Oncogene (2001) 20, 358 ± 366 ã 2001 Nature Publishing Group All rights reserved 0950 ± 9232/01 $15.00 www.nature.com/onc Inactivation of farnesyltransferase and geranylgeranyltransferase I by caspase-3: Cleavage of the common a subunit during apoptosis

Ki-Woo Kim1,5, Hyun-Ho Chung2,5, Chul-Woong Chung1, In-Ki Kim1, Masayuki Miura3, Suyue Wang4, Hong Zhu4, Kyung-Duk Moon2, Geun-Bae Rha2, Jy-Hyun Park2, Dong-Gyu Jo1, Ha-Na Woo1, Yu-Hyun Song1, Byung Ju Kim1, Junying Yuan4 and Yong-Keun Jung*,1

1Department of Life Science, Kwangju Institute of Science and Technology, Puk-gu, Kwangju 500-712, Korea; 2Biotech Research Institute, LG Chem./Research Park, Taejeon, Korea; 3Department of Neuroanatomy, Osaka University Medical School, Osaka 565, Japan; 4Department of Cell Biology, Harvard Medical School, 240 Longwood Avenue, Boston, Massachusetts, MA 02115, USA

Caspase plays an important role in apoptosis. We report (O'Reilly and Strasser, 1999; Tatton and Olanow, 1999; here that farnesyltransferase/geranylgeranyltransferase Zhivotovsky et al., 1999). Caspases, a highly conserved (FTase/GGTase)-a, a common subunit of FTase (a/ family of cysteine proteases, play a critical role in bFTase) and GGTase I (a/bGGTase), was cleaved by mammalian cell death (Yuan et al., 1993). Thus far, 14 caspase-3 during apoptosis. FTase/GGTase-a (49 kDa) members of the caspase family have been identi®ed, was cleaved to 35 kDa (p35) in the Rat-2/H-ras, W4 and each with distinct substrate recognition properties Rat-1 cells treated with FTase inhibitor (LB42708), anti- (Alnemri et al., 1996; Humke et al., 1998; Ahmad et Fas antibody and etoposide, respectively. This cleavage al., 1998). The critical role of caspases in apoptosis has was inhibited by caspase-inhibitors (YVAD-cmk, DEVD- been shown in cell culture and animal models with cho). Serial N-terminal deletions and site-directed inhibitors or caspase mutation. Caspase inhibitors have mutagenesis showed that Asp59 of FTase/GGTase-a been shown to suppress apoptosis induced by various was cleaved by caspase-3. The common FTase/GGTase- signals including Fas, tumor necrosis factor (TNF)-a, a subunit, but not the b subunits, of the FTase or growth factor withdrawal, etoposide, or disruption of GGTase I protein complexes puri®ed from baculovirus- the extracellular matrix (Thornberry and Lazebnik, infected SF-9 cells was cleaved to be inactivated by 1998; Nunez et al., 1998). In addition, caspase-3, -9, puri®ed caspase-3. In contrast, FTase mutant protein and -12 (7/7) mutant mice show decreased apoptosis, complex [(D59A)a/bFTase] was resistant to caspase-3. resulting in altered nervous system development (Kuida Expression of either the cleavage product (60-379) or et al., 1996, 1998; Hakem et al., 1998; Nakagawa et al., anti-sense of FTase/GGTase-a induced cell death in Rat- 2000). 2/H-ras cells. Furthermore, expression of (D59A)FTase/ Caspases are synthesized as inactive proenzymes and GGTase-a mutant signi®cantly desensitized cells to activated to speci®cally cleave both relevant cellular etoposide-induced death. Taken together, we suggest substrates and zymogens to progress apoptosis (Mar- that cleavage of prenyltransferase by caspase contributes golin et al., 1997; Thornberry and Lazebnik, 1998; to the progression of apoptosis. Oncogene (2001) 20, Nicholson, 1999). Thus, a key to understand the 358±366. molecular basis of apoptosis likely lies in the identi®cation and characterization of critical caspase Keywords: apoptosis; caspase; farnesyltransferase; substrates. Identi®ed caspase substrates include the geranylgeranyltransferase ICAD (Sakahira et al., 1998), amyloid-b precursor protein (Gervais et al., 1999), the retinoblastoma gene product (Janicke et al., 1996) and gelsolin (Kothakota Introduction et al., 1997). While it is known that in some cases proteolysis activates caspase substrates and in others it Apoptosis is a highly regulated cellular suicide inactivates or destroys them, the crucial substrate mechanism that controls development and homeostasis proteins that coordinate cell death are as yet unchar- in multicellular organisms. Inappropriate onset or acterized. defects in sensitivity to an apoptotic stimulus can give Protein prenylation, which is catalyzed by FTase and rise to a number of clinical conditions, including GGTase, modi®es and regulates the activity of many neurodegenerative disorders, autoimmunity, and cancer proteins important for cell proliferation and survival, including Ras, Rho, and brain type I inositol 1, 4, 5- triphosphate 5-phosphatase (Zhang and Casey, 1996; De Smedt et al., 1996). Inhibition of prenylation with *Correspondence: Y-K Jung inhibitors including FTI, GGTI-298, or lovastatin has 5These two authors contributed equally to this work Received 25 September 2000; revised 9 November 2000; accepted 9 recently shown to cause growth arrest and promote November 2000 apoptosis (Bernhard et al., 1996; Jansen et al., 1997; Cleavage of protein prenyltransferase by caspase-3 K-W Kim et al 359 Suzuki et al., 1998; Stark et al., 1998; Du et al., 1999; protein into a 35 kDa fragment (DFT/GGT-a) was Sun et al., 1999; Ghosh et al., 1999; Du and ®rst detected in cells showing 88% viability. The Prendergast, 1999; Feldkamp et al., 1999), suggesting cleavage pattern of FTase/GGTase-a was further that prenylation is an important step in the survival examined in Rat-1 and Rat-2 ®broblasts transformed signaling cascade. Here we show that FTase/GGTase- with oncogenic H-ras (Rat-2/H-ras) (Figure 1b,c). a, a common subunit of heterodimeric FTase (a/bFTase) Incubation of Rat-1 cells with etoposide readily and GGTase I (a/bGGTase), is cleaved by caspase-3 in induced cleavage of FTase/GGTase-a, which was vitro and in vivo to be inactivated during apoptosis, already detectable in cells with 94% viability (Figure which may help us to understand how the prenylation 1b). Exposure of Rat-2/H-ras cells to FTase inhibitor signaling responds to apoptosis. (FTI: LB42708, IC50=2.5 nM) produced a 35 kDa fragment in apoptotic cells (50% viability at 10 mM FTI) (Figure 1c). These results indicate that FTase/ GGTase-a subunit is cleaved during Fas-, etoposide-, Results and FTI-mediated apoptosis. To investigate whether FTase/GGTase-a was cleaved FTase/GGTase-a subunit is cleaved by caspase during by caspase, W4 cells were incubated with Jo-2 antibody apoptosis in the presence of YVAD-cmk or DEVD-cho (Figure In order to evaluate FTase/GGTase-a as a potential 1a). Preincubation of W4 cells with caspase inhibitor caspase substrate, proteolytic cleavage of FTase/ suppressed Fas-mediated apoptosis (viability, 18 to GGTase-a was examined during apoptosis. W4 cells, 84 ± 87%) and the cleavage of FTase/GGTase-a in the a mouse lymphoma cell line expressing the Fas cells, indicating that FTase/GGTase-a is cleaved by receptor, were induced to undergo apoptosis by caspases in the apoptotic cells. Examination of treatment with anti-Fas antibody (Jo-2). After deter- proteolytic activation of caspase with Western blot mining cell viability, cell lysates were prepared to analysis showed that caspase-3 was activated in W4 examine the cleavage of FTase/GGTase-a by Western cells showing 88% viability (Figure 2) and pre- blot analysis (Figure 1). Viability of the W4 cells was incubation of cells with YVAD-cmk inhibited proteo- 88% at 1 h and decreased to 18% at 4 h after lytic processing of caspase-3. In contrast, addition of treatment with anti-Fas antibody (Figure 1a). The DEVD-cho, which suppressed the cleavage of FTase/ cleavage of full-length (49 kDa) FTase/GGTase-a GGTase-a (Figure 1a), did not inhibit the processing of

Figure 1 Cleavage of the a subunit of the prenyltransferase, FTase and GGTase I, by caspase during apoptosis. (a) FTase/ GGTase-a (FT/GGT-a) cleavage by caspase in W4 cells undergoing apoptosis. W4 cells were preincubated with or without 100 mM YVAD-cmk (YVAD) or DEVD-cho (DEVD) for 2 h and then treated with 300 ng/ml anti-mouse Fas antibody (Jo-2) for 0, 1, 2 and 4 h. Cell lysates were analysed by Western blot using anti-FTase/GGTase-a antibody. For each time point, the corresponding cell viability, as determined with trypan blue assay, is indicated at the top. DFT/GGT-a indicates the truncated 35 kDa fragment of FTase/GGTase-a.(b) Cleavage of FTase/GGTase-a in Rat-1 cells. Rat-1 cells were incubated for the times indicated at the top with etoposide (30 mM) and analysed for cell viability and cleavage of FTase/GGTase-a.(c) Rat-2/H-ras cells were treated with 5 or 10 mM FTase inhibitor (FTI: LB42708) for 8 h (lanes 2,3) and recombinant FTase/GGTase-a subunit was incubated with 10 nM caspase-3 (casp-3) (lane 5). The reaction products were examined for the cleavage of FTase/GGTase-a with Western blot analysis

Oncogene Cleavage of protein prenyltransferase by caspase-3 K-W Kim et al 360

Figure 2 Proteolytic activation of caspase-3 in W4 cells under- going apoptosis. W4 cells were treated with anti-Jo-2 antibody in the presence or absence of 100 mM caspase inhibitors (YVAD-cmk and DEVD-cho) and the cell lysates were examined for activation of caspase-3 by Western blot using anti-hamster caspase-3 and anti-tubulin antibodies. For each time point, the corresponding cell viability is indicated at the top. The size of the detected proteins is indicated (p32 and p18)

caspase-3, indicating that a YVAD-inhibitable protease may act upstream of caspase-3 to process caspase-3 into its active subunits.

Mapping of FTase/GGTase-a cleavage site: Caspase-3 cleaves at Asp59 of FTase/GGTase-a To con®rm that a 35 kDa fragment (DFT/GGT-a)isa cleavage product of FTase/GGTase-a, COS-7 cells were transfected with rat FTase/GGTase-a tagged with hemagglutinin (HA) and exposed to etoposide or staurosporine (Figure 3). As shown in Figure 3b, anti-rat FTase/GGTase-a antibody only recognized the transfected rat FTase/GGTase-a and detected a Figure 3 Determination of the FTase/GGTase-a cleavage site in 35 kDa fragment in the apoptotic COS-7 cells (Figure COS-7 cells. (a) A schematic diagram of full size (FT/GGT-a), N- 3b, ®rst panel), indicating that a 35 kDa fragment is terminal deletions (DFT/GGT-a), and mutant (D59A)FT/GGT-a the cleavage product of rat FTase/GGTase-a. Whereas of FTase/GGTase-a containing HA at the N-terminus. Asp59 of anti-HA antibody recognized only full-length FTase/ FTase/GGTase-a was mutated to Ala in (D59A)FT/GGT-a. Numbers indicate Asp (D) positions in the primary sequence of GGTase-a which was eciently cleaved (Figure 3b, FTase/GGTase-a and the arrow indicates the putative cleavage second panel), no cleavage of FTase b subunit was site (Asp59) of FTase/GGTase-a (54-GFLSLDS-60) by a caspase. observed (Figure 3b, third panel). (b) COS-7 cells were transfected with HA-pcDNA3 expressing Caspase cleavage requires an Asp residue at posi- either full size or serial N-terminal deletions of FTase/GGTase-a. tion+1 relative to the cleavage site in the recognition After 24 h, cells were treated for 16 h with staurosporine (stauro. 1 mM) or etoposide (etopo. 60 mM) and then analysed for the motif of caspase (Margolin et al., 1997). The anti- cleavage of the transfected rat FTase/GGTase-a with Western FTase/GGTase-a antibody recognized the C-terminus blot: anti-rat FTase/GGTase-a antibody (®rst panel), anti-HA of FTase/GGTase-a, indicating that cleavage occurred antibody (second panel), anti-rat FTase-b antibody (third panel), near the N-terminus. There are three Asp residues near and anti-tubulin antibody (last panel). Asterisk (*) indicates the cleavage product (p35) of FTase/GGTase-a.(c) COS-7 cells the N-terminus of FTase/GGTase-a for the potential transfected with FTase/GGTase-a or (D59A)FTase/GGTase-a cleavage site. To determine the caspase cleavage site, mutant were exposed to staurosporine (1 mM) for 12 h and three N-terminal deletions starting at each of these Asp analysed for the cleavage with Western blot residues in FTase/GGTase-a were constructed (Figure 3a). COS-7 cells were then transfected with these N- terminal deletions (DFT/GGT-a) and subsequently We then replaced Asp59 with Ala using site-directed exposed to staurosporine or etoposide. Western blot mutagenesis, and the resulting (D59A)FTase/GGTase-a analysis with anti-rat FTase/GGTase-a antibody mutant were expressed in COS-7 cells. Subsequent showed that the 35 kDa fragment migrated between exposure to staurosporine failed to induce cleavage of the Asp59 and Asp82 deletion constructs (Figure 3b). the (D59A)FTase/GGTase-a mutant (Figure 3c). There- Because HA tag (16 amino acids) was attached at the fore, Asp59 in FTase/GGTase-a may be the caspase N-terminus of the FTase/GGTase-a deletion, the HA- cleavage site, though we could not observe apparent tagged proteins were expected to migrate slower than cleavage of D(1-53)FT/GGT-a. Incubation of 35S- the corresponding cleavage products. labeled FTase/GGTase-a with bacteria extracts con-

Oncogene Cleavage of protein prenyltransferase by caspase-3 K-W Kim et al 361 taining activities of caspase-1, -3, -8, or -11 showed using monoclonal antibodies to FTase/GGTase-a (data that caspase-3 substantially cleaved the 49 kDa FTase/ not shown). When enzyme activities of FTase and GGTase-a into 35 kDa fragment (Figure 4a). However, GGTase I were measured from the reaction mixtures, (D59A)FTase/GGTase-a mutant was found to be cleavage of FTase/GGTase-a subunit by caspase-3 resistant to the cleavage by the puri®ed caspase-3 abolished FTase and GGTase I activities in vitro (Figure 4b), indicating that caspase-3 cleaves FTase/ (Figure 5b). FTase mutant [(D59A)a/bFTase] was then GGTase-a at Asp59 in vitro. co-puri®ed by expressing (D59A)FTase/GGTase-a and FTase-b subunits together (Figure 6a). Incubation of the FTase mutant [(D A)a/b ] with caspase-3 FTase and GGTase I enzymes are inactivated by 59 FTase neither induced any cleavage of the (D A)FTase/ caspase-3 59 GGTase-a mutant Figure 6b) and FTase-b subunit In order to examine whether the cleavage a€ects FTase (Figure 6c) nor reduced FTase activity (Figure 6d). and GGTase I activities, FTase/GGTase-a subunit was co-expressed with FTase-b subunit (bFTase) or GGTase I-b subunit (bGGTase) in baculovirus infected SF-9 cells and puri®ed. The heterodimeric FTase (a/bFTase)and GGTase I (a/bGGTase) were incubated with caspase-3 and visualized with Coomassie blue staining after SDS ± PAGE (Figure 5a). Caspase-3 cleaved the common 49 kDa FTase/GGTase-a subunit into a 35 kDa fragment (lane 2 and 5), which was blocked by the caspase inhibitor DEVD-cho (lane 3 and 6). In contrast, the b subunits of FTase and GGTase I were not cleaved by caspase-3. The cleavage of FTase/ GGTase-a was con®rmed by Western blot analysis

Figure 5 Inactivation of FTase and GGTase I by caspase-3 in vitro. FTase (a/bFTase) and GGTase I (a/bGGTase) enzyme complexes were puri®ed from baculovirus-infected SF-9 cells as described in Materials and methods. (a) Caspase-3 cleaves only the shared a subunit of both FTase and GGTase I enzymes in vitro. After incubation with puri®ed caspase-3, reaction mixtures were subjected to SDS ± PAGE and visualized by Coomassie blue staining. Lane 1 and 4, FTase and GGTase I, respectively; lane 2 Figure 4 Cleavage at Asp59 of FTase/GGTase-a by caspase-3 in and 4, FTase and GGTase I incubated with 10 nM caspase-3; lane vitro.(a) FTase/GGTase-a was translated in vitro in the presence 3 and 6, FTase and GGTase I incubated with 10 nM caspase-3 in of 35S-methionine and incubated with 4 ml (20 ± 25 mg) of E. coli the presence of 500 nM DEVD-cho, respectively. (b) Caspase-3 extracts containing recombinant caspases (caspase-1, -3, -8 or inactivates the prenyltransferase. FTase and GGTase I were -11). The reaction mixtures were separated by SDS ± PAGE and untreated or incubated with caspase-3 and each reaction mixture exposed to X-ray ®lm. The arrows indicate 35S-labeled cleavage was assayed for prenyltransferase activity. Lane 1 and 3, FTase product (p53 and N-terminal fragment). (b) Wild-type (FT/GGT- and GGTase I; lane 2 and 5, FTase and prenyltransferase activity. a and mutant [(D59A)FT/GGT-a] of FTase/GGTase-a were Lane 1 and 3, FTase and GGTase I; lane 2 and 4, FTase and labeled with 35S-methionine and incubated with puri®ed cas- GGTase I treated with 10 nM caspase-3. Bars represent means pase-3 or -7 (10 nM) +s.d. from three independent experiments

Oncogene Cleavage of protein prenyltransferase by caspase-3 K-W Kim et al 362 The notion that FTase/GGTase-a was cleaved in 11%), expression of FTase/GGTase-a (anti-sense) vitro and in vivo by caspase led us to investigate increased cell death twofold (22%). Interestingly, whether protein prenylation was reduced during expression of D(1 ± 59)FTase/GGTase-a signi®cantly apoptosis. We examined proteolytic processing of Ras increased cell death rate to 18%. at the C-terminal CAAX motif, which was induced by We then examined the e€ects of expression of the farnesylation, in Rat-2/H-ras cells upon treatment with cleavage resistant mutant, (D59A)FTase/GGTase-a,on etoposide (Figure 7). In the presence of lovastatin, which blocks Ras processing by inhibiting a rate limiting step in isoprenoid biosynthesis, unprocessed Ras protein migrated more slowly than the processed Ras of control cells. Treatment with 50 mM etoposide reduced Ras processing, indicating that Ras prenyla- tion was reduced in cells undergoing apoptosis.

Expression of the cleavage product or cleavage resistant mutant of FTase/GGTase-a affects cell viability Figure 7 Inhibition of Ras processing in cells undergoing apoptosis by treatment with etoposide. Rat-2/H-ras cells were We have then addressed whether overexpression of the incubated with lovastatin or etoposide for 2 h in the presence of cleavage product of the FTase/GGTase-a exerted 35S-methionine. After additional incubation for 20 h, Ras was e€ects on cell death (Figure 8a). While Rat-2/H-ras immunoprecipitated from cell extracts with monoclonal antibody (Y13-259), resolved by SDS ± PAGE, and then detected by cells transfected with pcDNA3 (control), FTase/ autoradiography. P, processed; U, unprocessed; C, control; L, GGTase-a (sense), or HA-tagged FTase/GGTase-a lovastatin (15 mM); lane 1, etoposide (50 mM); lane 2, etoposide (sense) showed background levels of cell death (9 ± (100 mM)

Figure 6 Resistance of the heterodimeric FTase mutant [mFTase: (D59A)a/bFTase)] to cleavage and inactivation by caspase-3. FTase (a/b/FTase) and FTase mutant [(D59A)a/bFtase)] were puri®ed and visualized with Coomassie-blue staining (a). Each protein complex was incubated with 10 nM caspase-3 and the reaction products were probed with Western blot using anti-FTase/GGTase-a antibody (b) and anti-FTase-b antibody (c). (d) Enzyme activities of FTase and FTase mutant exposed to caspase-3 were measured from three independent experiments and the activity obtained from wild-type FTase was set to 100

Oncogene Cleavage of protein prenyltransferase by caspase-3 K-W Kim et al 363 apoptosis. Rat-1 cells were transiently transfected with plasmids expressing FTase/GGTase-a or (D59A)FTase/ GGTase-a mutant and then exposed to etoposide (Figure 8b). Determination of cell viability showed that Rat-1 cells expressing (D59A)FTase/GGTase-a mutant became resistant to cell death compared to FTase/GGTase-a (21 and 31% of cell death, respec- tively). These results suggest that the cleavage of FTase/GGTase-a may contribute to progress of apoptosis.

Discussion

Inhibition of the protein prenylation step is known to cause tumor regression in transgenic mice (Barrington et al., 1998; Norgaard et al., 1999) and to induce apoptosis in vitro (Sun et al., 1995; Lebowitz et al., 1997; Miquel et al., 1997; Bernhard et al., 1998; Du et al., 1999) leading to development of a new class of anti-cancer drugs. In our results, prenylation appears to be required for prosurvival signals: While inhibition of prenylation activity with FTI or expression of anti- sense FTase/GGTase-a induced cell death, expression of the cleavage resistant FTase mutant attenuated cell death evoked by etoposide in Rat-1 and Rat-2/H-ras cells. We also observed inhibitory e€ects of apoptosis on the processing of Ras, caused probably by reducing the farnesylation, in apoptotic cells and inactivation of protein prenylation enzymes by caspase in vitro, implicating that in vivo cleavage of FTase/GGTase-a leads to inactivation of prosurvival function of prenylating enzymes and enhances or accelerates the apoptotic process. Of many substrates for prenylation, farnesylation and geranylgeranylation of Ras and Rho subfamily, respectively have been shown to exert survival e€ects on various apoptotic signals. For example, tumor cells and transformed cells expressing the ras oncogene are highly resistant to apoptosis induced by ionizing radiation, E1A, c-myc, or disruption of epithelial cell- matrix (anoikis) (Lin et al., 1995; McKenna et al., 1996; Kau€mann-Zeh et al., 1997; Rosen et al., 2000), though Gulbins et al. (1996) reported that Ras is required for Fas-induced apoptosis. Inhibition of Ras Figure 8 E€ects of D(1 ± 59)FTase/GGTase-a and (D59A)FT/ prenylation may interfere with far down-stream GGT-a mutant expression on cell death. (a) Expression of the apoptotic activities such as Bcl-2 family (Kinoshita et cleavage product, D(1 ± 59)FT/GGT-a, and anti-sense FTase/ al., 1995; Scheid et al., 1999; Rosen et al., 2000). In GGTase-a induced apoptosis in Rat-2/H-ras cells. Rat-2/H-ras addition, increasing numbers of evidences for the cells were transfected with both pbactgal expressing b-galactosi- dase and either pcDNA3 (control), FT/GGT-a, HA-FT/GGT-a, potential roles of the geranylgeranylated protein such D(1 ± 59)FT/GGT-a, or FT/GGT-a (anti-sense). Cells were ®xed as R-Ras, RhoA, Rac1 and Cdc42Hc in apoptosis have 2.5 days later, stained for b-galactosidase activity, and viability been accumulated (Miquel et al., 1997; Stark et al., was determined based on cell morphology. Asterisk indicates the 1998; Ghosh et al., 1999; Du et al., 1999). statistical signi®cance with respect to FT/GGT-a construct (F) While both peptide inhibitors, YVAD-cmk and (p50.05). (b) Expression of the cleavage-resistant (D59A)FT/ GGT-a mutant reduced cell death induced by etoposide. Rat-1 DEVD-cho, inhibited cleavage of FTase/GGTase-a cells were transfected with both pbactgal and either pcDNA3 during Fas-induced apoptosis of W4 cells, Western (control), FT/GGT-a,or(D59A)FT/GGT-a mutant. One day blot analysis showed that YVAD-cmk, but not DEVD- later, cells were treated with etoposide (60 mM) for 16 h and the cho, inhibited the proteolytic activation of caspase-3. viability of b-galactosidase-positive cells was determined based on cell morphology. Percentage of cell death is shown with This result suggests that YVAD-inhibitable protease is mean+s.d. from four independent experiments and asterisk (*) upstream of DEVD-sensitive caspase-3 (Figure 2), indicates the statistical di€erence with respect to FT/GGT-a consistent with a delineation of caspase-cascade, construct (F)(P50.1)

Oncogene Cleavage of protein prenyltransferase by caspase-3 K-W Kim et al 364 although the exact pathway of caspase cascade CCCGAATTCTCGCCCACCTATGTC (D1 ± 59), or CCC- activation is not yet fully understood (Thornberry GAATTCGGCCCCAGTCCAGTG (D1 ± 82), respectively, and Lazebnik, 1998; Nunez et al., 1998). DEVD-cho and CGCTCTAGAGTCCACTTCTTCCAGCC. PCR pro- failed to suppress the proteolytic processing of caspase- ducts were subcloned into the EcoRI and XbaI sites of HA- 3 but might inhibit the cleavage of FTase/GGTase-a pcDNA3 and pcDNA3 (Invitrogen) for transfection and in vitro transcription/translation. Full-length human FTase/ probably by inhibiting caspase-3 activity. GGTase-a was also inserted between the NdeI and XhoI Expression of the cleavage product of FTase/ sites of pBacPAK8 (Clontech) for construction of a GGTase-a (p35) increased cell death probably by recombinant baculovirus (pBP8 ± FT/GGTa). FTase-b and interfering with the prenyl enzyme activities. The crystal GGTase-b were similarly inserted into pBacPAK8 to yield structure of heterodimeric mammalian FTase deter- pBP8-FTb and pBP8-GGTb, respectively. A FTase/GGTase- mined at 2.25AÊ resolution showed that, although the N- a (D59A) mutant was constructed by site-directed mutagenesis terminal proline rich domain of the a-subunit (residue 1 using Gene Editor mutagenesis system (Promega) and a synthetic oligonucleotide containing the mutation to 54) is disordered in the crystal, two 310 helices (55 ± 69 and 70 ± 73) and a short b-strand (89 ± 91) are (GGTTTGTGAGCCTGGCCTCGCCCTCCTTATTC). Cas- important for forming a stable heterodimeric complex pase cDNAs were ampli®ed by PCR using oligonucleotide primers CGCGGATCCTGGCACATTTCCAGGAC and (Park et al., 1997; Park and Beese, 1997), implying that CGCGGATCCTAAGGAAGTATTGGC for p30 domain cleavage of the common a subunit at the residue Asp59 of mouse caspase-1, CGCGGATCCGGAGAACACTGAA- results in disruption of heterodimerization and conse- AACTC and CTCGGATCCTACCATCTTCTCACTTGG quently inactivation of the enzymes. In addition, we for full-length of human caspase-3, and CGGGATCCTAGT- have examined interaction between FTase and Ras and GAATCACAGACTTTG and CCGCAAGCTTATCAGAA- mapped the region of the FTase/GGTase-a subunit that GGGAGACAAG for p30 domain of human caspase-8. The is responsible for binding to Ras using peptide PCR products were subcloned into the BamHI site (caspase-1 competition experiments (manuscript in preparation). and -3) or the BamHI and HindIII sites (caspase-8) of pET- Two peptides, including residues 25 ± 46 and 49 ± 70 of 15b (Novagen). The EcoRI fragment of caspase-11 cDNA FTase/GGTase-a, inhibited farnesylation of Ras by derived from pBSNO12 was inserted into pTrcHis (Introgen). All PCR products were con®rmed by DNA sequencing. FTase, suggesting that the N-terminal region containing these two peptide sequences may be involved in the binding to Ras. Although we cannot rule out the Cell culture, DNA transfection, and apoptosis assay possibility that cleavage products of FTase/GGTase-a Sf-9 cells were obtained from ATCC (American Type Culture may interact with other unidenti®ed prenylation- Collection) and maintained in Grace's medium (GIBCO), independent signaling pathways, our results imply that supplemented with 3.3 mg/ml lactoalbumin hydrolysate the N-terminus of FTase/GGTase-a is important for (Difco), 3.3 mg/ml yeasttolate (Difco), 10% (v/v) fetal calf catalytic activity of the enzyme complexes and thus, serum (FCS) (Hyclone Laboratories ), 50 mg/ml Gentamycin, that cleavage of FTase/GGTase-a during apoptosis and 0.1% Pluronic F-68 (GIBCO) in 125 ml Spinner ¯asks facilitates apoptosis by blocking a prenylation-depen- (Techne, Princeton, NJ, USA). W4 cells were grown in RPMI-1640 medium (Life Technologies, Inc.) supplemented dent survival signal. An analysis of prenylation-down- with 10% FCS. Rat-1, Rat-2/H-ras, and Cos-7 cells were stream events will help to further clarify the biological grown in Dulbecco's Modi®ed Eagle Medium (DMEM) (Life signi®cance of these observations in apoptosis. Technologies, Inc.) with 10% FCS. Cells were subcultured at a density of 26105 per well in 6 well dishes 1 day before transfection. For each well, 1 mg of DNA and 8 mgof lipofectamine reagent (Life Technologies, Inc.) were used following a protocol from Gibco. To induce apoptosis, cells Materials and methods were treated with 300 ng/ml anti-Fas antibody (Jo-2), etopo- side (60 m ), staurosporine (1 mM), and FTI (LB42708), and Reagents M cell viability was determined with 0.4% trypan blue staining Acetyl-Tyr-Val-Ala-Asp-chloromethylketone (YVAD-cmk) or FACS analysis. and acetyl-Asp-Glu-Val-Asp-aldehyde (DEVD-cho) were obtained from Bachem (Torrance, CA, USA). Etoposide, Production of recombinant virus staurosporine, and all others were purchased from Sigma (St. Louis, MO, USA). Anti-mouse Fas antibody (Jo-2) and To generate recombinant baculovirus. Sf-9 cells (26106) were FTase inhibitor (LB42708) were obtained from Pharmingen transfected with 0.5 mg/ml of BaculoGold wild-type viral (San Diego, CA, USA) and LG Chem. (Taejeon, Korea), DNA (Pharmingen) and each 2 mg of pBP8-FT/GGTa or respectively. pBP8-GGTb. The virus from each transfection was harvested after 4 days and screened using a plaque assay as described (Summers and Smith, 1998). Recombinant viruses obtained Plasmid construction from this screen were subjected to two further rounds of cDNAs for the human FTase/GGTase-a, FTase-b and plaque puri®cation. GGTase I-b subunits were cloned by reverse transcription- polymerization chain reaction (PCR) from human Colo205 Expression and purification of prenyl protein transferases in Sf-9 cells. Full-length and N-terminal deletions of rat FTase/ cells GGTase-a were generated by PCR with the primers either one of CCCGAATTCATGGCGGCCACTGAG (full- Puri®ed recombinant viruses were used to infect Sf-9 cells at length), CCCGAATTCGGGTTTCTGAGCCTG (D1 ± 53), a multiplicity of infection of 2. Cells were harvested 48 h

Oncogene Cleavage of protein prenyltransferase by caspase-3 K-W Kim et al 365 post-infection by centrifugation at 8006g for 15 min, washed containing 0.5% Nonidet P-40, 20 mM HEPES (pH 7.4), once with phosphate-bu€ered saline (PBS), and the resulting 100 mM NaCl, and 20 mM DTT for 1 h at 378C and cleavage cell pellet was ¯ash-frozen in a dry ice/ethanol bath. Cell products were separated by SDS ± PAGE and detected by extracts were prepared by thawing the cell suspension in 5 autoradiography. volumes of 20 mM Tris-HCl (pH 8.0), 1 mM EGTA, 1 mM DTT and by incubating the suspension on ice for 1 h, FTase and GGTase I assay followed by homogenization in a Dounce homogenizer. The resulting extract was centrifuged for 1 h at 30 0006g, and Prenyltransferase activity was determined by quantitating the the supernatant containing each prenyl protein transferase amount of [3H]prenyl diphosphate, farnesyl diphosphate, or was subjected to column chromatography as described geranylgeranyl diphosphate incorporated into Ras proteins previously with minor modi®cations (Reiss et al., 1990). (Reiss et al., 1992). The standard reaction mixture contained Both FTase and GGTase I were puri®ed to essential the following components in a ®nal volume of 50 ml; 50 mM homogeneity. Tris-HCl (pH 7.5), 5 mM MgCl2,5mM ZnCl2,2mM DTT, 4 mM Ras-CVLS (for FTase) or RAS-CVIL (for GGTase I) proteins, 2 m [3H]prenyl diphosphate (typically at Preparation of cell lysates, antibodies, and Western blot anlaysis M 2000 c.p.m./pmol), and the reaction mixtures containing Cell lysates were prepared and analysed by Western blot as 0.2 mg of FTase or GGTase I were incubated for 15 min at described previously (Jung et al., 1996) using anti-mouse (Y- 378C with or without caspase-3. 53) or anti-human (C-19) FTase/GGTase-a and anti-FTase-b antibodies (SC-137) (Santa Crutz). Human and hamster CPP32 monoclonal antibodies were from Transduction Laboratory (Lexington, KY, USA) and Dr J Goldstein Abbreviations (Cornell University), respectively. Anti-a-tubulin and anti-HA FTase, farnesyltransferase; FTI, farnesyltransferase inhibi- antibodies were purchased from Sigma and Boehringer tor; GGTase I, geranylgeranyltransferase I; YVAD-cmk, Mannheim, respectively. acetyl-Tyr-Val-Ala-Asp-chloromethylketone; DEVD-cho, acetyl-Asp-Glu-Val-Asp-aldehyde; DMEM, Dulbecco's Modi®ed Eagle Medium; FCS, fetal calf serum; PCR, In vitro caspase cleavage reaction polymerase chain reaction; SDS ± PAGE, sodium dodecyl For preparation of caspase extracts, bacterial plasmids sulfate-polyacrylamide gel electrophoresis; PBS, phosphate expressing caspase were transformed into E. coli BL21(DE3). bu€ered saline; HA, hemagglutinin. Exponentially growing cells were induced with 0.2 mM isopropyl-1-thio-a-D-galactopyranoside (IPTG) for 2 h, har- vested, and lyzed by sonication in a bu€er containing 0.05% Nonidet P-40, 20 mM HEPES (pH 7.4) and 100 mM NaCl. The lysates were cleared by centrifugation and the protein Acknowledgments concentration was determined with the Bio-Rad protein We thank S Nagata and J Goldstein for W4 cells and assay. When necessary, caspase-3 and -7 were puri®ed on hamster caspase-3 antibody, respectively. We thank N Ni-agarose from bacterial extracts. Proteins from the Spoerel for critical reading of this manuscript. This work plasmids were translated in vitro using the TNT system was supported by Brain Korea 21 project and in part by (Promega) in the presence of 35S-methionine (Amersham). In grants from the KOSEF (97-0401-07-01-5) and Molecular vitro cleavage assay reactions were performed in a bu€er Medicine Research.

References

Ahmad M, Srinivasula SM, Hedge R, Mukattash R, Feldkamp MM, Lau N and Guha A. (1999). Oncogene, 18, Fernandes-Alnemri T and Alnemri ES. (1998) Cancer 7514 ± 7526. Res., 58, 5201 ± 5205. Gervais FG, Xu D, Robertson GS, Vailancourt JP, Zhu Y, Alnemri ES, Livingston DJ, Nicholson DW, Salvesen G, Huang J, LeBlanc A, Smith D, Rigby M, Shearman MS, Thornberry NA, Wong WW and Yuan J. (1996). Cell, 87, Clark EE, Zheng H, Van Der Poleg LH, Ry€olo SC, 171. Thornberry NA, Xanthodarkis S, Zamboni RJ, Roy S and Barrington RE, Subler MA, Rands E, Omer CA, Miller PJ, Nicholson DW. (1999). Cell, 97, 395 ± 406. Hundley JE, Koester SK, Troyer DA, Bearss DJ, Conner Ghosh PM, Ghosh-Choudhury N, Moyer ML, Mott GE, MW, Gibbs JB, Hamilton K, Koblan KS, Mosser SD, Thomas CA, Foster BA, Greenberg NM and Kreisberg JI. O'Neil TJ, Schaber MD, Senderak ET, Windle JJ, Oli€ A (1999). Oncogene, 18, 4120 ± 4130. and Kohl NE. (1998). Mol. Cell. Biol., 18, 85 ± 92. GulbinsE,CoggeshallKM,BrennerB,SchlottmanK, Bernhard EJ, Kao G, Cox AD, Sebti SM, Hamilton AD, Linderkamp O and Lang F. (1996). J. Biol. Chem., 271, Muschel RJ and McKenna WG. (1996). Cancer Res., 56, 26390 ± 26394. 1727 ± 1730. Hakem R, Hakem A, Duncan GS, Henderson JT, Woo M, Bernhard EJ, McKenna WG, Hamilton AD, Sebti SM, Qian Soengas MS, Elia A, de la Pompa JL, Kagi D, Khoo W, Y and Muschel RJ. (1998). Cancer Res., 58, 1754 ± 1761. Potter J, Yoshida R, Kaufman SA, Lowe SW, Penninger De Smedt F, Boom A, Pesesse X, Schi€mann SN and Erneux JM and Mak TW. (1998). Cell, 94, 339 ± 352. C. (1996). J. Biol. Chem., 271, 10419 ± 10424. Humke EW, Ni J and Dixit VM. (1998). J. Biol. Chem., 273, Du W, Lu A and Prendergast GC. (1999). Cancer Res., 59, 15702 ± 15707. 4208 ± 4212. Janicke RU, Walker PA, Lin XY and Porter AG. (1996). Du W and Prendergast GC. (1999). Cancer Res., 59, 5492 ± EMBO J., 15, 6969 ± 6978. 5496.

Oncogene Cleavage of protein prenyltransferase by caspase-3 K-W Kim et al 366 Jansen B, Schlagbauer-Wadl H, Eichler HG, Wol€ K, van O'Reilly LA and Strasser A. (1999). In¯amm. Res., 48, 5±21. Elsas A, Schrier PI and Pehamberger H. (1997). Cancer Park HW and Beese LS. (1997). Curr. Opin. Struct. Biol., 7, Res., 57, 1846 ± 1850. 873 ± 880. Jung Y-K, Miura M and Yuan J. (1996). J. Biol. Chem., 271, ParkHW,BoduluriSR,MoomawJF,CaseyPJandBeese 5112 ± 5117. LS. (1997). Science, 275, 1800 ± 1804. Kau€mann-Zeh A, Rodriguez-Viciana P, Ulrich E, Gilbert Reiss Y, Brown MS and Goldstein JL. (1992). J. Biol. Chem., C, Co€er P, Downward J and Evan G. (1997). Nature, 385, 267, 6403 ± 6408. 544 ± 548. Reiss Y, Goldstein L, Seabra MC, Casey PJ and Brown MS. Kinoshita T, Yokota T, Arai K and Miyajima A. (1995). (1990). Cell, 62, 81 ± 88. Oncogene, 10, 2207 ± 2212. Rosen K, Rak J, Leung T, Dean NM, Kebel RS and Filmus Kothakota S, Azuma T, Reinhard C, Klippel A, Tang J, Chu J. (2000). J. Cell. Biol., 149, 447 ± 456. K, McGarry TJ, Kirschner MW, Koths K, Kwiatkowski Sakahira H, Enari M and Nagata S. (1998). Nature, 391, 96 ± DJ and Williams LT. (1997). Science, 278, 294 ± 298. 99. KuidaK,ZhengTS,NaS,KuanC-Y,YangD,Karasuyama Scheid MP, Schubert KM and Duronio V. (1999). J. Biol. H, Rakic P and Flavell RA. (1996). Nature, 384, 368 ± 372. Chem., 274, 31108 ± 31113. Kuida K, Haydar TF, Kuan CY, Gu Y, Taya C, Stark Jr WW, Blaskovich MA, Johnson BA, Qian Y, Karasuyama H, Su MS, Rakic P and Flavell RA. (1998). Vasudevan A, Pitt B, Hamilton AD, Sebti SM and Davies Cell, 94, 325 ± 337. P. (1998). Am. J. Physiol., 275, 55 ± 63. Lebowitz PF, Sakamuro D and Prendergast GC. (1997). Summers MD and Smith GE. (1998). Texas Agric. Exp. St. Cancer Res., 57, 708 ± 713. Bull., 1555, 29 ± 31. Lin HJ, Eviner V, Prendergast GC and White E. (1995). Mol. SunJ,QianY,ChenZ,MarfurtJ,HamiltonADandSebti Cell. Biol., 15, 4536 ± 4544. SM. (1999). J. Biol. Chem., 274, 6930 ± 6934. Margolin N, Raybuck SA, Wilson KP, Chen W, Fox T, Gu Sun J, Qian Y, Hamilton AD and Sebti SM. (1995). Cancer Y and Livingston DJ. (1997). J. Biol. Chem., 272, 7223 ± Res., 55, 4243 ± 4247. 7228. Suzuki N, Urano J and Tamanol F. (1998). Proc. Natl. Acad. McKenna WG, Bernhard EJ, Markiewicz DA, Rudoltz MS, Sci. USA, 95, 15356 ± 15361. Maity A and Muschel RJ. (1996). Oncogene, 12, 237 ± 245. Tatton WG and Olanow CW. (1999). Biochem. Biophys Miquel K, Pradines A, Sun J, Qian Y, Hamilton AD, Sebti Acta., 1410, 195 ± 213. SM and Farve G. (1997). Cancer Res., 57, 1846 ± 1850. Thornberry NA and Lazebnik Y. (1998). Science, 281, NakagawaT,ZhuH,MorishimaN,LiE,XuJ,YanknerBA 1312 ± 1316. and Yuan J. (2000). Nature, 403, 98 ± 103. Yuan J, Shaham S, Ledoux S, Ellis HM and Horvitz HR. Nicholson DW. (1999). Cell Death Di€er., 6, 1028 ± 1042. (1993). Cell, 75, 641 ± 651. NorgaardP,LawB,JosephH,PageDL,ShyrY,MaysD, Zhang FL and Casey PJ. (1996). Annu. Rev. Biochem., 65, Pietenpol JA, Kohl NE, Oli€ A, Co€er JR, Poulson HS 241 ± 269. and Moses HL. (1999). Clin. Cancer Res., 5, 35 ± 42. Zhivotovsky B, Joseph B and Orrenius S. (1999). Exp. Cell. Nunez G, Benedict MA, Hu Y and Inohara N. (1998). Res., 248, 10 ± 17. Oncogene, 17, 3237 ± 3245.

Oncogene