Caspase-3 Mediated Cleavage of BRCA1 During UV-Induced Apoptosis

Caspase-3 mediated cleavage of BRCA1 during UV-induced apoptosis

Qimin Zhan1,2, Shunqian Jin2, Bobby Ng2, Jordan Plisket1, Sanjeev Shangary1, Abhilasha Rathi1, Kevin D Brown3 and Rajasekaran Baskaran*,1

1Department of Molecular Genetics and Biochemistry, University of Pittsburgh, School of Medicine, Pittsburgh, Pennsylvania, PA 15261, USA; 2Pittsburgh Cancer Institute, University of Pittsburgh, School of Medicine, Pittsburgh, Pennsylvania, PA 15261, USA; 3Department of Biochemistry and Molecular Biology and the Stanley S. Scott Cancer Center, Louisiana State University, New Orleans, Louisiana, LA 70112, USA

The breast cancer suppressor protein, BRCA1 plays an have shown that, irrespective of the origin and the important role in mediating cell cycle arrest, apoptosis pathways that activate it, apoptosis proceeds through a and DNA responses to DNA damage signals. In this common mechanism involving predictable, reproduci- study, we show that BRCA1 level is downregulated ble sequence of events that includes DNA during UV-induced apoptosis by caspase-3 mediated fragmentation, chromatin condensation, membrane cleavage. Cleavage of BRCA1 by caspase-3 produced a blebbing, cell shrinkage and disassembly into fragment that contained the C-terminal of the molecule. membrane-enclosed vesicles (Hengartner, 2000). Accordingly, treatment of cells with caspase-3 inhibitor Central components of the apoptotic machinery are or mutation of a specific caspase-3 cleavage site (DLLD) a family of cysteine-proteases termed caspases. at amino acid 1151 – 1154 of BRCA1 abolished cleavage Caspases dismantle the cell by targeting a panoply of and consequential accumulation of the BRCA1 C- proteins for limited proteolysis (Earnshaw et al., 1999). terminal fragment. Whereas expression of the non- Generally synthesized as an inactive proform, they are cleavable BRCA1 (D/A 1154) mutant conferred the activated by proteolytic cleavage. Out of the 14 resistance phenotype to UV-induced cell death, expres- different caspases that have been identified, caspase-3, sion of the cleaved BRCA1 C-terminus induced cell thus far, remains the most common and important death in the absence of UV. Examination of the effector of apoptotic processes. Observations made in mechanism of C-terminus-induced cell death revealed caspase-3-deficient mice indicate that this protein is that the cleaved fragment triggers the apoptotic response necessary for specific apoptotic processes such as DNA through activation of BRCA1 downstream effectors, degradation, nuclear condensation, plasma membrane GADD45 and JNK. Altogether, results of our study blebbing, and proteolysis of certain substrates (Woo et demonstrate a functional role for caspase-3 mediated al., 1998). Generally, caspase-3 cleavage results in cleavage of BRCA1 during UV-induced apoptosis. functional inactivation of its targets; however, a Oncogene (2002) 21, 5335 – 5345. doi:10.1038/sj.onc. handful of proteins are activated by caspase-3 cleavage. 1205665 For example, Poly ADP-Ribose polymerase-1 (PARP- 1) when proteolyzed by caspase-3 produces a c- Keywords: DNA repair; BADD45; JNK terminal fragment that totally inhibits full-length PARP activity to allow apoptotic advancement (D’Amours et al., 2000; Kim et al., 2000). Similarly, Introduction cleavage of protein kinase C (PKC) d and j generates a 40 kDa catalytically active fragment that when Apoptosis is a normal physiological process that plays overexpressed results in events typical of apoptotic an important role in embryonic development and tissue advance (Ghayur et al., 1996). Thus, caspase-3 homeostasis maintenance (Afford and Randhawa; contributes to apoptotic progression not only through Meier et al., 2000). Because apoptosis is genetically functional inactivation but also through activation of programmed, its deregulation often contributes to target proteins. pathologies such as tumor promotion, immunodefi- The breast cancer suppressor protein, BRCA1 is a ciency diseases, and neurodegenerative disorders multifunctional protein implicated in DNA repair, cell (Thompson, 1995; Yuan and Yankner, 1999). cycle regulation and transcription (Vijver, 1993; Yu, Concerted efforts to characterize the apoptotic process 2000). BRCA1 is a 220 kDa nuclear protein, whose expression and phosphorylation is regulated in a cell cycle-dependent manner (Chen et al., 1996; Gudas et *Correspondence: R Baskaran, Molecular Genetics and Biochem- al., 1996; Ruffner and Verma, 1997). On its N- istry, E1205 Biomedical Science Tower, University of Pittsburgh terminus, BRCA1 contains a cysteine/histidine rich Medical Center, Pittsburgh, Pennsylvania, PA 15261, USA; E-mail: [email protected] RING finger domain that mediates its binding to Received 18 February 2002; revised 9 May 2002; accepted 14 May another RING finger containing protein, termed Bard1 2002 (Jin et al., 1997). The C-terminus of BRCA1 contains a Cleavage of BRCA1 by caspase-3 Q Zhan et al 5336 duplicated globular domain, the BRCT domain, that is A previous study determined that the breast cancer shared in many DNA repair proteins including tumor suppressor protein, BRCA1 level is down- XRCC1, DNA ligase III and IV, and the cell cycle regulated both at mRNA and protein level by UV-C checkpoint protein, Rad9 (Deng and Brodie, 2000). radiation in a dose- and time-dependent manner (Fan Exposure to genotoxic agents such as UV and ionizing et al., 1998). Another study showed that the steady- radiation induces BRCA1 phosphorylation by PI-3 state level of BRCA1 is regulated by proteolysis and kinase related family of kinases, ATM and ATR that N-acetyl-leu-leu-norleucinal (ALLN), an inhibitor (Cortez et al., 1999; Chen, 2000). It is generally of the proteasome caused accumulation of BRCA1 thought that phosphorylation activates its DNA repair protein (Blagosklonny et al., 1999). In this study we capability since mutations of the ATM/ATR phos- show that BRCA1 is cleaved by caspase-3 during UV- phorylation site on BRCA1 abrogates its DNA repair induced apoptosis. Cleavage of BRCA1 by caspase-3 function. A recent study showed that BRCA1 coexists caused accumulation of a C-terminal 90 kDa fragment with a number of DNA repair proteins including the of BRCA1 that when overexpressed induced cell death mismatch repair (MMR) proteins MLH1, MSH6, and in the absence of UV. Correlating well with activation MSH2, the Ataxia telangiectasia mutated (ATM) of cell death response the C-terminus of BRCA1 protein, and the Bloom’s syndrome protein (BLM), in activated GADD45/JNK signaling cascade. Conver- a multiprotein complex termed BASC (BRCA1- sely, expression of a non-cleavable BRCA1 mutant associated genome surveillance complex) (Wang et al., conferred protection against cell death and blocked 2000). Further, cells impaired in BRCA1 function GADD45 and JNK activation by UV radiation. These display impaired transcription-coupled repair (TCR) of results identify a functional role for caspase-mediated oxidative DNA damage arising from IR and H2O2 cleavage of BRCA1 in the regulation of cell death exposure (Gowen et al., 1998). Such findings support a response to UV irradiation. genome protection function for BRCA1 through its DNA repair capability. Results A distinct role for BRCA1 in transcriptional regulation has been proposed (Chapman and Verma, 1996; Cleavage of BRCA1 during UV-induced apoptosis Monteiro et al., 2000). The C-terminus of BRCA1, when fused to the GAL4 DNA binding domain activated To assess the temporal effect of genotoxic agent- GAL4 promoter. In transient transfection assays, induced apoptosis on the integrity of BRCA1, HL-60 BRCA1 increased p53-dependent and independent cells (a well-studied model system for apoptosis) were transcription from the p21WAF1/CIP1 and GADD45 induced to undergo apoptosis by exposure to 60 joules/ promoters (Chai et al., 1999; Jin et al., 2000). In m2 of UV-C radiation. Induction of apoptosis was addition, BRCA1 caused accumulation of p53 after evident by the appearance of a detectable population gamma-irradiation through transcriptional stimulation of cells containing sub-2N complement of DNA after of the p53-responsive promoter, MDM2. A number of irradiation (Figure 1a). The time-dependent cleavage of transcription regulatory proteins including ATF1, p53, the apoptotic marker proteins, caspase-3 (Figure 1b) BARD1, CtIP, CBP/p300, and RNA Polymerase II have and poly-ADP-polymerase (PARP) (Figure 1c) further been shown to interact with BRCA1 (Monteiro et al., confirmed apoptotic induction in irradiated cells. 2000). BRCA1 may also regulate transcription through Immunoprecipitation followed by immunoblotting of interaction with the SWI/SNF chromatin-remodeling BRCA1 revealed the presence of a *220 kDa BRCA1 complex (Bochar et al., 2000). band that, following UV-irradiation, displayed retarded A major transcriptional target of BRCA1 is the electrophoretic migration (Figure 1d, compare lanes 1 DNA damage-responsive gene, GADD45. Overexpres- and 2), due to UV-stimulated phosphorylation (Thomas sion of BRCA1 caused transcriptional stimulation of et al., 1997). In addition to full-length BRCA1, an the DNA damage-responsive gene GADD45 (Harkin et immunoreactive band of *90 kDa MW was detected in al., 1999). Studies involving high-density cDNA array UV-irradiated cells but not in control population (lane screening identified many genes, including GADD45,as 2). Concomitant with an increase in the abundance of the BRCA1-regulated genes (MacLachlan et al., 2000). 90 kDa band, a decrease in full-length Brca1 protein was Induction of BRCA1 triggers apoptotic response in also observed at 5 and 7 h post-irradiation (lanes 3 and U2OS correlated well with activation of c-Jun N- 4). Cleavage of BRCA1 is not due to non-specific terminal kinase/stress-activated protein kinase (JNK/ activation of proteases during lysate preparation since SAPK) a signaling pathway potentially linked to under the same condition cleavage of the BRCA1-related GADD45 gene family members. These findings have protein, BRCA2 was not observed (not shown). established a potential role for BRCA1 in apoptotic Together, these results indicate that BRCA1 is specifi- initiation through a pathway involving GADD45 and cally cleaved during UV-induced apoptosis. JNK. Further, BRCA1 mediated induction of The detection of the 90 kDa Brca1 fragment by an GADD45 and JNK/SAPK-dependent apoptosis is antibody that recognizes the C-terminus of Brca1 (C- shown to be p53 independent (Thangaraju et al., 20) but not the N-terminus of BRCA1 (N-12) (not 2000). Additionally, serum deprivation or calcium shown) suggested that the UV-induced cleavage results ionophore treatment of BRCA1 transfectants resulted in accumulation of a 90 kDa C-terminal fragment of in programmed cell death (Shao et al., 1996). Brca1. To confirm this, HeLa cells ectopically expres-

Oncogene Cleavage of BRCA1 by caspase-3 Q Zhan et al 5337

Figure 1 Cleavage of BRCA1 during UV-induced apoptosis. HL-60 cells were exposed to UV-C (60 joules/m2), collected at the indicated time points and their sub-2N DNA content was assessed by ﬂow cytometry (a). Lysates prepared from mock and treated cells were resolved on a 10 – 15% SDS – PAGE, transferred onto an Immobilon-P and probed with anti-caspase-3 (b) or PARP antibody (c). Extracts were also used for immunoprecipitation of BRCA1 followed by immunoblotting against anti-BRCA1 (d) UV-induced cleavage of C-terminally Flag-tagged BRCA1 (e)

sing a C-terminally Flag-tagged BRCA1 were exposed inhibitor, completely abrogated BRCA1 proteolysis to UV and its cleavage was examined. Results (Figure (Figure 3a, lane 5). As expected, neither IETD-fmk 1e) showing the unambiguous recognition of the nor VETD-fmk had an effect on UV-induced caspase-3 90 kDa cleaved polypeptide by anti-FLAG antibodies cleavage (Figure 2b, lanes 2 and 3); however, caspase-3 in UV-irradiated cells establish the 90 kDa band as the inhibitor completely blocked its activation (lane 5). The C-terminus of BRCA1. The fate of the N-terminal presence of a low but detectable level of BRCA1 fragment of BRCA1 resulting from this proteolytic cleavage in LEHD-treated cells (Figure 2a, lane 4) cleavage is not known; however, it is possible that this could be due to caspase-9-dependent activation of fragment, which contains the RING finger domain caspase-3 (Figure 2b, lane 4). Nevertheless, results of may be targeted for degradation by the ubiquitin- these experiments support the cleavage of BRCA1 by mediated proteosome pathway. caspase-3 protease. To provide direct support for the involvement of caspase-3 in BRCA1 cleavage, 35S-methionine-labeled Cleavage of BRCA1 is mediated by caspase-3 BRCA1 prepared by transcription-coupled translation To determine if cleavage of BRCA1 is mediated by in vitro was tested for its ability to serve as caspase-3 caspase class of proteases, HL-60 cells were induced to substrate. While abundant full-length BRCA1 was undergo apoptosis by UV irradiation in the presence of detected both in the crude reaction and after various caspase inhibitors and cleavage was assessed. immunoprecipitation from the in vitro translated Treatment of cells with IETD-fmk, a caspase-8 products (Figure 3a, lanes 2 and 5) incubation with inhibitor, did not affect BRCA1 cleavage (Figure 2a, caspase-3 resulted in the appearance of a *70 kDa lane 2). Similarly, VETD-fmk, a caspase-6 inhibitor, proteolytic fragment (Figure 3a, lanes 3 and 6). The also did not interrupt BRCA1 cleavage (lane 3). identity of this band as a C-terminal proteolytic However, treatment with LEHD-fmk, a caspase-9 fragment was unambiguous since this moiety was inhibitor, resulted in *10% inhibition of BRCA1 immunoprecipitated by the anti-BRCA1 antibody (C- cleavage (lane 4). ZVED-fmk, a caspase-3 specific 20) that recognizes the C-terminus. A control reaction

Oncogene Cleavage of BRCA1 by caspase-3 Q Zhan et al 5338

Figure 3 In vitro cleavage of full-length BRCA1 by purified caspase-3. 35S-labeled FLAG-tagged BRCA1 (and PARP) prepared by in vitro transcription/translation was incubated with purified Figure 2 Effects of caspase inhibitors on BRCA1 cleavage. HL- caspase-3 for 90 min at 308C. The reaction products were ana- 60 cells were mock-treated or exposed to UV-C (60 joules/m2)in lysed directly by 6% SDS – PAGE or used for immunoprecipita- the presence of 100 mM of the cell permeable caspase inhibitors, tion of BRCA1 using anti-FLAG antibody and then analysed IETD, VETD, LEHD, or ZVED-fmk. Twelve hours post-UV, ly- by SDS – PAGE. The gel was dried, subjected to fluorography sates were prepared, and used for immunoprecipitation followed and exposed for autoradiography (a). Cleavage of in vitro trans- by immunoblotting against anti-BRCA1 antibody (C-20) (a). lated PARP by caspase-3 (b) Lysates were also probed with anti-caspase-3 antibody (b)

showing the cleavage of PARP by caspase-3 shows the caspase-3 deficient MCF-7 cells with wt or D/A specificity of in vitro cleavage reaction (Figure 3b). mutants of BRCA1 along with procaspase-3. Similar to what was observed with endogenous BRCA1, cells expressing wt and D/A414 yielded a 90 kDa fragment Identification of the caspase-3 cleavage site in BRCA1 (Figure 4b, lanes 2 and 3). In contrast, no detectable Previous studies established that substrates of caspase- proteolysis of the D/A1154 mutant of BRCA1 was 3 are recognized and cleaved due to the presence of the observed (lane 4). The inability of the D/A1154 mutant consensus DXXD sequence (Earnshaw et al., 1999). to be cleaved despite the presence of activated caspase- Inspection of BRCA1sequence revealed the presence of 3 (Figure 4c), manifests itself as caspase-3 cleavage site. two putative caspase-3 cleavage sites (DVLD and DLLD) at amino acids positions 411 – 414 and 1151 – Involvement of BRCA1 cleavage in apoptosis 1154, respectively (Figure 4a). Consistent with our results concerning the digestion of BRCA1 by caspase- To determine the functional significance of the BRCA1 3 in vitro, cleavage of BRCA1 at position at 1154 is cleavage, HCT 116 cells stably expressing low levels of predicted to generate a fragment of *70 kDa. In order wt or D/A mutants of BRCA1 (Figure 6b) were to evaluate this motif as the caspase-3 recognition site, examined for UV-induced apoptotic response. Under we mutated aspartic acids at position 1151 and 1154 to the conditions assayed, expression of wt-BRCA1 alanine (designated as D/A1154). Additionally, a caused *20% spontaneous cell death. However, as second mutant possessing an analogous mutation at shown in Figure 5a, exposure to 60 joules/m2 of UV position 411 and 414 (D/A414) was also constructed. resulted in a 2.7-fold increase in Annexin V positive The mutant constructs were tagged with the Flag cells. A similar result was obtained with D/A 414 epitope and their ability to serve as caspase-3 mutant of BRCA1. Interestingly, cells expressing the substrates in vivo was tested by transfection of D/A1154 mutant showed only a 1.8-fold increase in cell

Oncogene Cleavage of BRCA1 by caspase-3 Q Zhan et al 5339

Figure 4 Identiﬁcation of the caspase-3 cleavage site in BRCA1. Schematic structure of BRCA1 protein showing the location of consensus caspase-3 recognition sites (a). MCF-7 cells were transfected with pCDNA3 plasmid containing FLAG-tagged wt,D/ A1154 or 414 mutants of BRCA1 along with pro-caspase-3 using lipofectamine-mediated transfection. Twenty-four hours post- transfection, lysates were prepared and resolved on 8% SDS – PAGE. The proteins were transferred to Immobilon-P membrane and immunoblotted against anti-Flag (b) or anti-caspase-3 antibody (c)

death following UV treatment. Consistent with these colonies were observed with the non-cleavable BRCA1 results, wt and D/A414 BRCA1 cells showed *4.5-fold mutant than with wt or D/A 414 mutant of BRCA1 increase in caspase-3 activity following UV-irradiation (Figure 6a,b). The increased UV survivability of the (Figure 5b). With the D/A 1154 mutant only a twofold non-cleavable mutant cells further confirmed the activation of caspase-3 activity was observed. To involvement of BRCA1 cleavage in mitogenic cell independently evaluate the effect of D/A mutations, death response. HeLa cells were transfected with wt and D/A mutants If cleavage of BRCA1 is required for apoptotic of BRCA1 along with GFP and UV-triggered cell advancement then it is a reasonable expectation that death response was assessed. Examination of the the cleaved product induces apoptosis. To determine chromatin morphology by Hoechst staining revealed the same, cDNA constructs encoding varying C- a large proportion of wt (40%) and D/A414 (41%) terminal segments of BRCA1 (BE and KE; Figure BRCA1 mutant cells with condensed chromatin, 7a) were expressed in HEK 293 cells, and their ability indicative of robust apoptotic advancement. Notably, to promote apoptosis was examined. As control, the N- fewer cells (12%) with condensed chromatin were terminal NK fragment of BRCA1 was also tested. observed in cells expressing the D/A 1154 mutant Unlike HCT 116 cells, statement of full-length BRCA1 (Figure 5c). Together, the results indicate that cleavage induced robust cell death, as demonstrated by 46% of BRCA1 is a requisite for apoptotic advancement of increase in Annexin V positive cells (Figure 7b). UV-irradiated cells. Expectedly, KE and BE fragments also induced We next examined if expression of the non-cleavable apoptosis with equivalent efficiency. In stark contrast, BRCA1 mutant protects cells from UV-mediated cell expression of the NK fragment had very little effect on killing by employing a clonogenic assay. HCT 116 cells cell death. In agreement with these findings, KE and expressing low levels of wt and D/A mutants (Figure BE but not NK fragment transfectants displayed high 6b) were exposed to low doses of UV, which induced levels of caspase-3 activity (Figure 7c). Together, the low levels of BRCA1 cleavage, and the percentage of results demonstrate that the C-terminus of BRCA1 surviving colonies after 14 days was assessed. Under contains pro-apoptotic property similar to that of full- the conditions assayed, *6 – 8-fold more surviving length BRCA1.

Oncogene Cleavage of BRCA1 by caspase-3 Q Zhan et al 5340

Figure 5 Expression of the BRCA1 D/A 1154 mutant protects cells against UV-induced apoptosis. HCT 116 cells expressing wt or D/A (414 and 1154) mutants of BRCA1 were exposed to UV-C (60 joules/m2) and 12 h after exposure the cell death response was assessed by propidium iodide/Annexin V staining followed by FACS analysis. The percentage of cells displaying Annexin V staining is given (a). Lysates prepared from these cells were adjusted for equal protein concentration and caspase-3 activity was measured using a ﬂuorogenic substrate as described in Materials and methods (b). HeLa cells transfected with wt or mutants of BRCA1 along with GFP plasmid were exposed to UV and after 12 h cells were stained with Hoechst and viewed under microscopy (c)

Induction of apoptosis by the C-terminus of Brca1 occurs via GADD45/JNK pathway A study by Harkin et al. (1999) showed that apoptotic initiation by full-length BRCA1 occurs via activation of the stress-responsive gene, GADD45. Further, in transient transactivation assays, BRCA1 activated the GADD45 promoter in a p53-independent manner (Jin et al., 2000; Thangaraju et al., 2000). Therefore, we examined if the pro-apoptotic activity of the C- terminus of BRCA1 relies, in part, on its ability to transactivate GADD45 in transient transfection assays utilizing CAT reporter under the control of GADD45 promoter (nt-994 to+1). Consistent with the report of Jin et al. (2000), expresssion of wt BRCA1 activated (*eightfold) GADD45 promoter (Figure 8b). Further, the C-terminal fragments, BE and KE displayed transactivation property similar to that wt-BRCA1 in that fold activation closer to eight and nine were obtained. Interestingly, despite similar levels of expression (Figure 8a) the D/A 1154 mutant showed signiﬁcantly reduced GADD45 transactivation. Figure 6 Cloning efﬁciency of transfected clones irradiated with ultraviolet. HCT 116 cells expressing wt or D/A mutants of The stress-activated protein kinase JNK is another BRCA1 were exposed to 2, 4, 8 and 10 joules/m2 of UV and their important mediator of UV-induced apoptosis (Tour- colony forming ability were examined as described in Materials nier et al., 2000). Like GADD45, a role for JNK in and methods. Data represent the average percentage of colony BRCA1-induced apoptosis is indicated by the observa- numbers relative to that of mock-irradiated control cells (a). Cell extracts were formed and analysed for recombinant BRCA1 tion that expression of a dominant negative JNK expression by immunoblotting the lysate against anti-Flag mutant abrogated BRCA1-induced apoptosis (Harkin antibody (b) et al., 1999). To determine if the apparent pro-

Oncogene Cleavage of BRCA1 by caspase-3 Q Zhan et al 5341

Figure 7 Induction of apoptosis by the C-terminus of BRCA1. Schematic representation of wt and C-terminal (NK (AA 1 – 810)) and N-terminal (KE (AA 810 – 1863) and BE (1352 – 1863)) deletion mutants of BRCA1 (a). Amino acid residue numbers within the primary structure of human BRCA1 are indicated. HEK 293 cells were transfected with vector, wt, NK, BE or KE fragments and collected after 48 h post-transfection. Cells were stained with propidium iodide/Annexin V staining and analysed by ﬂow-cytometry. The percentage of cells displaying Annexin V staining is given; error bars represent s.d. from the mean (b). Lysates prepared from transfected cells were adjusted for equal protein concentration and immunoblotted against caspase-3 antibody (c)

apoptotic activity of the C-terminal fragments involves vitro generated a C-terminal fragment of *70 kDa. JNK activation, cells transfected with wt, D/A, or The lower MW of the cleavage product could be due to truncation mutants of BRCA1 were assayed for the lack of extensive phosphorylation that BRCA1 activation of JNK by immunoblotting the lysate with undergoes on its C-terminus after UV irradiation in an antibody that recognizes dually (Threonine and vivo. Finally, mutation of a caspase-3 speciﬁc cleavage Tyrosine) phosphorylated active JNK. The striking site (D/A1154) rendered BRCA1 non-cleavable. correlation of the KE and BE fragment to induce Overstatement of BRCA1 results in induction of apoptosis and to activate JNK, as shown in Figure 8c, high levels of spontaneous apoptosis in U2OS cells provide further support for the involvement of the (Harkin et al., 1999). Additionally, in ovarian and cleaved C-terminus in the apoptotic advancement of prostate cancer cell lines, BRCA1 enhanced apoptosis UV-irradiated cells. induced by other stimuli such as growth factor withdrawal, substratum detachment, ionizing radiation (Thangaraju et al., 2000). While induction of apoptosis Discussion by full-length BRCA1 suggests a more upstream role for BRCA1, results presented here are consistent with In this study, we demonstrate the cleavage and the BRCA1 acting downstream of caspase-3. It should be consequential downregulation of full-length BRCA1 by noted, however, that forced expression often causes caspase-3 during UV-induced apoptosis. Cleavage of accumulation of non-physiological levels and cells BRCA1 by caspase-3 produces a fragment that respond to such toxic levels of proteins by mounting contains the C-terminal portion of the molecule. a death response. Unlike overexpression-induced Several lines of evidence indicate that cleavage is apoptosis, UV-induced BRCA1 cleavage and the caspase-3 mediated. First, treatment of cells with consequential accumulation of its C-terminal fragment, caspase-3 inhibitor blocked BRCA1 cleavage. Second, as reported here, are observed in cells carrying incubation of full-length BRCA1 with caspase-3 in physiological levels of BRCA1. Of note, low levels of

Oncogene Cleavage of BRCA1 by caspase-3 Q Zhan et al 5342

Figure 8 Effect of C-terminal fragment of BRCA1 on GADD45/JNK pathway. HCT 116 cells were co-transfected with expression plasmids encoding wt, D/A, or C-terminal fragments BE or KE along with a CAT reporter plasmid containing 7994 to +1 of the proximal promoter region of the GADD45 gene. Thirty-six hours post-transfection, cells were harvested, normalized for equal protein content, and analysed for transgene expression (a) or CAT activity (b). Reporter gene activity was first normalized to b-galac- tosidase activity obtained by cotransfection of an internal control pSV40-b-gal expression plasmid. Values represent per cent acetylation of [14C]chloromphenicol, normalized for transfection efficiencies. The mean from five independent experiments is shown. Error bars represent s.d. from the mean. Lysates were also probed for JNK activation by immunoblotting with anti-phospho-specific JNK antibody (c)

the cleaved C-terminus were observed during over- Role of BRCA1 cleavage in DNA repair and cell cycle expression-induced apoptosis (not shown). Thus, low levels of caspase-3, activated during overexpression What is the functional consequence of caspase-3 may cause cleavage of BRCA1 ultimately leading to cleaved BRCA1? As mentioned above, abrogation of the activation of cell death program. Moreover, the BRCA1’s repair function through caspase-3 mediated increased sensitivity of BRCA1-mutated cells cleavage could render cells that sustain extensive (HCC1937) to UV and other DNA damaging agents damage to cell death rather than be transformed. suggest a role for BRCA1 in DNA repair rather than Presently, the mechanism by which BRCA1 partici- triggering the apoptotic response to DNA damage pates in DNA repair is not clear. However, one study (Gowen et al., 1998). Further, our observation that determined that the N-terminus of BRCA1 binds DNA BRCA1 is cleaved during apoptosis is also consistent repair proteins such as the mammalian RecA homolog with cleavage and inactivation of a number of DNA Rad51 and its binding partner Rad52 (Yu, 2000) and repair proteins including DNA-PK and PARP that this interaction with Rad family members may observed during activation of cell death program mediate its repair functions. A more recent study (Aﬀord and Randhawa, 2000; Earnshaw et al., 1999). determined that the inherent property of BRCA1 to Together, such ﬁndings argue the presence of BRCA1 bind branched DNA structures and form protein-DNA downstream of caspase-3 in the UV induced cell death complexes cooperatively between multiple DNA pathway. strands may be important for its role in DNA repair

Oncogene Cleavage of BRCA1 by caspase-3 Q Zhan et al 5343 (Paull et al., 2001). The coexistence of BRCA1 with a embryonic fibroblasts refractory to UV-induced apop- number of repair proteins further suggests that BRCA1 tosis (Tournier et al., 2000), suggests that BRCA1 C- may co-ordinate multiple activities during DNA repair terminus-induced apoptosis involves catalytic activa- (Wang et al., 2000). These findings raise the possibility tion of JNK activity. that cleavage of BRCA1 may render it incapable of In summary, evidence provided here shows that binding to structured DNA or repair proteins BRCA1 is downregulated during UV-induced apopto- neutralizing its role in DNA repair. sis by caspase-3 mediated cleavage. While expression of Beyond its documented role in DNA repair and a non-cleavable mutant of BRCA1 raised the cellular apoptosis, BRCA1 has been implicated in establish- resistance to UV-induced cell death, the C-terminal ment of the G2/M cell cycle checkpoint following DNA fragment activated the cell death program in the damage. In support of this view, cells from BRCA1- absence of UV irradiation. Correlating these findings deficient mice display defective G2/M arrest response well, the C-terminus activated the GADD45/JNK to DNA damaging agent treatment (Xu et al., 1999). pathway whereas the non-cleavable mutant down- Interestingly, expression of a C-terminal fragment of regulated it. These data support a functional role for BRCA1 (amino acids 1293 – 1863) abrogated colchi- caspase-3 mediated cleavage of BRCA1 in the cine-induced G2/M arrest (Larson et al., 1997). The regulation of cell death response to UV irradiation. latter observation suggests a dominant negative role for the C-terminal domain in the G2/M arrest response. Additionally, a number of cell cycle regulatory proteins, including RB and p21, are cleaved during apoptosis and this event is thought to abolish the roles Materials and methods these proteins play in cell cycle advance (Earnshaw et Cell culture and reagents al., 1999; Walter et al., 1998). Thus, an exit from cell cycle arrest that may be necessary for the initiation of HeLa, HL-60, HEK293 and HCT 116 cells were purchased apoptosis may also be accomplished by cleavage. from American Tissue Type Collection (ATCC) and maintained in Dulbecco’s modified Eagle’s medium (DMEM) containing 10% heat inactivated fetal bovine serum (FBS) or Role of BRCA1 cleavage in UV-induced apoptosis RPMI containing 15% FBS. Stable transfectants were obtained by transfecting HCT 116 cells with 3 – 5 mgof The apoptotic resistance phenotype displayed by D/ plasmid using lipofectamine and selecting with G418 (at a A1154 mutant suggests that cleavage of BRCA1 is a concentration of 0.4 mg/ml) or Zeomycin as per the requisite for apoptotic progression. Consistent with this manufacturer’s protocol. Leupeptin, aprotinin and phenyl- result, expression of the cleaved C-terminus but not the methylsulf-onyl fluoride (PMSF) were purchased from Sigma. N-terminus induced apoptosis. In fact, expression of Caspase inhibitors, ZVED-fmk, LEHD-fmk, IETD-fmk, and the NK fragment protected cells from cell death VETD-fmk were purchased from BioMol. Anti-BRCA1 (C20 triggered by UV exposure. Studies by Thangaraju et and N-17), anti-caspase-3 (H277), and anti-PARP (F2) were al. (2000) showed an anti-apoptotic role for the N- purchased from Santa Cruz Biotechnology. Anti-Flag (M2) antibody was from Sigma-Aldrich. Horseradish peroxidase- terminus of BRCA1 since a truncation mutant of conjugated goat anti-rabbit and anti-mouse were purchased BRCA1 lacking the C-terminus suppressed apoptosis, from Bio-Rad. The fluorogenic caspase-3 substrate including spontaneous apoptosis. Therefore, cleavage Ac-DEVD-AFC (7-amino-4-trifluoromethylcoumarin) was of BRCA1 may activate the pro-apoptotic function of obtained from Pharmingen. the C-terminus through removal of the anti-apoptotic N-terminus. Additionally, BRCA1 cleavage could Plasmids activate its GADD45 transcriptional stimulatory function that could potentially contribute to apoptosis. The BRCA1 C-terminal fragments KE and BE were Although the precise role of BRCA1-induced GADD45 constructed by digestion of a wildtype (wt) BRCA1 cDNA transcription in UV-induced apoptosis remains to be clone with KpnI and EcoRI (KE fragment) or with BamHI and EcoRI (BE fragment) and subcloning the released elucidated, studies by Rishi et al. (1999) showed that fragments into pcDNA4 His-Max vector from Invitrogen. post-transcriptional upregulation of GADD45 corre- The BRCA1 N-terminal fragment, NK was obtained by lates with apoptotic induction by synthetic retinoids in digestion with HindII and KpnI. D/A 414 and D/A 1154 both human breast and lung carcinoma cells. Further- BRCA1 mutants were created utilizing the Gene Editor more, GADD45 upregulation is observed during mutagenesis system as described by the manufacturer apoptosis induced by agents that cause endoplasmic (Promega). The mutagenic oligonucleotides used for the reticulum stress (Sheikh et al., 2000). Such findings creation of D/A mutants were 5-GCCAAAGTAGCTGC- support the GADD45 transcriptional stimulatory TGTATTGGCCGTTCTAAATGAGGTAG-3 for D414A function as a mediator of BRCA1 C-terminus-induced and 5-GAGACACCTGCCCTGTTAGCTGGTGAAATAA- apoptosis. In addition to GADD45 transactivation, AGG-3 for D/A1154 mutant. forced expression of the C-terminal fragment activated JNK, as judged by an increased abundance of dually Immunofluorescence (tyrosine and threonine) phosphorylated JNK. This HeLa cells were plated at a density of 26105 cells/plate and observation, when taken with the finding that targeted transfected with Flag-tagged BRCA1 by lipofectamine- disruptions of all functional JNK genes rendered mediated trasfection. Transfection efficiency was monitored

Oncogene Cleavage of BRCA1 by caspase-3 Q Zhan et al 5344 by including 0.5 mg of pEGFP plasmid and determining the Immunoprecipitation and immunoblotting percentage of cells expressing EGFP by fluorescence microscopy (Nikon). Thirty-six hours post-infection, cells were At the indicated time points, cells were lysed in 16lysis 2 either mock-treated or exposed to UV (60 joules/cm ) and buffer containing 25 mM Tris pH 7.6, 150 mM NaCl, 1 mM collected at the indicated time points. Cells were fixed on EDTA, 1 mM NaF, 1% Triton X-100, 1 mM PMSF, 1 mM cover slips with 4% formaldehyde in PBS and stained with DTT, 1 mM sodium orthovanadate, and 1 mg/ml of pepstatin, Hoechst stain. Cover slips were then mounted onto slides and leupeptin, and aprotinin. Lysate was centrifuged for 30 min apoptotic morphology was examined under a Ziess fluores- at 13 000 r.p.m. Anti-BRCA1 (C-20 or N-17) or anti-Flag cence microscope. antibody was added at a concentration of 0.1 mg/ml to the supernatant and immune-complexes were collected by binding to 50 ml of Protein A-Sepharose beads (5 mg/ml). Apoptosis The beads were washed twice with 16lysis buffer containing Cultures of cells at *70% confluency were exposed to UV-C 500 mM NaCl, twice with lysis buffer containing 25 mM NaCl (60 joules/m2) and harvested at the indicated time points. and twice with 16lysis buffer. Equal volume of SDS sample Cells were washed in phosphate-buffered saline (PBS), and buffer was added and the samples were resolved by SDS – fixed with ice-cold 70% ethanol overnight. Cells were then PAGE on 6% acrylamide gels. Proteins were transferred onto suspended in 16PBS containing RNase A (20 mg/ml) and Immobilon-P (Millipore) membrane using a Bio-Rad semi- propidium iodide (20 mg/ml). Fluorescence-activated cell dry transfer apparatus. Membranes were then blocked in PBS sorting (FACS) was performed with a Becton Dickinson containing 0.5% Tween-20 and 5% dry milk for 1 h at RT FACSort apparatus and used to quantitate the apoptotic and probed with primary and secondary antibody. Blots were population based on DNA levels. For assessing phosphati- developed using Enhanced Chemiluminescence from NEN. dylserine redistribution, cells were incubated with 0.1 mM Fluorescein-conjugated human Annexin V (Pharmingen) in UV survival assays PBS containing 2% fetal bovine serum for 15 min at 378C prior to flow-cytometric analysis. HCT 116 cells expressing wt and mutants of BRCA1 were plated at a confluency of 16105 cells per well in 60-mm plates and exposed to varying doses of UV radiation (2, 4, 6 Cleavage assays and 10 joules/m2) then plated in 100-mm tissue culture Full-length BRCA1 was translated using TNT-reticulocyte dishes. The cultures were maintained until the surviving cells lysate system (Promega) with T7 RNA polymerase in the formed colonies. The colonies that survived after incubation presence of 35S-methionine. Five ml of the translation mix were then stained with crystal violet, counted, and relative was incubated with 1 ml (30 units) of purified caspase-3 colony numbers were obtained. (BioMol) in a 20 ml reaction buffer containing 100 mM HEPES pH 7.5, 10% sucrose and 0.1% 3-[(3-cholamidopro- pyl) dimethylammonio]-1-propanesulfonate. Following incu- Abbreviations bation at 308C for 2 h, labeled proteins were resolved by UV, Ultraviolet; DSB, Double strand break; FACS, fluores- SDS-polyacrylamide gel electrophoresis followed by fluoro- cence-activated cell sorting; PI, propidium iodide; PAGE, graphy. polyacrylamide gel electrophoresis

Caspase-3 activity Acknowledgments Caspase activity in whole-cell extracts was determined by The authors thank Don Jhonson at Pittsburgh Cancer treating cells with a solution containing 50 mM HEPES Institute for PARP constructs, Edison Liu at National (pH 7.4), 1% Triton X-100, 5 mM EDTA, and 2 mM DTT Institute of Health for the HA-tagged BRCA1. R Baskaran and incubating them with 50 mM acetyl-Asp-Glu-Val-Asp-7- is supported by a grant from National Institute of Health amino-4-methylcoumarin (AcDEVD-amc) for 30 min at 378C (GM 60945) and a research project grant (RPG) from as described (Tournier et al., 2000). Fluorescence in the linear American Cancer Society. Q Zhan is supported by a NIH range of DEVDase activity was measured with a plate reader grant (CA 93640). The critical reading of the manuscript by (Tecan). Tom E Smithgall, Grace and Erin Child is acknowledged.

References

Afford S and Randhawa S. (2000). Mol. Pathol., 53, 55 – 63. D’Amours D, Sallmann FR, Dixit VM and Poirier GG. Blagosklonny MV, An WG, Melillo G, Nguyen P, Trepel JB (2001). J. Cell Sci., 14, 3771 – 3778. and Neckers LM. (1999). Oncogene, 18, 6460 – 6468. Deng CX and Brodie SG. (2000). Bioessays, 22, 728 – 737. BocharDA,WangL,BeniyaH,KinevA,XueY,LaneWS, Earnshaw WC, Martins LM and Kaufmann SH. (1999). Wang W, Kashanchi F and Shiekhattar R. (2000). Cell, Annu. Rev. Biochem., 68, 383 – 424. 102, 257 – 265. Fan S, Twu NF, Wang JA, Yuan RQ, Andres J, Goldberg ID Chai YL, Cui J, Shao N, Shyam E, Reddy P and Rao VN. and Rosen EM. (1998). Int. J. Cancer, 77, 600 – 609. (1999). Oncogene, 18, 263 – 268. Ghayur T, Hugunin M, Talanian RV, Ratnofsky S, Quinlan Chapman MS and Verma IM. (1996). Nature, 382, 678 – 679. C,EmotoY,PandeyP,DattaR,HuangY,KharbandaS, Chen J. (2000). Cancer Res., 60, 5037 – 5039. Allen H, Kamen R, Wong W and Kufe D. (1996). Exp. Chen Y, Farmer AA, Chen CF, Jones DC, Chen PL and Lee Med., 184, 2399 – 2404. WH. (1996). Cancer Res., 56, 3168 – 3172. Gudas JM, Li T, Nguyen H, Jensen D, Rauscher FJ and Cortez D, Wang Y, Qin J and Elledge SJ. (1999). Science, Cowan KH. (1996). Cell Growth Differ., 6, 717 – 723. 286, 1162 – 1166.

Oncogene Cleavage of BRCA1 by caspase-3 Q Zhan et al 5345 Gowen LC, Avrutskaya AV, Latour AM, Koller BH and Shao N, Chai YL, Shyam E, Reddy P and Rao VN. (1996). Leadon SA. (1998). Science, 281, 1009 – 1012. Oncogene, 13, 1–7. Harkin DP, Bean JM, Miklos D, Song YH, Truong VB, Sheikh MS, Hollander MC and Fornance Jr AJ. (2000). Englert C, Christians FC, Ellisen LW, Maheswaran S, Biochem. Pharmacol., 59, 43 – 45. Oliner JD and Haber DA. (1999). Cell, 97, 575 – 586. Thomas JE, Smith M, Tonkinson JL, Rubinfeld B and Hengartner MO. (2000). Nature, 407, 770 – 776. Polakis P. (1997). Cell Growth Differ., 8, 801 – 809. Jin S, Zhao H, Fan F, Blanck P, Fan W, Colchagie AB, Thompson CB. (1995). Science, 267, 1456 – 1462. Fornace Jr AJ and Zhan Q. (2000). Oncogene, 19, 4050 – Thangaraju M, Kaufmann SH and Couch FJ. (2000). J. Biol. 4057. Chem., 275, 33487 – 33496. JinY,XuXL,YangMC,WeiF,AyiTC,BowcockAMand Tournier C, Hess P, Yang DD, Xu J, Turner TK, Nimnual A, Baer R. (1997). Proc. Natl. Acad. Sci. USA, 94, 12075 – Bar-Sagi D, Jones SN, Flavell RA and Davis RJ. (2000). 12080. Science, 288, 870 – 874. Kim JW, Kim K, Kang K and Joe CO. (2000). J. Cell Sci., Vijver MJ. (1993). Adv. Cancer Res., 61, 25 – 56. 275, 8121 – 8125. Walter BN, Huang Z, Jakobi R, Tuazon PT, Alnemri ES, Larson JS, Tonkinson JL and Lai MT. (1997). Cancer Res., Litwack G and Traugh JA. (1998). J. Biol. Chem., 273, 57, 3351 – 3355. 28733 – 28739. MacLachlan TK, Somasundaram K, Sgagias M, Shifman Y, Wang Y, Cortez D, Yazdi P, Neff N, Elledge SJ and Qin J. Muschel RJ, Cowan KH and El-Deiry WS. (2000). J. Biol. (2000). Genes Dev., 14, 927 – 939. Chem., 275, 2777 – 2785. Woo M, Hakem R, Soengas MS, Duncan GS, Shahinian A, Meier P, Finch A and Evan G. (2000). Nature, 407, 796 – 801. Kagi D, Hakem A, McCurrach M, Khoo W, Kaufman SA, Monteiro AN. (2000). Trends Biochem. Sci., 25, 469 – 474. Senaldi G, Howard T, Lowe SW and Mak TW. (1998). Paull TT, Cortez D, Bowers B, Elledge SJ and Gellert M. Genes Dev., 12, 806 – 819. (2001). Proc. Natl. Acad. Sci. USA, 98, 6086 – 6091. XuX,WeaverZ,LinkeSP,LiC,GotayJ,WangXW,Harris RishiAK,SunRJ,GaoY,HsuCK,GeraldTM,SheikhMS, CC, Ried T and Deng CX. (1999). Mol. Cell., 3, 389 – 395. DawsonMI,ReichertU,ShrootB,FornaceJrAJ,Brewer Yuan J and Yankner BA. (1999). Nat. Cell Biol., E44 – E45. G and Fontana JA. (1999). Nucleic Acids Res., 27, 3111 – Yu W. (2000). Breast Cancer Res., 2, 82 – 85. 3119. Ruffner H and Verma IM. (1997). Proc. Natl. Acad. Sci. USA, 94, 7138 – 7143.

Oncogene