Caspase-3-Mediated Cleavage of Rad9 During Apoptosis

Caspase-3-mediated cleavage of Rad9 during apoptosis

Michael W Lee1,2, Itaru Hirai1,3 and Hong-Gang Wang*,1,2,3

1Drug Discovery Program, H. Lee Mofﬁtt Cancer Center and Research Institute, USA; 2Department of Pharmacology & Therapeutics, University of South Florida College of Medicine, 12902 Magnolia Drive, Tampa, FL 33612, USA; 3Department of Interdisciplinary Oncology, University of South Florida College of Medicine, 12902 Magnolia Drive, Tampa, FL 33612, USA

The activation of caspases is a critical event for the distinct from necrosis.Grossly, apoptotic cells exhibit a execution phase of programmed cell death. Caspases are variety of identifiable characteristics such as membrane highly specific in their ability to activate or inhibit many blebbing, chromatin condensation, and typically, crucial proteins in the cell via cleavage. In this study, we fragmentation into membrane-enclosed vesicles report the identification of several caspase-3-like cleavage (Earnshaw et al., 1999). Biochemical changes include sites in the cell-cycle checkpoint protein Rad9. We the externalization of phosphatidylserine, degradation demonstrate that human Rad9 can be specifically cleaved of chromosomal DNA into high-molecular-weight in cells induced to enter apoptosis by both DNA damage oligonucleosomal fragments, and cleavage of specific and staurosporine treatment. Indeed, we show that human polypeptides such as PARP, ICAD, and antiapoptotic Rad9 can be effectively cleaved both in vitro and in vivo, proteins such as Bcl-2 and Bcl-XL. which can be inhibited by either a pan-caspase inhibitor or Many stimuli that trigger apoptosis without ligating a caspase-3-specific inhibitor. Additionally, no cleavage death receptors cause mitochondria to release the of Rad9 can be seen in the caspase-3-deficient cell line electron transport protein cytochrome c to the cytosol MCF-7. Site-directed mutagenesis of three of the most (Green and Reed, 1998).Apaf-1 binds the released conserved cleavage sites dramatically abrogates cleavage cytochrome c, and then undergoes a nucleotide-depen- of Rad9 by caspase-3 in vitro, and in intact cells after dent conformational change that allows the binding of DNA damage. Expression of the cleavage-resistant procaspase-9.As a consequence of this binding, mutant Rad9 DDD/AAA appears to protect the cell from procaspase-9 is cleaved to active caspase-9, which in DNA damage-induced apoptosis. Immunofluorescence turn leads to activation of the downstream effector studies of Rad9 localization before and after induction caspase-3, -6, and -7.These effector caspases then of apoptosis showa translocation of Rad9 from the directly cause the morphological changes associated nucleus to the cytosol, concomitant to the appearance of with apoptosis by specifically cleaving ‘death substrates’ apoptotic morphology. Furthermore, analysis of a trun- (Earnshaw et al., 1999). Caspases cleave a limited cated Rad9 mutant that corresponds to a putative N- number of cellular proteins, and the process is one of terminal cleavage fragment shows that the N-terminal limited proteolysis (Salvesen and Dixit, 1997).Owing to portion of Rad9 localizes in the cytosol, binds to Bcl-XL, their substrate specificity, caspase-mediated cleavage and induces apoptosis. These results support a dual role can result in activation or inactivation of their protein for cleavage of Rad9: (1) the liberation and translocation substrates, but never degradation (Salvesen and Dixit, of the BH3 domain-containing N-terminus of Rad9 to the 1997). cytosol, as a means of promoting apoptosis via antagon- The Bcl-2 family is comprised of well over a dozen ism of Bcl-XL, and (2) the disruption of the Rad9-Rad1- proteins, which have been classified into three functional Hus1 DNA damage checkpoint complex. groups (Wang and Reed, 1998).Members of the first Oncogene (2003) 22, 6340–6346. doi:10.1038/sj.onc.1206729 group such as Bcl-2 and Bcl-XL contain four conserved Bcl-2 homology (BH) domains (BH1–BH4), and possess Keywords: caspase-3; Rad9; DNA damage; cell-cycle antiapoptotic activity.In contrast, group II consists of checkpoint; apoptosis proapoptotic Bcl-2 family members including Bax and Bak, which have a similar overall structure to group I proteins but lack the N-terminal BH4 domain.Group III consists of a large and diverse collection of Introduction proapoptotic proteins whose only common feature is the presence of the BH3 domain.The Bcl-2 family Apoptosis is an evolutionarily conserved process of cell of proteins appears to control the ‘decision’ step of suicide that is both morphologically and biochemically apoptosis, determining whether certain caspases will or will not become activated (Reed, 1998).Antiapoptotic *Correspondence: H-G Wang, Drug Discovery Program, H.Lee members of the Bcl-2 family tend to prevent activation Moffitt Cancer Center and Research Institute, 12902 Magnolia Drive, of these terminal effector proteases by blocking the Tampa FL 33612, USA; E-mail: wanghg@moffitt.usf.edu Received 13 February 2003; revised 17 April 2003; accepted 22 April release of apoptogenic molecules such as cytochrome c 2003 from the mitochondria.Conversely, proapoptotic Caspase-3 cleavage of Rad9 MW Lee et al 6341 Bcl-2 family members facilitate caspase activation by promoting the release of cytochrome c from the mitochondria. In response to DNA damage, which can arise spontaneously from replication errors or from exogen- ous agents, a rapid-surveillance signaling cascade called the DNA damage checkpoint is initiated (Caspari and Carr, 2002).Activation of this cascade results in cell- cycle arrest, which provides time for the damage to be repaired.Deregulation of this checkpoint can lead to genomic instability that is associated with genetic diseases such as cancer (Hartwell and Kastan, 1994). The Rad family of checkpoint proteins have emerged not only as critical mediators of this damage response pathway but also as putative apical sensors in the DNA damage checkpoint.Indeed, three members of the Rad Figure 1 Human Rad9 is degraded after treatment with high-dose IR.U937 cells were treated with various doses of IR and cultured family Rad9, Rad1, and Hus1 have been shown to for 5 h after treatment.Lysate was prepared and subjected to interact with each other in a manner similar to the SDS—PAGE, and transferred to nitrocellulose filters for analysis of proliferating cell nuclear antigen (PCNA) (O’Connell Rad9 expression using a polyclonal anti-Rad9 antibody (top).The et al., 2000). Furthermore, a fourth Rad family member, same blot was then stripped and reprobed with anti-caspase-3 Rad17, bears structural similarity to replication factor C antibody (bottom) (RFC), the clamp loader which places PCNA onto DNA (Tsurimoto, 1999; Venclovas and Thelen, 2000).It et al., 2000b; St Onge et al., 2001), when resolved on is believed that, in times of damage, these four proteins SDS–polyacrylamide gel (SDS–PAGE), Rad9 migrates form a putative DNA damage sensor, which encircles as several distinct species ranging from 42 to 60 kDa damaged DNA and relays the damage signal to down- even in the absence of genotoxic stress.These different stream effectors (Zhou and Elledge, 2000; Wahl and bands are likely the result of phosphorylation, and can Carr, 2001).In addition, it was shown that Rad9 could be seen in both treated and untreated cells (St.Onge interact with Bcl-2 and Bcl-XL via a conserved BH3 et al., 1999, 2001; Volkmer and Karnitz, 1999; Komatsu domain located in the N-terminus of the Rad9 protein et al., 2000b; Chen et al., 2001). As shown in Figure 1, to antagonize the protective function of these proteins, for example, Rad9 appeared as three major species, promoting apoptosis (Komatsu et al., 2000a, b; Yoshida designated Rad9a, Rad9b, and Rad9g, on Western et al., 2002). It has also been demonstrated that Rad9 blots.Interestingly, after treating U937 cells with can be phosphorylated on Tyr-28 within the BH3 incrementally increasing doses of ionizing radiation domain by c-Abl, and that this phosphorylation (IR), we observed that the lesser modified forms (b, g) promotes the interaction of Rad9 with Bcl-XL (Yoshida of Rad9 decreased in a dose-dependent manner et al., 2002). (Figure 1).This decrease does not directly translate into In this study, we demonstrate that Rad9 possess the increases in the highly modified form (Rad9a), several potential caspase-3 cleavage sites and can be suggesting that Rad9, especially the lesser modified cleaved by caspase-3 both in vitro and in vivo.After forms (Rad9b and Rad9g), might undergo proteolysis exposure to DNA-damaging agents or staurosporine after DNA damage.Reprobing the same blot with anti- (STS), Rad9 undergoes proteolysis and cytosolic trans- caspase-3 antibody revealed that the decrease in Rad9 location, which can be blocked by caspase inhibitors. protein level was concomitant to the activation of Mutating the caspase-3 cleavage sites in Rad9 blocks caspase-3 (Figure 1).Upon analysis of the Rad9 protein cleavage in vivo, and also appears to protect the cell sequence, we noticed that human Rad9 contains several from death induced by DNA-damaging agents.Further- potential caspase-3 cleavage sites (Figure 2a).For more, Rad9DC, a mutant Rad9 that lacks the C- example, the sequence DDID (amino acids 301–304) is terminus and thus mimics a potential N-terminal quite similar to the consensus motif DXXD for caspase- cleavage fragment of Rad9, retains the ability to bind 3 cleavage (Thornberry et al., 1997). Bcl-XL, is localized to the cytosol, and promotes apoptosis. Caspase-3 but not caspase-6 cleaves Rad9 in vitro To determine whether caspase-3 can specifically cleave in vitro 35 in vitro Results Rad9 , [S]methionine-labeled tran- scribed/translated (IVT) Rad9 was incubated with either Rad9 undergoes proteolysis after DNA damage purified active caspase-3 or caspase-6 (Stennicke and Salvesen, 1997).After incubation for 1 h at 37 1C, the To further characterize the role of Rad9 in apoptosis, resultant cleavage products were resolved on SDS– the stability and modification of Rad9 were examined PAGE (15% gel) and exposed to X-ray film.As shown after treatment with genotoxic stress.Consistent with in Figure 2b, there were multiple species produced when previous results (Volkmer and Karnitz, 1999; Komatsu incubated with caspase-3, whereas incubation with

Oncogene Caspase-3 cleavage of Rad9 MW Lee et al 6342

Figure 2 Rad9 possesses several potential caspase-3 cleavage sites, and is cleaved by caspase-3 in vitro.(a) Sequence analysis of Rad9 identiﬁed one very conserved (DDID) and two moderately conserved (EEAD and SDTD) caspase-3 cleavage sites, as compared to the ideal consensus sequence (Thornberry et al., 1997).( b) IVT 35S-labeled Rad9 was incubated with either puriﬁed active caspase-3 or caspase-6 for 30 min, and analysed by SDS– PAGE.( c, d) 35S-labeled Rad9 or Rad9 mutants were incubated with ( þ ) or without (À) active caspase-3 for 30 min, and analysed by SDS–PAGE

caspase-6 produced no cleavage fragment.The fact that there are multiple cleavage fragments generated after incubation with caspase-3 did not come as a surprise for, Figure 3 IR induces Rad9 cleavage in vivo, which can be blocked by caspase inhibitors.U937 cells were treated with 50 Gy IR in the as indicated from the sequence analysis, there are several absence (a) or presence of 100 mm caspase inhibitor z-VAD-fmk (b), potential caspase cleavage sites in the Rad9 protein.To or z-DEVD-fmk (c).After incubation for the indicated times, cell further assess caspase-3 cleavage of Rad9, we generated lysates were prepared and subjected to SDS–PAGE/immunoblot two point mutants: Rad9D304A, a single-point mutant analysis with antibodies specific for Rad9 or caspase-3 in which the critical aspartic acid residue of DDID (AA: 301–304) was changed to alanine, and Rad9 DDD/ AAA, a triple-point mutant in which three aspartic acid residues of EEAD (AA: 184–187), SDTD (AA: 266– coincided with the appearance of several faster migrat- 269), and DDID (AA: 301–304) were changed to ing cleaved fragments (Figure 3a).These cleavage alanines.The 35S-labeled IVT Rad9 proteins were fragments most likely represent C-terminal fragments, incubated with or without caspase-3, and the resulting due to the fact that our Rad9 antibody recognizes the C- fragments were resolved on a 15% SDS–PAGE gel and terminal portion of the protein.Also, there was a subtle detected by autoradiography.Despite these mutations, modulation in the most highly phosphorylated form some cleavage can still be seen (Figure 2c and d), (Rad9a), although this species seems more resistant to suggesting either the presence of minor cleavage sites or IR-induced cleavage.The same blot was then stripped nonspecific proteolysis in vitro. and reprobed with anti-caspase-3 antibody to show activation of caspase-3 concomitant to the appearance of the cleaved fragments of Rad9.To confirm that Rad9 is cleaved by caspase-3 in vivo caspase-3 is the caspase responsible for Rad9 cleavage, To determine whether Rad9 is cleaved in intact cells, we U937 cells were exposed to 50 Gy IR in the presence or examined the processing of endogenous Rad9 in the absence of either a pan-caspase inhibitor (z-VAD-fmk) human myeloid leukemia cell line U937 after treatment or a caspase-3-specific inhibitor (z-DEVD-fmk).As seen with IR.U937 cells were treated with high-dose IR in Figure 3b, incubation of cells with z-VAD-fmk but (50 Gy), and analysed by Western blot for both cleavage not the DMSO control abolished the proteolysis of of Rad9 and activation of caspase-3.Consistent with the Rad9 induced by IR treatment.Similar results were results seen in Figure 1, the less-modified forms of Rad9 obtained when using the caspase-3-specific inhibitor z- (Rad9b and Rad9g) disappeared over time, and this DEVD-fmk (Figure 3c).Moreover, STS, a protein

Oncogene Caspase-3 cleavage of Rad9 MW Lee et al 6343

Figure 4 STS induces Rad9 cleavage in U937 cells but not in caspase-3-deﬁcient MCF-7 cells.Both U937 ( a) and MCF-7 cells (b) were treated with 1 mm STS for the indicated times, prior to preparation of cell lysates and SDS–PAGE/immunoblot analysis with anti-Rad9 and anti-caspase-3 antibodies

Figure 5 Mutation of caspase-3 cleavage sites in Rad9 blocks kinase C inhibitor that activates caspase-3 and induces cleavage in vivo, and protects cells from death induced by DNA- apoptosis, also caused the cleavage of Rad9 in U937 damaging agents.( a) HeLa cells were cotransfected with either cells (Figure 4a) but not the caspase-3-deficient cell line Myc-Rad9 (w/t) or Myc-Rad9 DDD/AAA mutant (3D), together MCF-7 (Figure 4b).These results, taken together, with Flag-Rad1 and HA-Hus1 or empty vectors as control.After 24 h, the cells were treated with 1.5 mm Doxorubicin for 20 h, prior indicate that Rad9 is specifically cleaved by caspase-3 to lysis and Western blot with anti-Rad9 antibody.The blot was in cells. stripped and reprobed with anti-caspase-3 polyclonal antiserum. (b) 293T cells were cotransfected with the same combinations of plasmid DNA as indicated in (a) for 24 h, and subjected to Blockade of Rad9 cleavage decreases cell death in coimmunoprecipitation with anti-Myc affinity beads (Sigma).In response to DNA-damaging agents addition to immune complexes, the whole-cell lysates (WCL) were applied directly to SDS–PAGE/immunoblot analysis.The blot was To ascertain the functional significance of Rad9 sequentially probed, striped, and reprobed with antibodies specific cleavage, we utilized the triple-point mutant Rad9 for c-Myc, Flag, and HA.*Denotes the residual signals from the DDD/AAA, which shows dramatic resistance to cas- previous immunoblotting.( c) HeLa cells were transfected as indicated above, and treated with 12.5 mg/ml etoposide for the pase-3-mediated cleavage in vitro (Figure 2d).We tested indicated times before trypan blue staining the efficacy of this mutant in HeLa cells cotransfected with either Myc-tagged wild-type or triple-mutant Rad9 DDD/AAA, together with Flag-Rad1 and HA-Hus1, Cleavage coincides with cytosolic redistribution of Rad9 and treated with 1.5 mm Doxorubicin for 20 h.As can be clearly seen in Figure 5a, the wild-type Rad9 protein is Previous findings have indicated that Rad9 can translo- cleaved quite well, whereas the Rad9 DDD/AAA cate to the cytosol after treatment with DNA-damaging mutant remains intact and unchanged as compared to agents, and that this translocation results in colocaliza- vehicle-treated cells.It is also important to note that tion of Rad9 with Bcl-2 (Komatsu et al., 2000b; Hayashi conversion of these three aspartic residues in Rad9 to et al., 2002). Since cleavage of Rad9 by caspase-3 alanines does not alter its ability to form a complex with generates N-terminal fragments that harbor the BH3 Rad1 and Hus1 (Figure 5b).Consistent with these domain, we reasoned that cleavage might contribute to results, overexpression of Rad9 DDD/AAA mutant the cytosolic translocation of Rad9.To test this revealed a significant protective effect from DNA hypothesis, we first treated the human breast cancer damage, as compared to wild-type Rad9 or control MDA-MB-468 cells with STS, and assayed for Rad9 vector-transfected cells (Figure 5c). cleavage and caspase-3 activation.As shown in

Oncogene Caspase-3 cleavage of Rad9 MW Lee et al 6344 redistribution of Rad9 after STS treatment.Immuno- ﬂuorescence analysis revealed that GFP-Rad9 was largely localized in the nucleus of untreated cells, but became a diffuse distribution throughout the cell after 2 h of STS treatment (Figure 6b).It is important to note that, after 2 h of STS treatment, the nuclear membrane was still intact, although chromatin was beginning to condense or fragment.In contrast to wild-type Rad9 (w/t), the caspase-3-resistant Rad9 mutant (DDD/ AAA) remained in the nucleus after STS treatment (Figure 6b).

The N-terminus of Rad9 interacts with Bcl-XL and induces apoptosis To further investigate the intracellular localization of the N-terminal fragment of Rad9, we generated a C- terminal truncated mutant of Rad9 that mimics an N- terminal fragment of the cleaved Rad9 protein, and is expressed as a GFP fusion protein in HeLa cells. Consistent with the previous report (Hirai and Wang, 2002), immunofluorescence microscopic analysis showed that GFP-Rad9DC was apparently distributed in the cytosol compared to the nuclear localization of GFP- Rad9 (Figure 7a).Cells transfected with the GFP control protein revealed a mostly diffuse pattern of fluorescence (Figure 7a).Coimmunoprecipitation ex- periments indicated that the N-terminal fragment of Rad9 (Rad9DC) in which the BH3 domain resides was sufficient to associate with Bcl-XL (Figure 7b), consistent with our previous results (Komatsu et al., 2000b). Moreover, expression of Rad9DC mutant in 293 cells induced more apoptosis compared to those of cells transfected with wild-type Rad9 (Figure 7c).

Discussion

Studies on substrate specificity, prodomain structure, and biological function have revealed that caspases are activated during apoptosis in a self-amplifying cascade Figure 6 Rad9 is cleaved in MDA-MB-468 cells, and translocates (Salvesen and Dixit, 1997; Zimmermann et al., 2001). from the nucleus to the cytosol after induction of apoptosis.( a) Activation of the upstream caspases such as caspase-2, m MDA-MB-468 cells were treated with 1 m STS for the designated -8, -9, and -10 by proapoptotic signals leads to times, and analysed by SDS–PAGE/immunoblot for Rad9 and caspase-3. b-Tubulin was used as a loading control.( b) MDA-MB- proteolytic activation of the downstream effector 468 cells growing on glass cover slips were transfected with GFP- caspase-3, -6, and -7.The most prevalent caspase in Rad9 (w/t) or GFP-Rad9 (DDD/AAA) for 24 h, and followed by the cell is caspase-3, and it is the one ultimately treatment with 1 mm STS for the allotted times.Cells were then responsible for the majority of the apoptotic effects fixed with 3.7% formaldehyde for 30 min, and mounted onto microscope slides with mounting media containing the nuclear dye (Zimmermann et al., 2001). Up to this point, several key DAPI, and analysed via fluorescent microscopy proteins involved in DNA damage checkpoints or repair have been identified as targets of caspase-3.These proteins include poly-(ADP-ribose) polymerase (PARP), the DNA damage kinase ATM, the DNA Figure 6a, a definite decrease in Rad9 protein level repair recombinase Rad51, and the catalytic subunit of occurred in MDA-MB-468 cells after 3 h of STS the DNA-dependent protein kinase (DNA-PKcs) (Ni- treatment, which was concomitant to caspase-3 activa- cholson, 1999).Our findings indicate that Rad9, a tion, although the cleavage fragments themselves were human protein involved in the regulation of a cell-cycle not as clearly visible as in U937 cells.Since we do not checkpoint, is a substrate of caspase-3. have antibodies available for detection of the N-terminal Previous work has provided data that support a role fragments of Rad9, we transfected MDA-MB-468 cells for Rad9, in a complex along with Rad family members with GFP-tagged Rad9 to study the intracellular Hus1 and Rad1, as a DNA damage sensor in the DNA

Oncogene Caspase-3 cleavage of Rad9 MW Lee et al 6345 Yoshida et al., 2002). Thus, the discovery that Rad9 is cleaved by caspase-3 adds a new, potentially interesting aspect to Rad9’s involvement in apoptosis.We demonstrate that Rad9 is cleaved very efficiently by caspase-3 but not caspase-6.In response to DNA-damaging agents or STS treatment, Rad9 undergoes proteolysis in intact cells to generate several small fragments.The cleavage of Rad9 in vivo coincides with cleavage and activation of caspase-3, which can be inhibited by both a pan-caspase inhibitor and a caspase-3-specific inhibitor. Moreover, by mutating the major caspase-3 cleavage sites in Rad9, we generated a cleavage-resistant mutant that, when overexpressed, helps protect cells from death induced by DNA damage, while still retaining the ability to form the Rad9-Rad1-Hus1 checkpoint complex. Unfortunately, because the presently available antibodies for Rad9 are all C-terminus-directed, we could not visualize the cleaved N-terminal fragments of endogenous Rad9 by immunoblot analysis.However, a Rad9 mutant that corresponds to a potential cleavage product can serve as a proxy fragment to allow functional analysis of caspase-3-mediated cleavage of Rad9.This C-terminal deletion mutant of Rad9 nevertheless retains the ability to bind Bcl-XL in the cytosol, and to promote apoptosis.This then poses a potential dual role for cleavage that places Rad9 at the crux of a delicate balance between cell survival and programmed cell death. Burtelow et al.(2001) showed that the N-terminus of Rad9 interacts with the C-terminus of Rad1, the N- terminus of Rad1 interacts with the C-terminus of Hus1, and the N-terminus of Hus1 interacts with the C- terminus of Rad9, thus forming a ‘doughnut’-like Figure 7 The N-terminus of Rad9 resides in the cytosol, and is heterotrimeric complex.Therefore, the action of cas- sufficient for binding to Bcl-XL and inducing apoptosis.( a) Hela cells were transiently transfected with GFP-Rad9, GFP-Rad9DC pase-3 on Rad9 could be to serve as an indirect method or control GFP by lipofectamine 2000 reagent.After 2 days, cells of halting DNA repair in times of unrepairable damage were fixed in 3.7% formaldehyde and analysed by fluorescence via disruption of the Rad9-Rad1-Hus1 complex.Indeed, microscopy.( b) 293 cells were cotransfected with Bcl-XL and Flag- many polypeptides involved in DNA replication or Rad9, Flag-Rad9DC, or control pCMV2-FLAG vector for 24 h. repair such as PARP, DNA-PKcs, RAD51, p21waf1/cip1, Cells were lysed and subjected to coimmunoprecipitation with anti- FLAG monoclonal antibody, and analysed by SDS–PAGE/ as well as the large subunit of RFC, are cleaved in immunoblot with anti-FLAG or anti-Bcl-XL polyclonal antibo- apoptotic cells (Earnshaw et al., 1999; Nicholson, 1999). dies.In addition to immunoprecipitates, the cell lysates (50 mg/lane) Thus, truncation of Rad9 would signal an abandonment were analysed directly.( c) 293 cells were transiently transfected of the ‘sensor’ role, as well as a loss of G2/M checkpoint with pcDNA3-Myc-Rad9, pcDNA3-Myc-Rad9DC or control vector pcDNA3-Myc for 2 days.Cells were stained with DAPI integrity. and counted by fluorescence microscopy for apoptosis Cleavage may also serve as a cue for translocation and activation of Rad9 as an apoptosis-promoting protein through interaction with Bcl-2/Bcl-XL.Indeed, both full-length Rad9 and Rad9DC mutant are capable of integrity checkpoint (Zhou and Elledge, 2000; Caspari binding to Bcl-XL, making a case for Rad9’s antagon- and Carr, 2002).This putative complex, as revealed by ism of the antiapoptotic Bcl-2 family proteins.In times molecular modeling analysis, bears sequence homology of DNA damage, a distinct pool of full-length Rad9 with the DNA-sliding clamp proliferating cell nuclear may function to inhibit Bcl-XL, while a separate pool antigen (PCNA).(Aravind et al., 1999; Venclovas and continues to associate with Rad1 and Hus1, and engages Thelen, 2000).The Rad9-Rad1-Hus1 complex is loca- in checkpoint control.Then, through an unknown lized to the nucleus by the Rad9 nuclear localization mechanism, when it is determined that damage is too sequence (NLS), where it then presumably initiates severe to survive or if it is energetically unfavorable to downstream signaling events for cell-cycle arrest in continue repair, Rad9 undergoes proteolysis, disrupting response to DNA damage (Hirai and Wang, 2002). the Rad9 checkpoint complex.The resulting N-terminal Another role outlined for Rad9 is as an arbiter of fragments with the conserved BH3 domain of Rad9 then apoptosis via a conserved BH3 domain in the N- translocate to the cytosol, and interact with Bcl-2 or Bcl- terminus of the Rad9 protein (Komatsu et al., 2000b; XL, promoting apoptosis.

Oncogene Caspase-3 cleavage of Rad9 MW Lee et al 6346 Materials and methods active caspase-3 or 1.5 mlof10mm active caspase-6 (Stennicke and Salvesen, 1997), and incubated at 371C for 30 min. Expression plasmids Proteins were then resolved on SDS–PAGE (15%) and detected by autoradiography. The pcDNA3-HA-Rad9, pCMV2-FLAG-Rad9, and pEGFP- C2-Rad9 have been previously described (Komatsu et al., 2000b).The single-point mutant Rad9 (D304A) was created Immunoprecipitation/immunoblot assays using a two-step polymerase chain reaction (PCR) method Cell lysates were prepared in NP-40 lysis buffer containing with pcDNA3-HA-Rad9 as a template.The triple-point protease and phosphatase inhibitors (Komatsu et al., 2000b), mutation Rad9 (DDD/AAA) was generated with the Strata- and normalized for protein content by BCA assay (Pierce). gene quick-change mutagenesis kit (Stratagene, CA, USA), Immunoprecipitations were performed by incubating 500 mgof using pcDNA3-HA-Rad9 as a template.The construction of cell lysate with 20 ml of protein G-agarose preadsorbed with pFLAG-Rad9DC, pcDNA-Myc-Rad9DC, and pEGFP- anti-FLAG M2 monoclonal antibody (Sigma) for 3 h at 41C. Rad9DC was accomplished by subcloning the PCR-ampliﬁed After multiple washes in lysis buffer, the beads were DNA fragment, encoding 1–267 amino acids of the human resuspended in 25 ml of Laemmli buffer, boiled, and subjected Rad9 protein into pFLAG-CMV2, pcDNA3-Myc or pEGFP- to immunoblotting analysis using polyclonal antibodies C2 (Clontech) vector. speciﬁc for FLAG (Sigma) or Bcl-XL (PharMingen).

Cell culture Apoptosis assay U937 cells were maintained in RPMI-1640 medium supple- 293 cells were transfected with pEGFP-C2, pEGFP-Rad9, or mented with 10% fetal calf serum (FCS), 100 U/ml penicillin pEGFP-Rad9DC plasmid DNA for 24 h.Cells were fixed with G, and 100 mg/ml streptomycin.MDA-MB-468, 293, HeLa, 3.7% formaldehyde, stained with 40,6-diamino-2-phenylinoo- and MCF-7 cells were cultured in Dulbecco’s modified Eagle’s dole (DAPI), and visualized by fluorescent microscopy.Cells medium (DMEM) supplemented with 10% FCS, 100 U/ml with clear condensed chromatin and/or fragmented nuclei were penicillin G, and 100 mg/ml streptomycin. determined as apoptosis. In vitro cleavage assays Acknowledgements [35S]methionine-labeled Rad9 proteins were generated using We thank Dr Guy S Salvesen (The Burnham Institute) for TNT-reticulocyte lysate kit (Promega), according to the kindly providing caspase-3 and -6 enzymes, and Dr Hirohito manufacturer’s recommendations. 35S-labeled proteins (5 ml) Yamaguchi for helpful discussion.This work was supported by were placed into 10 ml of caspase reaction buffer (50 mm grants CA82197 and CA90315 from the National Institutes of HEPES, pH 7.2, 100 mm NaCl, 1 mm EDTA, 0.1% Chaps, Health.Michael W Lee is a fellow of the American Heart 5mm DTT, 10% sucrose) containing either 0.75 mlof20mm Association.

References

Aravind L, Walker DR and Koonin EV.(1999). Nucleic Acids St Onge RP, Besley BD, Park M, Casselman R and Davey S. Res., 27, 1223–1242. (2001). J. Biol. Chem., 276, 41898–41905. Burtelow MA, Roos-Mattjus PM, Rauen M, Babendure JR St Onge RP, Udell CM, Casselman R and Davey S.(1999). and Karnitz LM.(2001). J. Biol. Chem., 276, 25903–25909. Mol. Biol. Cell, 10, 1985–1995. Caspari T and Carr AM.(2002). Curr. Biol., 12, R105–R107. Stennicke HR and Salvesen GS.(1997). J. Biol. Chem., 272, Chen MJ, Lin YT, Lieberman HB, Chen G and Lee EY. 25719–25723. (2001). J. Biol. Chem., 276, 16580–16586. Thornberry N, Rano T, Peterson E, Rasper D, Timkey T, Earnshaw WC, Martins LM and Kaufmann SH.(1999). Annu. Garcia-Calvo M, Houtzager V, Nordstrom P, Roy S, Rev. Biochem., 68, 383–424. Vaillancourt J, Chapman K and Nicholson D.(1997). J. Green DR and Reed JC.(1998). Science, 281, 1309–1312. Biol. Chem., 272, 17907–17911. Hartwell LH and Kastan MB.(1994). Science, 266, 1821–1828. Tsurimoto T.(1999). Front Biosci., 4, D849–D858. Hayashi K, Kuniyasu H, Oue N, Shigeishi H, Kuraoka K, Venclovas C and Thelen MP.(2000). Nucleic Acids Res., 28, Nakayama H and Yasui W.(2002). Pathobiology, 70, 40–46. 2481–2493. Hirai I and Wang HG.(2002). J. Biol. Chem., 277, 25722– Volkmer E and Karnitz LM.(1999). J. Biol. Chem., 274, 25727. 567–570. Komatsu K, Hopkins KM, Lieberman HB and Wang H-G. Wahl GM and Carr AM.(2001). Nat. Cell Biol., 3, (2000a). FEBS Lett., 481, 122–126. E277–E286. Komatsu K, Miyashita T, Hang H, Hopkins KM, Zheng W, Wang H-G and Reed JC.(1998). Histol. Histopathol., 13, Cuddeback S, Yamada M, Lieberman HB and Wang HG. 521–530. (2000b). Nat. Cell Biol., 2, 1–6. Yoshida K, Komatsu K, Wang HG and Kufe D.(2002). Mol. Nicholson DW.(1999). Cell Death Differ., 6, 1028–1042. Cell. Biol., 22, 3292–3300. O’Connell MJ, Walworth NC and Carr AM.(2000). Trends Zhou BB and Elledge SJ.(2000). Nature, 408, 433–439. Cell Biol., 10, 296–303. Zimmermann KC, Bonzon C and Green DR.(2001). Reed JC.(1998). Oncogene, 17, 3225–3236. Pharmacol. Ther., 92, 57–70. Salvesen GS and Dixit VM.(1997). Cell, 91, 443–446.

Oncogene