CAD Leading to DNA Fragmentation Caspase-Activated Dnase

Granzyme M Directly Cleaves Inhibitor of Caspase-Activated DNase (CAD) to Unleash CAD Leading to DNA Fragmentation

This information is current as Hongxia Lu, Qiang Hou, Tongbiao Zhao, Honglian Zhang, of September 24, 2021. Qixiang Zhang, Lianfeng Wu and Zusen Fan J Immunol 2006; 177:1171-1178; ; doi: 10.4049/jimmunol.177.2.1171 http://www.jimmunol.org/content/177/2/1171 Downloaded from

References This article cites 50 articles, 23 of which you can access for free at: http://www.jimmunol.org/content/177/2/1171.full#ref-list-1 http://www.jimmunol.org/ Why The JI? Submit online.

• Rapid Reviews! 30 days* from submission to initial decision

• No Triage! Every submission reviewed by practicing scientists

• Fast Publication! 4 weeks from acceptance to publication

by guest on September 24, 2021 *average

Subscription Information about subscribing to The Journal of Immunology is online at: http://jimmunol.org/subscription Permissions Submit copyright permission requests at: http://www.aai.org/About/Publications/JI/copyright.html Email Alerts Receive free email-alerts when new articles cite this article. Sign up at: http://jimmunol.org/alerts

The Journal of Immunology is published twice each month by The American Association of Immunologists, Inc., 1451 Rockville Pike, Suite 650, Rockville, MD 20852 Copyright © 2006 by The American Association of Immunologists All rights reserved. Print ISSN: 0022-1767 Online ISSN: 1550-6606. The Journal of Immunology

Granzyme M Directly Cleaves Inhibitor of Caspase-Activated DNase (CAD) to Unleash CAD Leading to DNA Fragmentation1

Hongxia Lu, Qiang Hou, Tongbiao Zhao, Honglian Zhang, Qixiang Zhang, Lianfeng Wu, and Zusen Fan2

Granzyme (Gzm)M is constitutively highly expressed in NK cells that may play a critical role in NK cell-mediated cytolysis. However, the function of GzmM has been less deﬁned. Just one report showed GzmM induces a caspase-independent death pathway. In this study, we demonstrate a protein transfection reagent Pro-Ject can efﬁciently transport GzmM into target cells. GzmM initiates caspase-dependent apoptosis with typical apoptotic nuclear morphology. GzmM induces DNA fragmentation, not

DNA nicking. GzmM can directly degrade inhibitor of caspase-activated DNase to release the nuclease activity of caspase-activated Downloaded from DNase for damaging DNA. Furthermore, GzmM cleaves the DNA damage sensor enzyme poly(ADP-ribose) polymerase to prevent cellular DNA repair and force apoptosis. The Journal of Immunology, 2006, 177: 1171–1178.

erforin/Granzyme (Gzm)3 pathway is a key mechanism for caspase cascade (13). GzmB can directly process several pro- cytotoxic lymphocytes to kill intracellular pathogens and caspases and some key caspase pathway substrates, such as Bid,

P tumor cells (1, 2). Perforin (also known as pore-forming inhibitor of caspase-activated DNase (ICAD), DNA-PKcs, NuMa, http://www.jimmunol.org/ protein) released from cytotoxic granules assists the entry of Gzms poly(ADP-ribose) polymerase (PARP), and laminB (3). GzmB into the target cell cytosol (3, 4). Gzms are a group of serine causes a caspase-dependent death pathway with DNA fragmenta- proteases that initiate target cell apoptosis. They include GzmA, B, tion. GzmB leads to DNA fragmentation through activation of the C, D, E, F, G, H, K, and M. Nine mouse and five human Gzms caspase-activated DNase (CAD) via degradation of its inhibitor have been identified (5). GzmA and B are the most abundant Gzms ICAD (14, 15). CAD exists as a heterodimer with ICAD in a in CTLs, and lymphokine-activated killer cells and their functions resting cell where CAD is inactive. ICAD is not cleaved in the have been well defined (3, 6). GzmH, K, and M in humans are induction of apoptosis in GzmB-deficient CTLs (16). The mech- called orphan Gzms because their roles are less defined. anism used by GzmB to initiate DNA fragmentation by inactivat-

The five human Gzms differ in their chromosome locations and ing the DNase inhibitor ICAD is reminiscent of the activation of by guest on September 24, 2021 substrate specificities. GzmA and K are tryptases that locate on nick-induced nuclease NM23H1 by GzmA cleavage of its inhibitor chromosome 5q11–12. GzmB is an aspase locating on 14q11.2. SET (12). GzmH is a chymase that is colocated with GzmB. GzmM is a GzmM is an orphan Gzm that cleaves preferentially after me- metase locating on chromosome 19p13.3 (7, 8). GzmA induces thionone, leucine, or norleucine (17, 18). GzmM is constitutively ssDNA nicks as well as apoptotic morphology and loss of cell highly expressed in NK cells, whereas it is not expressed in CD4ϩ membrane integrity (9). GzmA initiates a caspase-independent or CD8ϩ T cells either constitutively or after stimulation (19). One death pathway through targeting an endoplasmic reticulum-asso- study compared the cytolysis of three NK cell lines and showed ciated SET complex that contains three GzmA substrates: the nu- that stronger cytotoxicity of NK cell lines is consistent with higher cleosome assembly protein SET, the DNA-bending protein expression of GzmM (20). It suggests GzmM may play an impor- HMG2, and the apurinic endonuclease Ape1 (10, 11). SET cleav- tant role in NK cell-mediated cytolysis for killing virally infected age releases the GzmA-activated DNase (NM23H1) to nick DNA or transformed cells. A recent study reported that GzmM induces (12). The execution of apoptosis by GzmB generally occurs by a novel form of perforin-dependent death without caspase activa- activation of the caspase family of cysteine proteases that triggers tion and DNA fragmentation (21). In this study, we found that GzmM triggers caspase-dependent apoptosis with typical DNA laddering. The DNA fragmentation was confirmed by TUNEL as- National Laboratory of Biomacromolecules and Center for Infection and Immunity, Institute of Biophysics, Chinese Academy of Sciences, Beijing, China say. GzmM can directly cleave ICAD to activate CAD, leading to Received for publication January 17, 2006. Accepted for publication April 26, 2006. DNA damage. Furthermore, GzmM degrades the DNA damage The costs of publication of this article were defrayed in part by the payment of page sensor enzyme PARP to prevent cellular DNA repair and force charges. This article must therefore be hereby marked advertisement in accordance apoptosis. with 18 U.S.C. Section 1734 solely to indicate this fact. 1 This work was supported by the National Natural Science Foundation of China and Materials and Methods the Outstanding Youth Grant (30525005 and 30470365) and the Hundred Talents Cell lines, Abs, and reagents Program of Chinese Academy of Sciences (to Z.F.). 2 Address correspondence and reprint requests to Dr. Zusen Fan, National Laboratory Jurkat cells were grown in RPMI 1640 medium supplemented with 10% of Biomacromolecules and Center for Infection and Immunity, Institute of Biophys- FCS, 50 mM 2-ME, 100 U/ml penicillin, and 100 ␮g/ml streptomycin. ics, Chinese Academy of Sciences, 15 Datun Road, Beijing 100101, China. E-mail HeLa cells were grown in DMEM medium containing 10% FCS, 20 mM address: [email protected] L-glutamine, 100 U/ml penicillin, and 100 ␮g/ml streptomycin. Commer- 3 Abbreviations used in this paper: Gzm, granzyme; Ad, adenovirus; CAD, caspase- cial Abs were rabbit antisera against caspase 3 (BD Pharmingen) and activated DNase; ICAD, inhibitor of CAD; PARP, poly(ADP-ribose) polymerase; PI, PARP (Cell Signaling Technology), mouse mAb against His-tag (Sigma- propidium iodide; PJ, Pro-Ject protein transfection reagent; STP, staurosporine. Aldrich), HRP-conjugated sheep anti-mouse IgG and HRP-conjugated

sheep anti-rabbit IgG (Santa Cruz Biotechnology), and Alexa 488-conju- Jurkat cells treated as above were fixed, permeabilized as described, and gated donkey anti-mouse IgG (Molecular Probes). Rabbit antiserum incubated with 5 ␮g/ml Hoechst 33342 (Sigma-Aldrich) in PBS for 10 min against ICAD was provided by Dr. R. P. Sekaly (Laboratoire in dark and plated on the slide for observation. Images were captured using d’Immunologie, Universite´ de Montreal, Montreal, Canada). Annexin a Nikon microscope camera. VFITC was from BD Pharmingen, and PI from Sigma-Aldrich. ProLong Antifade kit was from Molecular Probes. In Situ cell death detection kit Klenow incorporation assay was from Roche Applied Science. Klenow fragment of DNA polymerase I (New England Biolabs) was used to label DNA nicks, as described (22). Briefly, 2 ϫ 105 Jurkat cells were Production of active and inactive GzmM, rICAD, and rPARP buffer treated or treated with 1 ␮M S-AGzmM or GzmM pretreated with NPase lysis buffer for 4 h. Treated cells were incubated with5UofKlenow The cDNA encoding mature human GzmM was PCR amplified from full- and 10 ␮Ci of [32P]dATP at 37°C for 1 h. Radiolabeled nuclei were washed length cDNA of GzmM (RZPD Deutsches Ressourcenzentrum) and sub- thoroughly, dotted, and detected with phosphor screen. The radioactivity cloned into the yeast expression vector pPICZ␣A (Invitrogen Life Tech- counts were recorded with Storm program (Pharmacia) and analyzed with nologies). The sequence of the forward primer containing XhoI site was ImageQuant software. The values shown were comparative [32P]dATP in- 5Ј-ACTCTCGAGAAAAGAATCATCGGGGGCCGGGAGGTG-3Ј. The corporations and expressed as mean Ϯ SEM. Radiolabeled DNA was also reverse primer containing XbaI site and polyhistidine tag for purification separated on 1% alkaline agarose gel electrophoresis, as described (12). was 5Ј-ATCTCTAGATCAATGATGGTGGTGATGATGGGCCGATC GGCCGGTGACCTTC-3Ј. The resulting construct permitted the sequence DNA fragmentation assay of mature human GzmM to immediately follow the Kex2 signal cleavage site of the Saccharomyces cerevisiae ␣-factor secretion signal. The vector A total of 5 ϫ 105 Jurkat cells lysed with 0.5% Nonidet P-40 buffer was was linearized with SacI and transformed into the X33 strain of Pichia buffer treated or treated with 1 ␮M S-AGzmM and indicated doses of ␮ pastoris, according to the manufacturer’s instruction. Clones with the in- GzmM in 50 l of HBSS supplemented with 1 mM MgCl2, 1 mM CaCl2,

tegrated human GzmM cDNA were selected by resistance to Zeocin (In- and 1 mg/ml BSA at 37°C for4hinthepresence or absence of ZVAD- Downloaded from vitrogen Life Technologies). After 3 days of induction with methanol, the fmk. Genomic DNA was extracted and analyzed by agarose gel electro- conditioned medium from the shaker flask cultures was isolated and loaded phoresis. Briefly, after incubation, 330 ␮l of buffer containing 100 mM onto the QuiStand benchtop system (Amersham Biosciences) for concen- Tris-HCl (pH 8.5), 5 mM EDTA, 0.2 M NaCl, 0.2% w/v SDS, and 0.2 tration and buffer exchange. One-step purification was performed using mg/ml proteinase K was added to each reaction and incubated at 37°C nickel affinity chromatography. The concentration of rGzmM was deter- overnight. NaCl was then added to a final concentration of 1.5 M, and the mined by bicinchoninic acid assay. GzmM catalytic site Ser182 was mu- nuclear debris was spun down for 15 min in a microcentrifuge at room tated to alanine by PCR mutagenesis. The mutated S-AGzmM sequence temperature. The DNA in the supernatant was precipitated with an equal was subcloned to pPICZ␣A as GzmM and verified by DNA sequencing. volume of 100% ethanol. The DNA precipitate was washed once with 70% http://www.jimmunol.org/ Recombinant S-AGzmM was expressed and purified, as above. Full-length (v/v) ethanol and resuspended in 20 ␮l of TE buffer (10 mM Tris⅐HCl (pH cDNA coding human ICAD was provided by Dr. X. Wang (Department of 7.4), 1 mM EDTA) supplemented with 200 ␮g/ml DNase-free RNase A at Molecular Biology and Oncology, University of Texas Southwestern Med- 37°C for 2 h, the DNA was loaded onto a 2% agarose gel with 2 ␮g/ml ical Center, Dallas, TX) and was subcloned into pET26bϩ. Full-length ethidium bromide, and electrophoresis was conducted at 5 V/cm for 2 h. cDNA of human PARP was subcloned into pET15b. Recombinant human Genomic DNA of Jurkat cells treated with 1 ␮M GzmM or S-AGzmM ICAD and PARP were expressed in BL21(DE3) and purified using nickel alone or with PJ at 37°C for4hinthepresence or absence of 100 ␮M affinity chromatography. ZVAD-fmk was also extracted and analyzed by agarose gel electrophoresis. GzmM loading with PJ Laser-scanning confocal microscopy by guest on September 24, 2021 Pro-Ject protein transfection reagent (PJ) kit (Pierce) was prepared and ali- Jurkat cells were treated with 1 ␮M S-AGzmM or GzmM in the presence quoted, according to the manufacturer’s instruction. GzmM or S-AGzmM was or absence of PJ for the indicated times, washed, and plated on polylysine- diluted in loading buffer (HBSS, 1 mM MgCl2, 1 mM CaCl2, and 1 mg/ml coated slides. After fixing with 4% paraformaldehyde for 20 min at room BSA) and incubated with appropriate volume of PJ at room temperature for temperature and permeabilizing with 0.1% Triton X-100 for another 20 5 min. Target cells were harvested and washed twice with cold HBSS. The ␮ ␮ min, the cells were incubated at room temperature for 1 h with 5 g/ml indicated number of cells was cultured with PJ/GzmM complex in 50 lof anti-His-tag mAb, 50 ␮g/ml donkey serum, and 100 ␮g/ml RNase I. After loading buffer at 37°C for 4 h. Cells were pelleted for subsequent assays. washing twice with cold PBS, the cells were stained with Alexa 488-conjugated donkey anti-mouse IgG and soaked for 5 min in PBS containing 0.1 Cytotoxicity assay ␮g/ml PI. The slides were mounted with ProLong Antifade reagent and Target cells were labeled with 100 ␮Ci of sodium 51Cr-chromate observed using laser-scanning confocal microscopy (Olympus FV500 ϫ 4 microscope). (Na2CrO4) at 37°C for 1 h. After being thoroughly washed, 1 10 - labeled cells were incubated with 1 ␮M S-AGzmM or GzmM alone or with Cleavage assay PJ in the presence or absence of 100 ␮M ZVAD-fmk at 37°C for 4 h. The supernatant from each tube was carefully transferred into a disposable tube. Cell lysates were prepared with 0.5% Nonidet P-40 lysis buffer at 4°C for Specific 51Cr release was counted on the 1450 MicroBeta counter (Phar- 15 min and centrifuged at 13,000 rpm for 15 min to remove nuclei and cell macia) and was calculated as follows: percentage of 51Cr release ϭ ((test debris. Cell lysates (equivalent to 2 ϫ 105 cells) or 0.5 ␮g of recombinant release Ϫ spontaneous release)/(maximum release Ϫ spontaneous protein were incubated with indicated doses of GzmM or S-AGzmM for release)) ϫ 100. indicated time intervals. For in vivo cleavage assay, cells loaded with GzmM/PJ were lysed with 0.5% Nonidet P-40 lysis buffer. The reactions Detection of apoptosis were terminated in 1ϫ SDS loading buffer before being run on SDS-PAGE gels and immunoblotted with specific Abs, as indicated. Jurkat cells were treated with 1 ␮M GzmM or S-AGzmM alone or with PJ For N-terminal sequencing, the cleaved products of rICAD were run on at 37°C for4hinthepresence or absence of 100 ␮M ZVAD-fmk. Jurkat SDS-PAGE gels and transferred to polyvinylidene difluoride membrane. cells cultured with 50 ng/ml staurosporine (STP) for 20 h served as a After staining with Coomassie R-250, the polyvinylidene difluoride band positive control. Treated cells were double stained with annexin V-Fluos was cut down and analyzed using ABI Procise 492cLC sequencer. The data and propidium iodide (PI) before being evaluated by flow cytometry using were analyzed using Model 610 software program supplied by Applied a FACSCalibur (BD Biosciences). Analysis was performed using Biosystems. CellQuest software (BD Biosciences). Flow-TUNEL assay with In Situ cell death detection kit (Roche Applied Results Science) was performed on treated cells, according to the manufacturer’s PJ can directly load GzmM into target cells causing apoptotic instruction. Briefly, treated cells were washed three times with PBS, fixed with 2% paraformaldehyde in PBS for1hatroom temperature, and per- nuclear morphological changes meabilized with freshly prepared 0.1% Triton X-100 in 0.1% sodium ci- Recombinant human active GzmM and inactive S-AGzmM were trate for 2 min on ice. A total of 50 ␮l of TUNEL reaction mixtures was added into each sample and incubated for1hat37°C in a humidified expressed in P. pastoris and purified by nickel affinity chromatog- atmosphere in the dark. After washing twice, cells were visualized using raphy. The enzymatic activity of rGzmM was detected by hydro- immunofluorescent microscopy. lysis of its synthetic tetrapeptide substrate Suc-AAPL-pNA (7). The Journal of Immunology 1173

rGzmM showed strong proteolytic activity even at low nanomolar By 6 h, the nuclei treated by GzmM with PJ occurred segmented concentration, whereas the mutant form S-AGzmM had no activity and collapsed. PJ and GzmM alone failed to make any change as (data not shown). mock-treated cells. Hoechst DNA staining obtained the similar ap- A recent report showed GzmM induces a novel form of cell optotic nuclear morphology in GzmM- and PJ-loaded cells (Fig. death that depends on perforin (21). Perforin can assist all the 1B). PJ or GzmM alone was without effect. Apoptosis-inducing Gzms into target cells to trigger death by a variety of assays (6, 23, agent STP-treated cells appeared characteristic apoptosis as a pos- 24). Perforin transfers Gzms into the target cells through endocy- itive control. Similar results were obtained by using Ad for GzmM tosis and causes release of Gzms into the cytosol of target cells (4). loading (data not shown). However, in vitro GzmB-loading experiments showed GzmB enters a target cell and induces apoptosis independently of perforin (25, 26). Bleackley and colleagues (25, 27) demonstrated that a GzmM induces a caspase-dependent death pathway replication-deﬁcient adenovirus type V (Ad) can efﬁciently trans- Kelly et al. (21) showed GzmM loaded with perforin caused 51 port GzmB into target cells and trigger cytolysis. We wanted to Cr release and higher annexin Vϩ/PIϩ double-positive cells and com- determine whether PJ can substitute for perforin or Ad to directly ϩ deliver Gzms into the target cell cytosol and induce cell death. PJ parable single annexin V cells. To detect the cytolysis by PJ is used to deliver biologically active proteins directly into a living direct loading of GzmM, we analyzed cytotoxicity in GzmM/PJ- 51 cell. Proteins loaded by PJ remain biologically active because the treated cells by a standard 4-h Cr release assay. In three inde- interaction between a protein and PJ is noncovalent. We loaded pendent experiments, Jurkat cells loaded with GzmM plus PJ in-

GzmM into Jurkat cells in the presence of an optimal dose of PJ duced 74 Ϯ 8% speciﬁc cytolysis (Fig. 2A). GzmM and PJ alone Downloaded from and observed the entry of GzmM. GzmM transferred into the cy- or S-AGzmM and PJ just got background lysis (generally Ͻ10%). tosol and the nuclei of treated cells within 4 h (Fig. 1A). The nuclei Similar results were found by using HeLa cells (data not shown). appeared condensed, shrunken, and with morphological changes. Jurkat cells loaded with different concentrations of GzmM plus PJ led to dynamic increase of cell death, as shown in Fig. 2B.Atthe concentration of 0.2 ␮M GzmM, cells began to undergo death ϩ ϩ ϩ (6.9% single annexin V and 7.4% annexin V /PI ). At 1 ␮M http://www.jimmunol.org/ GzmM, death rate increased to the equivalent rate of annexin Vϩ/ PIϩ cells by Kelly’s (21) observations (33.4 vs 36%). Moreover, 1 ␮M GzmM induced high rate of single annexin Vϩ cells (24.9%). By contrast, Kelly et al. (21) showed GzmM did not initiate single annexin Vϩ cells. PJ and GzmM alone or inactive S-AGzmM/PJ just got background annexin Vϩ/PIϩ and single annexin Vϩ cells. These results are representative of at least three independent

experiments. by guest on September 24, 2021 To determine whether GzmM/PJ-induced death is dependent on caspase activation, cells were preincubated with the pan caspase inhibitor ZVAD-fmk before GzmM loading. A total of 100 ␮M ZVAD-fmk blocked GzmM/PJ-induced specific lysis measured by a standard 4-h 51Cr release assay (14 Ϯ 2vs74Ϯ 8%) (Fig. 2A) and annexin V/PI staining (data not shown). These data are representative of at least three separate experiments. To further verify GzmM-induced death is caspase dependent, we analyzed caspase 3 cleavage by immunoblotting. Caspase activation is a key regulator of the apoptotic pathway for multiple inducers of cell death (28). Procaspases are processed at aspartic acid residues between the subunits of a heterotetramer and can be activated by other caspases or by GzmB (29, 30). Caspase 3 is a central executioner in the induction of apoptosis. Caspase 3 can be activated by processing its proenzyme procaspase 3 undergoing apoptosis. We found procaspase 3 was processed in GzmM-treated cell lysates with a dose- and time-dependent fashion (Fig. 3A). Caspase 3 began to be processed with a low dose of 10 nM GzmM at very early time of 5 min. The active form p17 band was detect- able at high dose of 1 ␮M GzmM, which was comparable to 0.3 ␮ FIGURE 1. GzmM is transported into the cytosol and the nuclei of tar- M GzmB treatment. The cleavage patterns by GzmM and B are get cells, leading to typical apoptotic nuclear morphology. A, Jurkat cells different. Inactive S-AGzmM was without effect. To verify the were treated with 1 ␮M GzmM in the presence of an optimal dose of PJ for caspase activation is physiologically relevant, Jurkat cells were the indicated times or left untreated. The cells were fixed, permeabilized, loaded with GzmM in the presence of PJ. Procaspase 3 was and incubated with anti-His6 mAb, followed by staining with the second cleaved to produce the active form, a p17 product in GzmM/PJ- Alex 488-conjugated goat anti-mouse Ab and PI. GzmM staining green loaded cells (Fig. 3B). PJ or GzmM alone or S-AGzmM/PJ failed fluorescence was shown at left, PI staining red in the middle, and the to cut procaspase 3. ␤-Actin was unchanged as a loading control. merged image at right. B, After incubating with PJ, 1 ␮M GzmM, or 1 ␮M GzmM/PJ at 37°C for 4 h, Jurkat cells were stained with Hoechst 33342. Taken together, GzmM-induced death is required for caspase STP-treated cells were used as a positive control. activation. 1174 GzmM-INDUCED DNA FRAGMENTATION

FIGURE 2. GzmM induces cell death that can be blocked by the pan caspase inhibitor ZVAD-fmk. A, 51Cr-labeled Jurkat cells were incubated with 1 ␮M GzmM or S-AGzmM for 4 h in the presence or absence of PJ. Speciﬁc 51Cr release was counted. For ZVAD treatment, cells were pretreated with 100 ␮M ZVAD-fmk at 37°C for 30 min before loading with 1 ␮M GzmM/PJ. The values are mean Ϯ SEM. B, Jurkat cells were treated with PJ, 1 ␮M GzmM, 1 ␮M S-AGzmM/PJ, or different doses of GzmM plus PJ for 4 h, followed by staining with annexin V and PI. Buffer and STP-treated cells served as negative and positive controls, respectively. Results shown are representative of three Downloaded from separate experiments. http://www.jimmunol.org/

GzmM fails to nick DNA GzmM directly cleaves the ICAD DNA damage is another feature of cell death. GzmA induces CAD, also known as DFF40, has been verified to be directly re- caspase-independent death with ssDNA nicking (6, 9). DNA nick- sponsible for DNA fragmentation. CAD exists as a heterodimer ing is supposed to be an early stage of cell death induced by cy- with its inhibitor ICAD (also called DFF45) that serves as both a tolytic lymphocytes. GzmM initiates a faster death pathway of specific molecular chaperone to mediate the correct folding of target cells consistent with Kelly’s report (21). We next wanted to DFF40 and an inhibitor of the nuclease activity of DFF40 in rest- by guest on September 24, 2021 look at whether GzmM initiates ssDNA nicking. The Klenow ing cells (31–33). CAD can be activated by cleavage of its inhib- polymerase labeling was used to assess DNA strand breaks, as itor ICAD to initiate DNA fragmentation after a cell receives a described (12). DNA nicks were not induced from GzmM-treated death signal. To further determine whether ICAD is degraded by Jurkat nuclei, even with 1 ␮M GzmM (Fig. 4A), whereas extensive GzmM treatment, ICAD was probed by immunoblotting with anti- DNA nicking occurred with 0.5 ␮M GzmA treatment under the ICAD Ab. ICAD was cleaved to generate a predominant 30-kDa same condition, which is in concert with our previous reports (12). fragment in GzmM-treated cell lysates in a dose-dependent man- PJ- or S-AGzmM-treated nuclei were not nicked. These results ner (Fig. 6A). S-AGzmM had no proteolytic activity. ICAD began were confirmed via denaturing alkaline gel electrophoresis, as to degrade within 30 min and was almost completely processed by shown in Fig. 4B. These data are representative of three separate 2 h with 0.5 ␮M GzmM treatment. S-AGzmM did not cause ICAD experiments. degradation. To investigate whether ICAD degradation occurs in GzmM-loaded cells, ICAD was probed in Jurkat cells loaded with GzmM triggers DNA fragmentation GzmM plus PJ. ICAD was cleaved in GzmM/PJ-loaded cells (Fig. To determine the form of DNA damage in GzmM-induced death, 6B). GzmM or PJ alone or S-AGzmM plus PJ failed to degrade genomic DNA was extracted from GzmM-treated nuclei. Charac- ICAD. The same blot was stripped and reprobed for ␤-actin. ␤-Ac- teristic apoptotic DNA ladders were visualized on agarose gels in tin was unchanged as a good loading control. GzmM-treated nuclei (Fig. 5A). DNA laddering occurred in a To further evaluate whether ICAD can be directly processed by dose-dependent manner. Mock- or S-AGzmM-treated nuclei were GzmM, human rICAD was treated with different concentrations of without effect. To further assess the DNA fragmentation induced GzmM. GzmM can directly cut rICAD to produce 11- and 30-kDa by GzmM is physiologically relevant, GzmM was loaded into fragments beginning at low nanomolar concentrations (Fig. 6C). Jurkat cells in the presence of PJ. Genomic DNA was extracted rICAD was completely degraded with 50 nM GzmM. rICAD and assayed to visualize DNA fragmentation from GzmM-loaded cleavage occurred very rapidly within 5 min at 100 nM concen- cells. DNA ladders appeared in GzmM/PJ-treated cells (Fig. 4B). tration of GzmM. rICAD was completely degraded by 15 min. GzmM and PJ alone or S-AGzmM/PJ failed to damage DNA. A S-AGzmM did not lead to rICAD degradation. Caspase 3 was used total of 100 ␮M ZVAD-fmk nearly blocked GzmM-induced DNA as a positive control. We performed N-terminal sequencing of the laddering that is consistent with the above observations. To further cleaved products and showed the cleavage site was after Ser107 confirm DNA damage induced by GzmM, fluorescein-dUTP-la- (TAWIS2QESFD). This is different from GzmB. beled TUNEL was used to detect the free 3Ј-OH DNA ends at a single cell level by fluorescence microscopy. GzmM/PJ-treated PARP is degraded by GzmM cells appeared TUNEL positive with strong staining (Fig. 5C). PJ In cells, a NADϩ-dependent signal transduction mechanism serves or GzmM alone was TUNEL negative. to protect cells against the genome-destabilizing effects after DNA The Journal of Immunology 1175

FIGURE 3. Caspase 3 is activated after GzmM treatment. A, Jurkat cell lysates (equivalent to 2 ϫ 105 cells) were treated with different concentrations of GzmM from 0.01 to 1.0 ␮M(lanes 3–8) or GzmB from 0.3 to 0.5 ␮M(lanes 9–12)for2h(top panel), or with 1.0 ␮M GzmM for the indicated times (bottom panel). The reaction products were probed with Downloaded from anti-caspase 3 Ab. B, A total of 2 ϫ 105 Jurkat cells was treated with buffer, 1.0 ␮M GzmM, or S-AGzmM alone or with PJ at 37°C for 4 h. The whole cell lysates were electrophoresed on a 15% SDS-PAGE gel and probed for caspase 3. ␤-Actin served as a negative control. breaks. The mechanism involves two nuclear enzymes, PARP-1 http://www.jimmunol.org/ and PARP-2, that sense DNA strand breaks. When activated by DNA breaks, these PARPs catalyze the transfer of ADP-ribose moiety from NADϩ to nuclear target substrates. Through recruit- ment of speciﬁc proteins at the site of damage and regulation of their activities, these polymers may either directly participate in the DNA repair process or coordinate repair through chromatin unfolding, cell cycle progression, and cell death (34, 35). PARP is FIGURE 5. GzmM mediates DNA fragmentation. A, The isolated Jur- a sensor of DNA damage, which is a downstream substrate of by guest on September 24, 2021 kat nuclei were untreated or treated with 1 ␮M S-AGzmM, or 0.5 ␮Mor caspase 3 and can be directly degraded by GzmB (36, 37). We 1 ␮M GzmM, respectively. Genomic DNA was extracted and visualized by wanted to determine whether PARP is degraded after GzmM treat- 2% agarose gel electrophoresis with ethidium bromide staining. B, Jurkat ment. Human rPARP was incubated with different doses of GzmM cells were incubated with 1 ␮M GzmM or S-AGzmM alone or with PJ for at 37°C for 2 h. PARP was cleaved beginning at a low concentra- 4 h. Cells were pretreated with 100 ␮M ZVAD-fmk at 37°C for 30 min tion of 30 nM GzmM (Fig. 7A). A total of 50 nM GzmM can before being loaded with GzmM/PJ, while DMSO pretreatment was used as a control. C, Jurkat cells were incubated with PJ, 1 ␮M GzmM, or 1 ␮M GzmM/PJ at 37°C for 4 h. Treated cells were ﬁxed and permeabilized, and TUNEL staining was performed. The images shown are representative of three independent experiments.

completely degrade PARP. PARP cleavage appeared rapidly within 15 min by 50 nM GzmM treatment. To further investigate whether PARP is a physiological substrate of GzmM, Jurkat cells were loaded with GzmM plus PJ and probed by immunoblotting with anti-PARP serum. PARP was completely cleaved to generate a 45-kDa product in GzmM-loaded cells within 4 h (Fig. 6B). GzmM and PJ alone or S-AGzmM/PJ was without effect. There- fore, PARP is a direct physiological substrate of GzmM.

Discussion Perforin was originally thought to generate pores in the target cell plasma membrane through which Gzms would get into cytosol (38, 39). However, this model for Gzm entry has been questioned. At FIGURE 4. GzmM does not cause ssDNA nicking. A, Jurkat nuclei low concentrations, perforin delivers Gzms to trigger apoptosis, were treated with GzmM (1 ␮M) or GzmA (1 ␮M) at 37°C for4hand and at high concentrations perforin induces necrotic death. Actu- radiolabeled by Klenow DNA polymerase. The radioactivity counts were determined, as described in Materials and Methods. The values shown ally, at a sublytic concentration the plasma membrane integrity is represent mean Ϯ SEM. B, Genomic DNA treated above was separated not disrupted; small dyes do not enter the cell (40, 41). Froelich et with 1% denaturing alkaline gel electrophoresis and visualized by autora- al. (25–27) showed that a replication-deficient Ad can efficiently diography. The data shown are representative of three separate transport GzmB into targets and lead to apoptosis. Recent reports experiments. verified Gzms are taken into cells by endocytosis and perforin or 1176 GzmM-INDUCED DNA FRAGMENTATION

FIGURE 6. ICAD is directly cleaved by GzmM. A, Jurkat cell lysates (equivalent to 2 ϫ 105 cells) were incubated with different doses of GzmM at 37°C for2h(lanes 1–6) or with 1.0 ␮M GzmM for the indicated times (lanes 7–12). Untreated or 1.0 ␮M S-AGzmM-treated sample served as a negative control. The products were electrophoresed on SDS gels and probed by anti- ICAD Ab. B,2ϫ 105 Jurkat cells were treated with 1.0 ␮M GzmM, S-AGzmM alone, or PJ at 37°C for 4 h. The whole cell lysates were immunoblotted with anti-ICAD serum. ␤-Actin was unchanged as a negative control. C, rICAD (0.5 ␮g) was treated with different doses of GzmM ranging from 5 to 100 nM (lanes 1–7) and 300 nM recombinant active caspase 3 for 2 h (lane 8). A total of 0.5 ␮g of rICAD was incubated with 100 nM GzmM for the indicated times (lanes 9–12). The reaction products were Downloaded from probed, as described above.

Ad causes release of Gzms into the cytosol (4, 42). Transient ex- number of caspases are processed by GzmB, caspases 3 and 7 are

pression of GzmB or direct transfer of GzmB by microinjection cleaved most rapidly in cells loaded GzmB with perforin (44, 45). http://www.jimmunol.org/ can initiate cell death (27, 43). It indicates that GzmB exerts its Caspase 3 is a central execution member of the caspase family. proapoptotic function inside a target cell independently of perforin. DNA damage is a key feature of cell death. We previously In this study, we showed a cationic lipid-based PJ can efficiently showed GzmA triggers cell death with ssDNA nicks by targeting transport GzmM into target cells and induce cell death. PJ binds to an endoplasmic reticulum-associated SET complex (6, 9). The proteins without covalent formation that moves proteins into cells, SET complex contains the GzmA-activated DNase NM23H1 that where proteins remain their physiological activities. GzmM alone binds to its inhibitor SET in a resting cell. Cleavage of SET by did not enter the target cells. GzmM entry needs assistance of PJ GzmA unleashes the nuclease activity of NM23H1 to nick DNA or Ad that is consistent with perforin-dependent entry of GzmM (12). DNA nicking is thought to be an initial stage of death induced (21). Actually, an optimal dose of PJ was used to load GzmA or B through cytotoxic lymphocytes, followed by DNA fragmentation by guest on September 24, 2021 into target cells, which led to similar apoptotic features as perforin- by other Gzms. In addition to GzmA, GzmC and K have been or Ad-delivered GzmA or B (data not shown). reported to cause DNA nicks (23, 46). Only GzmB has been de- A recent study showed that GzmM causes a third major death fined to make DNA fragmentation (3). We found GzmM did not pathway besides GzmA and B (21). GzmM-induced death is in- cause DNA nicking, but led to typical DNA laddering. The DNA dependent of caspase activation and mitochondrial damage. fragmentation induced by GzmM was verified by TUNEL assay. GzmM does not trigger chromosomal DNA fragmentation by an- To further gain insight into the mechanism, we demonstrated that alyzing DNA content via PI staining. However, we found that ICAD/DFF45 can be directly cleaved by GzmM. CAD/DFF40 has GzmM initiates a caspase-dependent death pathway. The pan been defined to be directly responsible for DNA fragmentation. caspase inhibitor ZVAD (100 ␮M) can completely block GzmM- CAD remains in an inactive form with its inhibitor ICAD of nor- induced cytolysis, DNA fragmentation, and procaspase 3 process- mal cells (31–33). Once a cell receives a death signal, caspase 3 is ing. Our observations are different from those of Kelly et al. (21). activated and it in turn cleaves ICAD, which activates the nuclease They showed that even high concentration of 200 ␮M ZVAD activity of CAD. ICAD acts as both an inhibitor of CAD and a failed to inhibit GzmM-induced death. GzmM-treated cells did not specific nuclear chaperone to assist the correct folding of CAD. cause caspase activation and phosphatidylserine externalization. ICAD-deficient mice have a reduced ability to make DNA frag- We found GzmM initiated high rate of single annexin Vϩ cells and mentation, which is partially resistant to in vitro and in vivo in- caspase activation. Although we used the different concentrations duced apoptosis (16). These indicate that CAD/ICAD complex of GzmM and different delivery agents, we obtained the similar plays an essential role in DNA fragmentation undergoing apopto- death rates by GzmM (1 ␮M)/PJ and GzmM (83 nM)/perforin sis. ICAD degradation by GzmB releases the nuclease activity of treatment. It indicates that the 1 ␮M concentration of GzmM we CAD to result in DNA fragmentation (14, 15). GzmM also directly used is equal in activity to 83 nM concentration used by Smyth degrades ICAD to activate CAD activity. ICAD was cleaved to group. Different GzmM activity might be due to the different ex- generate 11- and 30-kDa fragments, which were from one cut after pression systems. We produced GzmM in P. pastoris yeast, Ser107 (TAWIS2QESFD). The cleavage pattern by GzmM is dif- whereas Smyth group generated GzmM from baculovirus sf21 ferent from that of GzmB, producing 28- and 31-kDa fragments at cells. The same death rates induced by GzmM (1 ␮M)/PJ and residues of D117 and D224. It has been reported GzmM prefers to GzmM (83 nM)/perforin might be also attributable to the delivery cleave after methionine, leucine, or norleucine by measuring of efficiency of the two different delivery agents. Caspases are central synthetic peptides (18). The cleavage sites of GzmM for physio- regulators of apoptotic pathway for multiple inducers of cell death logical substrates remain to be further investigated. (13). Caspases are synthesized as a single polypeptide, act as a PARP is a sensor enzyme of DNA damage. It has three func- heterotetramer consisting of two large and two small subunits, and tional domains, an N-terminal DNA binding domain, a C-terminal can be initiated by other caspases or GzmB (29, 30). Although a catalytic domain, and an automodification domain in the middle The Journal of Immunology 1177

function of NK cells will provide the basis for their potential clin- ical applications.

Acknowledgments We thank Drs. X. Wang and R. P. Sekaly for their reagents, and Chunchun Liu and Yan Teng for their technical help. Disclosures The authors have no ﬁnancial conﬂict of interest. References 1. Russell, J. H., and T. J. Ley. 2002. Lymphocyte-mediated cytotoxicity. Annu. Rev. Immunol. 20: 323–370. 2. Lieberman, J. 2003. The ABCs of granule-mediated cytotoxicity: new weapons in the arsenal. Nat. Rev. Immunol. 3: 361–370. 3. Trapani, J. A., and V. R. Sutton. 2003. Granzyme B: pro-apoptotic, antiviral and antitumor functions. Curr. Opin. Immunol. 15: 533–543. 4. Keefe, D., L. Shi, S. Feske, R. Massol, F. Navarro, T. Kirchhausen, and J. Lieberman. 2005. Perforin triggers a plasma membrane-repair response that facilitates CTL induction of apoptosis. Immunity 23: 249–262. 5. Grossman, W. J., P. A. Revell, Z. H. Lu, H. Johnson, A. J. Bredemeyer, and

T. J. Ley. 2003. The orphan granzymes of humans and mice. Curr. Opin. Im- Downloaded from munol. 15: 544–552. ␮ 6. Lieberman, J., and Z. Fan. 2003. Nuclear war: the granzyme A-bomb. Curr. FIGURE 7. PARP is a direct substrate of GzmM. A, 0.5 g of rPARP Opin. Immunol. 15: 553–559. was untreated or treated with different doses of GzmM ranging from 0.01 7. Mahrus, S., and C. S. Craik. 2005. Selective chemical functional probes of gran- to 0.5 ␮Mfor2hat37°C (upper panel), or with 0.3 ␮M GzmM for 15–120 zymes A and B reveal granzyme B is a major effector of natural killer cell- min (lower panel). The products were run on SDS-PAGE gels and probed mediated lysis of target cells. Chem. Biol. 12: 567–577. ϫ 5 ␮ 8. Pilat, D., T. Fink, B. Obermaier-Skrobanek, M. Zimmer, H. Wekerle, P. Lichter, for PARP. B, A total of 2 10 Jurkat cells was treated with 1.0 M and D. E. Jenne. 1994. The human Met-ase gene (GZMM): structure, sequence, http://www.jimmunol.org/ GzmM or S-AGzmM and/or PJ at 37°C for 4 h. The products were im- and close physical linkage to the serine protease gene cluster on 19p13.3. Genom- munoblotted with anti-PARP serum. NS, Nonspeciﬁc band. ics 24: 445–450. 9. Fan, Z. S., and Q. X. Zhang. 2005. Molecular mechanisms of lymphocyte-mediated cytotoxicity. Cell Mol. Immunol. 2: 259–264. 10. Fan, Z., P. J. Beresford, D. Zhang, and J. Lieberman. 2002. HMG2 interacts with part. Once DNA breaks, PARP is activated to bind to the DNA the nucleosome assembly protein SET and is a target of the cytotoxic T-lym- strand interruptions through its DNA binding domain. Activated phocyte protease granzyme A. Mol. Cell. Biol. 22: 2810–2820. 11. Fan, Z., P. J. Beresford, D. Zhang, Z. Xu, C. D. Novina, A. Yoshida, Y. Pommier, PARP catalyzes formation of long and branched polymers of and J. Lieberman. 2003. Cleaving the oxidative repair protein Ape1 enhances cell ADP-ribose using NAD as a substrate (47). These polymers recruit death mediated by granzyme A. Nat. Immunol. 4: 145–153. speciﬁc proteins at the DNA damage site and enhance their activ- 12. Fan, Z., P. J. Beresford, D. Y. Oh, D. Zhang, and J. Lieberman. 2003. Tumor

suppressor NM23–H1 is a granzyme A-activated DNase during CTL-mediated by guest on September 24, 2021 ities to participate in DNA repair or coordinate repair via regula- apoptosis, and the nucleosome assembly protein SET is its inhibitor. Cell 112: tion of other cellular DNA-related functions (34, 35). PARP is 659–672. specifically activated at a very early stage of apoptosis, while the 13. Los, M., S. Wesselborg, and K. Schulze-Osthoff. 1999. The role of caspases in development, immunity, and apoptotic signal transduction: lessons from knock- cell continues to retain membrane integrity and the majority of the out mice. Immunity 10: 629–639. proteins are not cleaved (48). PARP is directly processed to pro- 14. Sharif-Askari, E., A. Alam, E. Rheaume, P. J. Beresford, C. Scotto, K. Sharma, D. Lee, W. E. DeWolf, M. E. Nuttall, J. Lieberman, and R. P. Sekaly. 2001. duce a 89-kDa product by caspase 3-like activity undergoing ap- Direct cleavage of the human DNA fragmentation factor-45 by granzyme B in- optosis (37), whereas PARP is cleaved to generate 89- and 64-kDa duces caspase-activated DNase release and DNA fragmentation. EMBO J. 20: fragments in GzmB-mediated cytolysis. The dual pattern for cleav- 3101–3113. 15. Thomas, D. A., C. Du, M. Xu, X. Wang, and T. J. Ley. 2000. DFF45/ICAD can age of PARP by GzmB indicates GzmB not only activates be directly processed by granzyme B during the induction of apoptosis. Immunity caspases, but also enters the nucleus to directly degrade PARP. 12: 621–632. GzmM can directly cleave PARP to produce a 45-kDa fragment, 16. Zhang, J., X. Liu, D. C. Scherer, L. van Kaer, X. Wang, and M. Xu. 1998. Resistance to DNA fragmentation and chromatin condensation in mice lacking which is different from that generated by caspase 3-like activity the DNA fragmentation factor 45. Proc. Natl. Acad. Sci. USA 95: 12480–12485. or GzmB. 17. Kelly, J. M., M. D. O’Connor, M. D. Hulett, K. Y. Thia, and M. J. Smyth. 1996. ϩ ϩ Cloning and expression of the recombinant mouse natural killer cell granzyme Unlike GzmB, GzmM is not expressed in CD4 or CD8 T Met-ase-1. Immunogenetics 44: 340–350. cells either constitutively or after activation (19). However, GzmM 18. Smyth, M. J., T. Wiltrout, J. A. Trapani, K. S. Ottaway, R. Sowder, is constitutively highly expressed in NK cells as is perforin. A L. E. Henderson, C. M. Kam, J. C. Powers, H. A. Young, and T. J. Sayers. 1992. Purification and cloning of a novel serine protease, RNK-Met-1, from the gran- recent report showed a human NK cell line KHYG-1 appears to ules of a rat natural killer cell leukemia. J. Biol. Chem. 267: 24418–24425. have greater killing activity that is consistent with higher GzmM 19. Sayers, T. J., A. D. Brooks, J. M. Ward, T. Hoshino, W. E. Bere, G. W. Wiegand, expression (20), although GzmA and B are undetectable. It sug- J. M. Kelly, and M. J. Smyth. 2001. The restricted expression of granzyme M in human lymphocytes. J. Immunol. 166: 765–771. gests that GzmM may play a crucial role in NK cell-mediated 20. Suck, G., D. R. Branch, M. J. Smyth, R. G. Miller, J. Vergidis, S. Fahim, and cytolysis. In fact, reports by us and Kelly et al. (21) demonstrated A. Keating. 2005. KHYG-1, a model for the study of enhanced natural killer cell cytotoxicity. Exp. Hematol. 33: 1160–1171. that GzmM-induced rapid cell death is in concert with the kinetics 21. Kelly, J. M., N. J. Waterhouse, E. Cretney, K. A. Browne, S. Ellis, J. A. Trapani, of cytolysis by NK cells. But we found that GzmM-induced cell and M. J. Smyth. 2004. Granzyme M mediates a novel form of perforin-depen- death is dependent on caspase activation and DNA fragmentation dent cell death. J. Biol. Chem. 279: 22236–22242. 22. Beresford, P. J., D. Zhang, D. Y. Oh, Z. Fan, E. L. Greer, M. L. Russo, M. Jaju, reminiscent of GzmB. NK cells are the effectors of the innate im- and J. Lieberman. 2001. Granzyme A activates an endoplasmic reticulum-asso- munity, acting as the first line of defense against viral infection and ciated caspase-independent nuclease to induce single-stranded DNA nicks. tumors (49). A recent report from NK cell depletion experiments J. Biol. Chem. 276: 43285–43293. 23. Johnson, H., L. Scorrano, S. J. Korsmeyer, and T. J. Ley. 2003. Cell death in- showed NK cells are critical for tumor immunity (50). They exert duced by granzyme C. Blood 101: 3093–3101. their effector function without prior sensitization. Manipulating 24. Shi, L., C. M. Kam, J. C. Powers, R. Aebersold, and A. H. Greenberg. 1992. Purification of three cytotoxic lymphocyte granule serine proteases that induce NK cells to eradicate tumors may be a powerful approach for can- apoptosis through distinct substrate and target cell interactions. J. Exp. Med. 176: cer immunotherapy. Therefore, to fully understand the biology and 1521–1529. 1178 GzmM-INDUCED DNA FRAGMENTATION

25. Froelich, C. J., K. Orth, J. Turbov, P. Seth, R. Gottlieb, B. Babior, G. M. Shah, ribose) polymerase to the 89-kDa apoptotic fragment and less abundant 64-kDa R. C. Bleackley, V. M. Dixit, and W. Hanna. 1996. New paradigm for lympho- fragment. Biochem. Biophys. Res. Commun. 227: 658–665. cyte granule-mediated cytotoxicity: target cells bind and internalize granzyme B, 38. Podack, E. R., J. D. Young, and Z. A. Cohn. 1985. Isolation and biochemical and but an endosomolytic agent is necessary for cytosolic delivery and subsequent functional characterization of perforin 1 from cytolytic T-cell granules. Proc. apoptosis. J. Biol. Chem. 271: 29073–29079. Natl. Acad. Sci. USA 82: 8629–8633. 26. Wang, G. Q., E. Wieckowski, L. A. Goldstein, B. R. Gastman, A. Rabinovitz, 39. Tschopp, J., D. Masson, and K. K. Stanley. 1986. Structural/functional similarity A. Gambotto, S. Li, B. Fang, X. M. Yin, and H. Rabinowich. 2001. Resistance between proteins involved in complement- and cytotoxic T-lymphocyte-mediated to granzyme B-mediated cytochrome c release in Bak-deficient cells. J. Exp. Med. cytolysis. Nature 322: 831–834. 194: 1325–1337. 40. Kawasaki, Y., T. Saito, Y. Shirota-Someya, Y. Ikegami, H. Komano, M. H. Lee, 27. Pinkoski, M. J., M. Hobman, J. A. Heibein, K. Tomaselli, F. Li, P. Seth, C. J. Froelich, N. Shinohara, and H. Takayama. 2000. Cell death-associated trans- C. J. Froelich, and R. C. Bleackley. 1998. Entry and trafficking of granzyme B in location of plasma membrane components induced by CTL. J. Immunol. 164: target cells during granzyme B-perforin-mediated apoptosis. Blood 92: 4641–4648. 1044–1054. 41. Metkar, S. S., B. Wang, M. Aguilar-Santelises, S. M. Raja, L. Uhlin-Hansen, 28. Danial, N. N., and S. J. Korsmeyer. 2004. Cell death: critical control points. Cell E. Podack, J. A. Trapani, and C. J. Froelich. 2002. Cytotoxic cell granule-me- 116: 205–219. diated apoptosis: perforin delivers granzyme B-serglycin complexes into target 29. Adrain, C., B. M. Murphy, and S. J. Martin. 2005. Molecular ordering of the cells without plasma membrane pore formation. Immunity 16: 417–428. caspase activation cascade initiated by the cytotoxic T lymphocyte/natural killer 42. Shi, L., D. Keefe, E. Durand, H. Feng, D. Zhang, and J. Lieberman. 2005. Gran- (CTL/NK) protease granzyme B. J. Biol. Chem. 280: 4663–4673. zyme B binds to target cells mostly by charge and must be added at the same time 30. Darmon, A. J., D. W. Nicholson, and R. C. Bleackley. 1995. Activation of the as perforin to trigger apoptosis. J. Immunol. 174: 5456–5461. apoptotic protease CPP32 by cytotoxic T-cell-derived granzyme B. Nature 377: 43. Zhao, J., L. H. Zhang, L. T. Jia, L. Zhang, Y. M. Xu, Z. Wang, C. J. Yu, 446–448. W. D. Peng, W. H. Wen, C. J. Wang, et al. 2004. Secreted antibody/granzyme B 31. Liu, X., P. Li, P. Widlak, H. Zou, X. Luo, W. T. Garrard, and X. Wang. 1998. The fusion protein stimulates selective killing of HER2-overexpressing tumor cells. 40-kDa subunit of DNA fragmentation factor induces DNA fragmentation and J. Biol. Chem. 279: 21343–21348. chromatin condensation during apoptosis. Proc. Natl. Acad. Sci. USA 95: 44. Talanian, R. V., X. Yang, J. Turbov, P. Seth, T. Ghayur, C. A. Casiano, K. Orth, 8461–8466. and C. J. Froelich. 1997. Granule-mediated killing: pathways for granzyme B- 32. Liu, X., H. Zou, C. Slaughter, and X. Wang. 1997. DFF, a heterodimeric protein initiated apoptosis. J. Exp. Med. 186: 1323–1331. Downloaded from that functions downstream of caspase-3 to trigger DNA fragmentation during 45. Yang, X., H. R. Stennicke, B. Wang, D. R. Green, R. U. Janicke, A. Srinivasan, apoptosis. Cell 89: 175–184. P. Seth, G. S. Salvesen, and C. J. Froelich. 1998. Granzyme B mimics apical 33. Enari, M., H. Sakahira, H. Yokoyama, K. Okawa, A. Iwamatsu, and S. Nagata. caspases: description of a unified pathway for trans-activation of executioner 1998. A caspase-activated DNase that degrades DNA during apoptosis, and its caspase-3 and -7. J. Biol. Chem. 273: 34278–34283. inhibitor ICAD. Nature 391: 43–50. 46. Shi, L., R. P. Kraut, R. Aebersold, and A. H. Greenberg. 1992. A natural killer 34. Ame, J. C., E. L. Jacobson, and M. K. Jacobson. 1999. Molecular heterogeneity cell granule protein that induces DNA fragmentation and apoptosis. J. Exp. Med. and regulation of poly(ADP-ribose) glycohydrolase. Mol. Cell. Biochem. 193: 175: 553–566.

75–81. 47. Duriez, P. J., and G. M. Shah. 1997. Cleavage of poly(ADP-ribose) polymerase: http://www.jimmunol.org/ 35. D’Amours, D., S. Desnoyers, I. D’Silva, and G. G. Poirier. 1999. Poly(ADP- a sensitive parameter to study cell death. Biochem. Cell Biol. 75: 337–349. ribosyl)ation reactions in the regulation of nuclear functions. Biochem. J. 342: 48. Kaufmann, S. H., S. Desnoyers, Y. Ottaviano, N. E. Davidson, and G. G. Poirier. 249–268. 1993. Specific proteolytic cleavage of poly(ADP-ribose) polymerase: an early 36. Andrade, F., S. Roy, D. Nicholson, N. Thornberry, A. Rosen, and marker of chemotherapy-induced apoptosis. Cancer Res. 53: 3976–3985. L. Casciola-Rosen. 1998. Granzyme B directly and efficiently cleaves several 49. Fan, Z., P. Yu, Y. Wang, M. L. Fu, W. Liu, Y. Sun, and Y. X. Fu. 2005. NK cell downstream caspase substrates: implications for CTL-induced apoptosis. Immu- activation by LIGHT triggers tumor specific CD8ϩ T cell immunity to reject nity 8: 451–460. established tumors. Blood 107: 1342–1351. 37. Froelich, C. J., W. L. Hanna, G. G. Poirier, P. J. Duriez, D. D’Amours, 50. Kelly, J. M., P. K. Darcy, J. L. Markby, D. I. Godfrey, K. Takeda, H. Yagita, and G. S. Salvesen, E. S. Alnemri, W. C. Earnshaw, and G. M. Shah. 1996. Granzyme M. J. Smyth. 2002. Induction of tumor-specific T cell memory by NK cell-me- B/perforin-mediated apoptosis of Jurkat cells results in cleavage of poly(ADP- diated tumor rejection. Nat. Immunol. 3: 83–90. by guest on September 24, 2021