were then immunoprecipitated with anti-hRad9 antisera followed by SDS-PAGE separation and immunoblotting.
Immunofluorescence-Intact and permeabilized cells were spun onto glass slides. Intact cells were fixed in 3.2% paraformaldehyde in PBS for 10 min at 23°C. Permeabilized cells were fixed in 3.2% paraformaldehyde in 10 mM potassium phosphate, pH 7.4. Fixed cells were then treated with buffer (20 mM HEPES, 50 mM NaCl, 3 mM MgCl 2 , 0.5% Triton X-100, 300 mM sucrose) for 5 min, washed twice with PBS, and blocked for 1 h in 2% bovine serum albumin in TBS containing 0.1% Tween 20. Cells were incubated overnight with affinity-purified hRad9 antibodies and washed three times with TBS containing 0.1% Tween 20. Immune complexes were then stained with fluorescein-conjugated donkey anti-rabbit antibodies. The DNA was counterstained with 2 g/ml Hoechst 33342 in PBS for 5 min. Stained cells were analyzed with a Zeiss LSM 510 laser-scanning confocal microscope.

RESULTS AND DISCUSSION
hRad9 Is a Nuclear Protein-Genetic studies in yeasts indicate that Rad9 functions early in a DNA damage-response pathway (5,6). Such a proximal function suggests that Rad9 might reside in the nucleus. To assess this possibility, we examined the cellular location of hRad9 by indirect immunofluorescence. K562 myeloid leukemia cells were stained with affinity-purified anti-hRad9 polyclonal antibodies that recog- FIG. 4. Members of the hRad9 checkpoint complex become extraction-resistant after irradiation. K562 cells were irradiated (50 Gy) and harvested at the indicated time points. Cells were then permeabilized, and cell-equivalent volumes of low and high salt nuclear extracts were analyzed by immunoblotting for hRad9, hHus1, hRad1, hRad17, and CREB-1.
FIG. 5. DNA damage triggers formation of discrete nuclear foci. K562 cells were treated with diluent, 50 Gy of IR, or 2 g/ml 4-NQO. One hour later cells were permeabilized, immobilized on glass microscope slides, fixed in paraformaldehyde, and stained with affinitypurified hRad9. Washed slides were incubated with fluorescein-conjugated donkey anti-rabbit antibody, and nuclei were counterstained with Hoechst 33342. Images were visualized with a confocal laser-scanning microscope.
nize a single band of hRad9 in whole cell lysates (Fig. 1B). The cells were then counterstained with the DNA binding dye Hoechst to visualize the nucleus. Confocal microscopy revealed that hRad9 was distributed throughout the nucleus (Fig. 1A). In agreement with this, hRad9 was also confined to the nucleus in 293 and A549 cells (data not shown).
Next, we considered the possibility that DNA damage might affect the cellular localization of hRad9. After treatment with the replication inhibitor hydroxyurea or 50 Gy of ionizing ␥-radiation (IR), cells were fixed and examined. Neither hydroxyurea nor IR altered hRad9 distribution after 1 (Fig. 1A) or 8 h (data not shown). Collectively, these results demonstrate that hRad9 is a nuclear protein and suggest that DNA damage does not regulate the cellular localization of hRad9.
To further analyze the cellular localization of hRad9, we permeabilized control, irradiated K562 cells, and separated the lysed cells by differential centrifugation into nuclei-, mitochondria-, and lysosome/microsome-enriched fractions as well as a soluble S100 fraction that contained the cytosolic contents. Surprisingly, in control cells, hRad9 was not associated with the nuclei or other particulate fractions (Fig. 2). Rather, hRad9 was released from the nuclei into the soluble fraction, suggesting that cell permeabilization readily extracted hRad9 from its normal nuclear location. In marked contrast, 1 h after irradiation, a significant fraction of hRad9 was retained in the nuclei. However, hRad9 was not found in any other particulate fraction up to 3 h after DNA damage. Taken together, these results indicate that hRad9 associates with nuclei but not other subcellular structures after DNA damage.
hRad9, hRad1, and hHus1 but Not hRad17 Are Inducibly Tethered within the Nucleus after Irradiation-Because DNA damage altered the association of hRad9 with the nuclear compartment, we further characterized this phenomenon. In Fig. 2, hRad9 was released from nuclei by denaturing and solubilizing proteins in SDS-PAGE sample buffer. Thus, we asked whether we could recover nuclei-retained hRad9 by high salt extraction in a form suitable for additional biochemical analysis. Control and irradiated (50 Gy) K562 cells were first permeabilized in a digitonin-containing low salt buffer. The permeabilized cells were centrifuged to generate a pellet, which contained the nuclei, and a soluble low salt extract (Fig. 3A, LS extract). The nuclei were washed with the low-salt buffer, and the wash was collected (Fig. 3, wash). The nuclear pellet was extracted with a high salt nuclear extraction buffer (conductivity equivalent to a 250 mM NaCl solution) to release a subset of nuclear proteins (Fig 3, high salt (HS) extract), and the insoluble nuclear remains were boiled in SDS-PAGE sample buffer (Nuclear remnant). We then analyzed cell-equivalent volumes of the low salt extract, the nuclear wash, the high salt nuclear extract, and the solubilized nuclear remains for hRad9. hRad9 was released from control cell nuclei by low salt extraction. In contrast, however, a substantial amount of hRad9 was extracted from the particulate fraction of irradiated cells with the high salt extraction buffer. Quantitation of the immunoblot demonstrated that 1 h after irradiation (50 Gy), approximately 30% of the hRad9 pool accumulated within the extractionresistant fraction. This fraction was stably bound to the nuclei (Fig. 3A), as washing the nuclei with permeabilization buffer (wash) released little additional hRad9. Because no additional hRad9 was recovered when the nuclei were boiled in SDS-PAGE sample buffer, the nuclear extraction buffer quantitatively removed hRad9. Collectively, these results demonstrate that DNA damage converts a portion of the cellular hRad9 into a less extractable nuclear form that could then be recovered by high salt extraction.
We then explored the dose of IR required to convert hRad9 into the extraction-resistant nuclear form. K562 cells and 293 cells were irradiated with the indicated doses, permeabilized, and extracted 1 h later (Fig. 3B). hRad9 was converted into the less extractable form after a 5-Gy dose in both cell lines, and it was retained in greater amounts, proportional to dose, up to at least 50 Gy. Thus, these results indicate that DNA damageinducible conversion of hRad9 to a less extractable form occurs in response to modest doses of ␥-radiation.
We also observed DNA damage-induced conversion of hRad9 to extraction-resistant complexes in human non-small lung carcinoma A549 cells and normal lymphoblasts (Fig. 3). Additionally, SV40-transformed GM847 fibroblasts, Jurkat T cells, and HCT-116 colon carcinoma cells also retained hRad9 in the nucleus after DNA damage (data not shown). Collectively, these results demonstrate that this DNA damage-inducible event occurs in numerous immortal and transformed epithelial, fibroblast, and hemopoietic human cell lines.
To determine whether other nuclear proteins were also converted to less extractable forms after DNA damage, we examined three other checkpoint proteins and, as a control, the transcription factor CREB-1. When control cells were permeabilized, hHus1 and hRad1, the binding partners for hRad9 (37)(38)(39), were found predominantly in the released extract (Fig.  4). However, after irradiation, hHus1 was also converted to a less extractable form with a similar time course. Similarly, a slower migrating, anti-hRad1-reactive band was also rapidly retained after DNA damage, suggesting that IR induces a post-translational modification of this checkpoint protein (see Fig. 7). We also examined the nuclear checkpoint protein hRad17, which, like hRad9, also leaked out of the nuclei in permeabilized cells. However, unlike hRad9, hRad17 was not inducibly retained by DNA damage. As predicted for a nuclear transcription factor, CREB-1 remained associated with the nuclear pellet after cell permeabilization, and the protein was extracted in the high salt nuclear extraction buffer. Taken together, these results indicate that hRad9, along with its binding partners hHus1 and hRad1 is rapidly and selectively converted to a less extractable nuclear form after irradiation.
Extraction-resistant hRad9 Accumulates in Nuclear Foci after DNA Damage-Data presented in Figs. 2 and 3 suggested that extraction-resistant hRad9 might be localized within the nucleus. To demonstrate this directly, we treated K562 cells with IR and the DNA-damaging agent 4-NQO, which also potently converts hRad9 to a less extractable form (see Fig. 8). However, before staining for hRad9, we permeabilized cells to extract the non-bound pool of hRad9. The permeabilized cells were then stained for hRad9 (Fig. 5). In agreement with the biochemical fractionation studies, hRad9 was lost from the nuclei of cells that had not been treated with DNA-damaging agents. In contrast, in cells treated with 50 Gy of IR or 4-NQO 1 h before permeabilization, hRad9 was retained in extractionresistant foci. Taken together, these results suggest that hRad9 is loosely tethered and readily extracted from the nucleus by FIG. 6. Micrococcal nuclease releases extraction-resistant hRad9. Nuclei were isolated from cells 1 h after irradiation (50 Gy). Washed nuclei were treated with the indicated amounts of micrococcal nuclease (MNase) for 30 min on ice. The nuclei were then centrifuged, and the supernatant was collected for analysis (supernatant). The nuclear pellet was then extracted in high salt nuclear extraction buffer (pellet). Cell-equivalent volumes of the nuclease-released fraction (supernatant) and the nuclear extract (pellet) were immunoblotted for hRad9. permeabilization in cells not treated with DNA-damaging agents. Upon DNA damage, extraction-resistant hRad9 complexes assemble in discrete foci that require increased ionic strength to disrupt.
Nuclei-retained hRad9 Associates with the Chromatin-The conversion of hRad9 into less-extractable forms might represent DNA damage-inducible hRad9 association with DNA or chromatin. To address this question, we tested whether nucleases could detach extraction-resistant hRad9 from nuclei that were isolated from irradiated cells. Results of this analysis revealed that micrococcal nuclease (Fig. 6) and DNase I (data not shown) released hRad9 from the nuclei. The nuclease susceptibility of nuclei-retained hRad9 suggests that hRad9 associates with DNA, either directly or indirectly, after DNA damage.
Nuclei-retained hRad9 Associates with hHus1 and Phosphorylated hRad1-Results in Fig. 3 showed that hRad9, hHus1, and modified hRad1 were all retained in K562 cell nuclei after irradiation. In view of the data that hRad9 interacted with hRad1 and hHus1 (37,38), we asked whether retained hRad9 also interacted with hRad1 and hHus1. One hour after irradiation, K562 cells were permeabilized and fractionated. hRad9 was immunoprecipitated from the low and high salt extracts, and the immunoprecipitates were immunoblotted for hHus1 and hRad1. Figs. 7, A and B, show that hRad9 extracted by the low and high salt extraction buffers was associated with hHus1 and hRad1. Interestingly, nuclei-bound hRad9 associated with a slower migrating form of hRad1, which resembled the slower migrating form of hRad1 that was retained in the nucleus after DNA damage (Fig. 4). To assess the possibility that the hRad9-associated hRad1 represented phosphorylated hRad1, hRad9 immunoprecipitates were treated with -protein phosphatase. Immunoblotting revealed that the phosphatase converted the slower migrating form of hRad1 to a form that exhibited mobility identical to that of unmodified hRad1 (Fig. 7C). This demonstrates that, like hRad9, hRad1 is inducibly phosphorylated following DNA damage. Additionally, phospho-hRad1 is significantly enriched in the less extractable hRad1 nuclear fraction. Collectively, these results demonstrate that the hRad9⅐hHus1⅐hRad1 checkpoint complex is converted to an extraction-resistant nuclear aggregate by DNA damage.
hRad9 Nuclear Retention Occurs in Response to DNA Damage but Not Other Cytotoxic Agents-Even 24 h after irradiation of K562 cells with 50 Gy of IR the cells were not morphologically apoptotic nor had they activated caspases (data not shown), suggesting that conversion of hRad9 to the extractionresistant form was not an apoptotic response. To further test this possibility, we assessed hRad9 nuclear retention in cells treated with drugs that are cytotoxic but do not directly dam- FIG. 7. Nuclei-retained hRad9 is associated with hHus1 and phospho-hRad1. Control (Ϫ) and irradiated (50 Gy) K562 cells were cultured for 1 h. Cells were then permeabilized, and low (LS) and high salt (HS) nuclear extracts were prepared. A, hRad1 immunoprecipitates (IP) were immunoblotted (IB) with hRad1 (left panel) or hRad9 (right panel). hRad9 immunoprecipitates were immunoblotted with hRad1 (center panel). B, hHus1 immunoprecipitates were immunoblotted with hHus1 (left panel) or hRad9 (right panel). hRad9 immunoprecipitates were immunoblotted with hHus1 (center panel). C, control and irradiated (50 Gy) K562 cells were harvested 1 h later. High salt extracts were prepared, and hRad1 was immunoprecipitated. One set of parallel hRad1 immunoprecipitates was treated with 125 units of -protein phosphatase (PPase), and another set was treated with -protein phosphatase and 1 mM sodium orthovanadate (VO 4 ). Immunoprecipitates were then immunoblotted for hRad1.
FIG. 8. DNA damage but not cell death triggers formation of nuclei-retained hRad9. A, K562 cells were irradiated (50 Gy) or treated with 1 M flavopiridol or 1 M paclitaxel. One and 24 h later cells were permeabilized, and low and high salt extracts were prepared. To correct for cell death and lack of cell proliferation, protein concentrations were determined, and equal amounts of protein were immunoprecipitated with an anti-hRad9 monoclonal antibody. Immunoprecipitates were then immunoblotted for hRad9. B, K562 cells were treated with 2 g/ml 4-NQO, 50 Gy of IR, 1 M VP-16, 10 mM hydroxyurea (HU), or 1 M camptothecin (CPT) for 2 h. age DNA. Cells were irradiated or treated with the microtubule disrupter paclitaxel and the Cdk inhibitor flavopiridol for 1 and 24 h using drug concentrations that reduce clonogenicity of K562 cells by more than 95 percent (data not shown). In contrast to IR, which transformed hRad9 into a less extractable form that persisted even 24 h later (Fig. 8A), a 1-h exposure to flavopiridol did not elicit nuclear association. Even at 24 h, when cells were morphologically apoptotic, hRad9 was not retained in nuclei. Likewise, paclitaxel did not stimulate hRad9 nuclear retention after 1 or 24 h of drug exposure. Therefore, these results demonstrate that the increased association of hRad9 with nuclei requires DNA damage and does not occur in response to other non-DNA-damaging cytotoxic agents.
Diverse DNA-damaging Agents Stimulate hRad9 Nuclear Retention-Ionizing radiation generates a variety of DNA lesions including large numbers of DNA strand breaks and oxidatively damaged bases. To address whether other DNA lesions might trigger nuclear retention, we treated K562 cells with a panel of DNA-damaging agents for 2 h (Fig. 8B). hRad9 was strongly retained after treatment with 4-NQO, which generates bulky base adducts and single-strand DNA breaks. The topoisomerase II poison etoposide (VP-16) and the topoisomerase I poison camptothecin, which trap their target enzymes in covalent complexes with DNA and create double-strand breaks when adjacent replication or transcription complexes collide with these covalent complexes, induced potent hRad9 nuclear retention. Hydoxyurea, which inhibits ribonucleotide reductase and causes nucleotide depletion, thereby stalling replication forks on the DNA, also provoked nuclear retention. Collectively, these results suggest that diverse types of DNA damage convert hRad9 to a less extractable nuclear complex.
Although genetically identified as key components of the checkpoint response in S. pombe, the biochemical functions of Rad9, Hus1, and Rad1 are not understood. Moreover, the subcellular localization of hRad9 is currently under debate. Our data as well as the data of St. Onge et al. (37) demonstrate that endogenous hRad9 is a nuclear protein. In contrast, however, Komatsu et al. (44) report that transiently expressed, epitopetagged hRad9 is localized in both the nucleus and the cytoplasm. This group also demonstrates that transiently expressed hRad9 exited the nucleus after DNA damage and redistributed to the mitochondria, where it interacted with Bcl-2 and promoted apoptosis. In the present study we found no evidence for extranuclear hRad9 after DNA damage when staining intact K562 or other cells. Also, we did not detect Bcl-2, Bcl-X L , or Mcl-1 in anti-hRad9 immunoprecipitates nor could we detect hRad9 in Bcl-2, Bcl-X L , or Mcl-1 immunoprecipitates (data not shown). The reasons for these discrepancies are not clear but they may relate to the differential behavior of endogenous and transiently expressed, epitope-tagged hRad9.
Alternatively, based on genetic and biochemical analyses, several models have proposed that Rad9 complex members are part of a surveillance mechanism that scans the genome for DNA damage. The data presented here expand upon and support such a model. First, hRad9 is a nuclear protein that is readily extracted from control cells, suggesting that it interacts only weakly with DNA. Second, after DNA damage, the hRad9 complex rapidly acquires new interactions, anchoring the complex to DNA. Given that Rad1, Rad9, and Hus1 have been proposed to structurally resemble PCNA-like clamps (39,42), it is tempting to speculate that conversion of the hRad9⅐hRad1⅐hHus1 complex to the less extractable form may reflect the loading of this clamp-containing complex onto sites of DNA damage. Once loaded, the checkpoint protein clamp may anchor DNA-processing proteins and other checkpoint proteins that activate the DNA damage-signaling pathway in mammalian cells. Because DNA damage converted the hRad9 complex to a less extractable form in numerous cell lines, this might be an event that occurs in all human cells after DNA damage. However, such a conclusion awaits further study.