Báo cáo khoa học: Apoptosis-inhibiting activities of BIR family proteins in Xenopus egg extracts ppt

Apoptosis-inhibiting activities of BIR family proteins in Xenopus egg extracts Yuichi Tsuchiya, Shin Murai and Shigeru Yamashita Department of Biochemistry, Toho University School of Medicine, Tokyo, Japan The survival of oocytes and eggs is crucial to the reproduction of multicellular organisms (reviewed in [1,2]). Whereas oocytes grow and survive in nurturing conditions, ovulated eggs lose environmental support and have to survive on their own until successful fertil- ization. Eggs that fail to be fertilized for a long time need to be removed because they may suffer from aging and degeneration of maternal materials, leading to abortive or abnormal development. Recent studies on various species suggest that eggs possess a maternal system of apoptotic execution [3–8]. This machinery is inhibited for a limited time by unknown mechanisms, and only successfully developed embryos can switch off this embryonic time bomb. However, abnormally developed embryos, aged unfertilized eggs, and parthenogenetically activated eggs fail to stop the timer and commit suicide by an intrinsic apoptotic mechan- ism [9,10]. In Xenopus laevis (referred to simply as Xenopus hereafter) embryos, a drastic developmental change occurs after the 12th cell division, called the midblastula transition (MBT) stage. Checkpoint sys- tems are overridden before MBT in this system because of the rapid embryonic cell cycle, and embryos challenged by environmental insults (DNA replication block, DNA damage, or transcriptional ⁄ translational inhibition) still continue to divide until MBT. When the somatic cell cycle starts after MBT, checkpoint sys- tems begin to work properly, but cells with an unre- coverable accumulation of defects activate the suicide program and die by apoptosis at this stage. Aged and Keywords apoptosis inhibitor; survivin; Xenopus egg extracts; XIAP; XLX Correspondence S. Yamashita, Department of Biochemistry, Toho University School of Medicine, 5-21-16 Omori-nishi, Ota-ku, 143-8540 Tokyo, Japan Fax: +81 3 5493 5412 Tel: +81 3 3762 4151 ext. 2356 E-mail: yamasita@med.toho-u.ac.jp Note Nucleotide sequence data are available in the DDBJ ⁄ EMBL ⁄ GenBank databases under the accession numbers AB197247, AB197249, AB197251, and AB197252. (Received 11 January 2005, revised 1 March 2005, accepted 7 March 2005) doi:10.1111/j.1742-4658.2005.04648.x In many animal species including Xenopus, ovulated eggs possess an intrin- sic apoptotic execution system. This program is inhibited for a limited time by some maternal apoptosis inhibitors, although their molecular properties remain uncharacterized. Baculovirus IAP repeat (BIR) family proteins con- tain evolutionarily conserved BIR domains and play important roles in apoptosis suppression, and are therefore good candidates as maternal apoptosis inhibitors. We identified four maternal BIR family proteins in Xenopus eggs and, using the biochemical advantages of egg extracts, exam- ined their physiological functions. These molecules included two survivin- related proteins, xEIAP ⁄ XLX, and a possible ortholog of XIAP named xXIAP. The addition of recombinant xXIAP greatly delayed apoptotic exe- cution, whereas the immunodepletion of endogenous xXIAP significantly accelerated the onset of apoptosis. In contrast, xEIAP ⁄ XLX was a poor apoptosis inhibitor, and neither of the survivin orthologs showed anti- apoptotic activity in our assay. Both xEIAP ⁄ XLX and xXIAP were degra- ded by activated caspases, and also by a novel proteolytic system that required the presence of C-terminal RING finger domain but was insensit- ive to proteasome inhibition. Our data suggest that the regulation of endogenous xXIAP concentration is important for the survival of Xenopus eggs. Abbreviations BIR, baculovirus IAP repeat; CSF, cytostatic factor; IAP, inhibitor of apoptosis protein; MAPK, mitogen-activated protein kinase; MBT, midblastula transition; MBP, maltose-binding protein; MG-132, carbobenzoxy- L-leucyl-L-leucyl-L-leucinal; Z-VAD-FMK, carbobenzoxy-L-valyl- L-alanyl-b-methyl-L-aspart-1-yl fluoromethane. FEBS Journal 272 (2005) 2237–2250 ª 2005 FEBS 2237 parthenogenetically activated eggs are also thought to be killed by the same maternal mechanism, albeit with different timing. Stack & Newport [3] hypothesized that a maternal apoptosis inhibitor functions until MBT, although the nature of this inhibitor is not known. A physiologically relevant, sudden initiation of apoptotic execution can be reproduced with an in vitro system derived from Xenopus egg extracts [11]. When interphase egg extracts are incubated in the presence of sperm nuclei and the protein synthesis inhibitor cycloheximide (to prevent cyclin B synthesis and arrest the cell cycle) for several hours at room temperature, spontaneous nuclear fragmentation begins to appear, followed by the disappearance of nuclear membranes and the scattering of fragmented chromatin [12,13]. This change in nuclear morphology is known to be due to the activation of a cytochrome c-induced apop- totic pathway, as this phenomenon is accompanied by the release of cytochrome c from mitochondria, oligo- nucleosomal degradation of genomic DNA, activation of apoptotic caspases, and proteolytic cleavage of caspase substrates. Although this process has been reported to be effectively inhibited by the addition of exogenous Bcl-2 and chemical caspase inhibitors [11– 19], the presence of endogenous apoptosis inhibitors has not been investigated. If they do exist, then it is necessary to evaluate their functional contribution to this widely used experimental system. The BIR (baculovirus IAP repeat, in which IAP stands for inhibitor of apoptosis protein) domain is an evolutionarily conserved zinc-binding motif. Many, but not all, proteins containing the BIR domain (BIR fam- ily proteins) play important roles in apoptosis suppres- sion (reviewed in [20–25]). These proteins possess one or more repeats of the BIR domain, some of which are important for caspase inhibition. Some members also contain the C-terminal RING finger domain, which is reported to be required for the ubiquitylation of them- selves and bound caspases [26–28]. They are therefore good candidates as maternal apoptosis inhibitors if expressed in eggs. In this study, we identified four members of the BIR family of proteins from Xenopus eggs and characterized their biochemical behavior and functions. Results By a combination of EST database searches, PCR, and cDNA library screening, we isolated four cDNAs coding for BIR family proteins from Xenopus oocytes and eggs. Of these, two clones are related to mamma- lian survivin. One cDNA encodes a 160-amino-acid protein with a single BIR domain, which we named xSvv1 for Xenopus survivin1. This clone is nearly iden- tical with the Xenopus survivin ortholog recently iden- tified by several groups [29–31]. The other cDNA, which we designated xSvv2 for Xenopus survivin2, encodes a similar but distinct molecule of 157 amino- acid residues containing a single BIR domain. As a very similar protein, named SIX, was recently identi- fied as an RXRa-interacting protein [32], we refer to this molecule as xSvv2⁄ SIX. Amino-acid sequence alignment of xSvv1 and xSvv2 ⁄ SIX with human survi- vin is shown in Fig. 1A. The residues critical for zinc co-ordination, as well as the Thr residue phosphorylated by Cdc2 ⁄ cyclin B, are conserved in xSvv1 and xSvv2 ⁄ SIX. Putative orthologs of xSvv1 and xSvv2 ⁄ SIX were also identified in diploid Xenopus tropicalis (GenBank CR761050 and CR761512, respectively). Thus, xSvv1 and xSvv2 ⁄ SIX are distinct proteins rather than the nonallelic isoforms of a single protein in pseudotetra- ploid X. laevis. Another cDNA encodes a 401-residue protein con- taining two N-terminal BIR domains and a C-terminal RING finger domain, which we initially termed xEIAP for egg or embryonic IAP because of its strong expres- sion in oocytes and eggs. This domain organization is most similar to those of IAPs from pig, insect or viral origins, but no corresponding orthologs have been identified from human or mouse. While we were pre- paring this manuscript, Holley et al . [33] reported a new Xenopus BIR family protein, XLX, which is almost identical with xEIAP, and we refer to this molecule as xEIAP ⁄ XLX hereafter. The fourth cDNA encodes a protein of 412 amino acids with three N-terminal BIR domains and a C-ter- minal RING finger domain, the composition of which is most closely related to that of mammalian XIAP. We therefore named this molecule xXIAP for Xenopus or- tholog of XIAP, although we did not examine whether this gene is also linked to the X chromosome. Amino- acid sequence alignment of xEIAP ⁄ XLX and xXIAP with human XIAP is shown in Fig. 1B. xEIAP ⁄ XLX is moderately similar to human XIAP but lacks one BIR domain. xXIAP shows significant homology with human XIAP along the entire length despite the shorter spacer region between BIR3 and RING finger domains. All of the zinc co-ordinating residues in BIR and RING finger domains are completely conserved in both clones. For both xEIAP ⁄ XLX and xXIAP, several highly related sequences were found in the Xenopus tropicalis EST database. We therefore conclude that xXIAP is the structural ortholog of mammalian XIAP and that xEIAP ⁄ XLX is a new member of the BIR family of proteins in Xenopus. Xenopus egg apoptosis inhibitors Y. Tsuchiya et al. 2238 FEBS Journal 272 (2005) 2237–2250 ª 2005 FEBS Notably, amino-acid sequence analyses using the PESTfind program (http://www.at.embnet.org/embnet/ tools/bio/PESTfind/) revealed potential PEST sequen- ces in xSvv1 (+14.17 in 71–86 residues) and xEIAP ⁄ XLX (+8.21 in 246–259 residues and +5.58 in 333– 346 residues). PEST sequence is defined as a region enriched in Pro ⁄ Glu ⁄ Ser ⁄ Thr, and some proteins containing PEST sequences are rapidly degraded [34]. The presence of PEST sequences in xSvv1 and xEIAP ⁄ XLX suggests their instability in vivo. To carry out biochemical characterization, we bacte- rially expressed and purified a pair of His 6 -tagged and maltose-binding protein (MBP)-tagged recombinant fusions for each Xenopus BIR family protein (Fig. 2A). His 6 -tagged recombinants were all insoluble and used only for antibody production and purification, whereas Fig. 1. Primary sequences of Xenopus BIR family proteins. (A) Amino-acid sequence alignment of xSvv1, xSvv2, and human survivin. (B) Amino-acid sequence alignment of xEIAP ⁄ XLX, xXIAP, and human XIAP. Identical and similar residues are boxed and shaded, respectively. Zinc-co-ordinating Cys ⁄ His are indicated by asterisks at the bottom. The putative Cdc2 ⁄ cyclin B phosphorylation site in survivin is shown by (p). Xl, Xenopus laevis;Hs,Homo sapiens. The GenBank accession numbers of xSvv1, xSvv2 ⁄ SIX, human survivin, xEIAP ⁄ XLX, xXIAP, and human XIAP are AB197247, AB197249, NM_001168, AB197251, AB197252, and NM_001167, respectively. Y. Tsuchiya et al. Xenopus egg apoptosis inhibitors FEBS Journal 272 (2005) 2237–2250 ª 2005 FEBS 2239 MBP-tagged recombinants were all soluble and used for functional analyses as below. Our affinity-purified antibodies were all able to recognize as little as 1 ng of the corresponding recombinant antigens (Fig. 2B), but we could not detect the endogenous proteins in egg extracts directly by western blot, suggesting their low content. To investigate the existence of these proteins, the immunoprecipitated materials from 100 lL egg extracts with specific antibodies were similarly ana- lyzed. Although previous reports identified endogenous xSvv1 in egg extracts [29,30], we failed to detect it with our antibody (Fig. 2B). However, xSvv2 ⁄ SIX, xEIAP ⁄ XLX, and xXIAP were clearly detected with this procedure. Up to 5–10 ng xXIAP and smaller amounts of xSvv2 ⁄ SIX and xEIAP ⁄ XLX could be pre- cipitated from 100 lL cytostatic factor (CSF)-arrested egg extracts, although the recoveries were variable. These results indicate that xSvv2 ⁄ SIX, xEIAP ⁄ XLX, and xXIAP are present in egg extracts at low concen- tration (less than 1 lgÆmL )1 ). Recent studies have firmly established that the mor- phological change in sperm nuclei is a hallmark of apoptotic execution in Xenopus egg extracts [11–19]. The incubation of egg extracts at room temperature for 3–5 h converts spherical nuclei (I, Fig. 3A left) into degenerated ones (A, Fig. 3A left), although the time required depends on egg quality. This apoptotic change is accompanied by chromatin DNA degrada- tion (data not shown). Moreover, caspase-dependent partial (but not complete) proteolysis of p150 Glued , an endogenous apoptotic substrate in Xenopus egg extracts, is reproducibly detectable with a commercially available antibody (Fig. 3A right and [35]). All these events occur simultaneously and can be specifically inhibited by a pan-caspase inhibitor carbobenzoxy- l-valyl-l-alanyl-b-methyl-l-aspart-1-yl fluoromethane (Z-VAD-FMK), as previously reported [11–19]. We therefore used nuclear morphology and p150 Glued pro- teolysis as reliable indicators of apoptotic execution in this study. The molecule capable of preventing these processes can be regarded as an inhibitor of apoptosis. As all recombinants were fused with MBP, we selected nonfused MBP as a control protein. Preliminary stud- ies indicated that the addition of up to 10 lgÆmL )1 MBP had no effect on apoptotic timing and nuclear morphology in egg extracts. Therefore, all recombi- nants were added at the same protein (rather than equimolar) concentration at 10 lgÆmL )1 . The presence of 10 lgÆmL )1 control MBP, MBP-xSvv1, or MBP- xSvv2 showed little inhibition of this process, and MBP-xEIAP showed only limited inhibition (Fig. 3B). In contrast, the addition of MBP-xXIAP effectively inhibited apoptotic nuclear fragmentation. Moreover, caspase-dependent partial proteolysis of p150 Glued was also prevented only by MBP-xXIAP (Fig. 3C). Finally, we examined the dose–response of MBP-xXIAP, the only effective molecule in our assay. As shown in Fig. 3D, anti-apoptosis activity of MBP-xXIAP was strong at 10 lgÆmL )1 but decreased at 3 lgÆmL )1 , and became undetectable at 1 lgÆmL )1 . These results indicate that only xXIAP shows strong apoptosis- inhibiting activity in our assay with recombinant proteins. A B Fig. 2. Western blot analysis of Xenopus BIR family proteins in egg extracts. (A) Purity of recombinant BIR family proteins. Purified His 6 -tagged (left four lanes) and MBP-fused (right four lanes) rec- ombinants were resolved by SDS ⁄ PAGE and stained with Coomas- sie Brilliant blue. Lanes S1, xSvv1; lanes S2, xSvv2; lanes E, xEIAP ⁄ XLX, and lanes X, xXIAP. Positions of size markers (Mr) are indicated on the left. (B) Quantification of BIR family proteins in egg extracts. Immunoprecipitated materials from 100 lL CSF-arres- ted egg extracts with control (C) and the respective affinity-purified specific (I) antibodies were resolved by SDS ⁄ PAGE, followed by western blot with the same antibodies. For the quantification of each protein, 10, 3, and 1 ng of the corresponding recombinant His 6 -tagged antigen was subjected to western blot analysis at the same time. Positions of size markers are indicated on the left. Immunoprecipitated antigens and IgG heavy chain are denoted by arrowheads and asterisks, respectively. Xenopus egg apoptosis inhibitors Y. Tsuchiya et al. 2240 FEBS Journal 272 (2005) 2237–2250 ª 2005 FEBS Although exogenous xXIAP strongly suppressed apoptotic change in egg extracts, it is still not certain whether the same molecule actually functions as an endogenous apoptosis inhibitor in vivo. To examine the physiological roles biochemically, we blocked the func- tions of endogenous BIR family proteins by immuno- depletion with specific antibodies. The efficiency of depletion could not be accurately estimated because of the low contents of endogenous molecules, but excess amounts of antibodies were used to deplete the corres- ponding proteins from egg extracts (see Experimental procedures). The first immunodepletion appeared near complete for each BIR family protein because no endogenous molecule was detected when immunode- pletion was repeated and the second immunoprecipi- tate was subjected to western blot (data not shown). The nuclei in xXIAP-depleted extract became fragmen- ted significantly earlier than those in other immunode- pleted samples (Fig. 4A). Western blotting of p150 Glued also revealed the accelerated apoptotic cleavage only in xXIAP-depleted extract (Fig. 4B). These results strongly suggest that the physiological role of xXIAP is as a maternal apoptosis inhibitor in Xenopus egg extracts. The above results reveal that xSvv1, xSvv2 ⁄ SIX, and xEIAP ⁄ XLX could not inhibit apoptosis in egg extracts (Fig. 3B). However, these proteins may be unstable and completely degraded before the onset of apoptotic change. To examine the stability of these proteins, we first tried to detect the endogenous mole- cules by western blot to follow their quantitative chan- ges. However, reliable assay was not possible because of the low concentrations of the endogenous mole- cules. As an alternative approach, each of four MBP fusions was added at 1 lgÆmL )1 at the start of incuba- tion, and the remaining recombinants were detected by western blot with anti-MBP serum. As shown in Fig. 3D, all MBP fusions at this low concentration did A B C D Fig. 3. Apoptosis-inhibiting activities of recombinant BIR family pro- teins in egg extracts. (A) Indications of apoptosis in Xenopus egg extracts. Left, typical images of normal interphase (I) and fragmen- ted apoptotic (A ) nuclei stained with 4,6-diamidino-2-phenylindole. Bar, 10 lm. Right, Partial cleavage of p150 Glued in apoptotic extracts. p150 Glued in interphase (I) and apoptotic (A) extracts were detected by western blot. (B) Time course of apoptotic nuclear fragmentation in the presence of recombinant BIR family proteins. Each recombinant MBP fusion protein was added to interphase egg extracts at 10 lgÆmL )1 , and morphological changes in 50 nuclei were counted every hour for each sample. The percentages of apoptotic nuclei are represented as mean ± SEM from three inde- pendent experiments. (C) Apoptotic cleavage of p150 Glued in the presence of recombinant BIR family proteins. The extracts at 4 h incubation in one experiment in (B) were analyzed by western blot as shown in Experimental procedures. In B and C, the addition of each protein is denoted as below: s (lane C), control MBP; n (lane S1), MBP-xSvv1; m (lane S2), MBP-xSvv2; h (lane E), MBP-xEIAP; (lane X), MBP-xXIAP. (D) Dose–response of MBP-xXIAP. Various amounts of MBP-xXIAP were added to egg extracts, and the mor- phological changes in nuclei were counted as in (B). s,0lgÆmL )1 ; d,1lgÆmL )1 ; h,3lgÆmL )1 ; ,10lgÆmL )1 . Data are represented as mean ± SEM from three independent experiments. Y. Tsuchiya et al. Xenopus egg apoptosis inhibitors FEBS Journal 272 (2005) 2237–2250 ª 2005 FEBS 2241 not affect the timing of apoptotic onset. The amount of MBP-xSvv1 decreased only slightly, and that of MBP-xSvv2 was virtually unchanged during a 4-h incubation (Fig. 5A, upper). These results indicate that both recombinants were stable and remained in suffi- cient amounts in extracts, but that they could not inhi- bit apoptosis (Fig. 3B,C). In contrast, the signal of MBP-xEIAP rapidly dimin- ished before the onset of apoptotic change and had completely disappeared after 4 h of incubation. MBP- xXIAP was also gradually degraded but more slowly than MBP-xEIAP. These results raised the possibility that anti-apoptotic activities of xEIAP ⁄ XLX and xXIAP may be fully or partially masked by their insta- bilities. To confirm the above results, another set of studies was carried out using recombinants translated in vitro in rabbit reticulocyte lysates. Although the sta- bility of His 6 -xSvv1 was difficult to analyze because of low translation efficiency, the three other recombinants showed similar behaviors to corresponding MBP fusions; His 6 -xSvv2 was stable, His 6 -xEIAP was rap- idly degraded, and His 6 -xXIAP was relatively stable at 2 h but degraded at 4 h (Fig. 5A, lower). Similar results were obtained with untagged recombinants (data not shown), excluding the possibility of struc- tural interference by the N-terminal tags. These data suggest that the lack of apoptotic inhibi- tion by MBP-xEIAP, as well as MBP-xXIAP at low concentration, may be due to its rapid proteolysis. Thus, we focused our attention on the stabilities of xEIAP ⁄ XLX and xXIAP. Recent studies have identi- fied at least four independent mechanisms that control the stability of BIR family proteins; RING finger- dependent ubiquitylation and subsequent degradation by proteasome [26–28], caspase-dependent cleavage [36–38], calpain-dependent degradation [39], and pro- teolysis by a serine protease Omi ⁄ HtrA2 released from mitochondria [40–42]. Moreover, proteolytic fragments may be further degraded by the N-end rule pathway in Drosophila [43]. To evaluate the contribution of these proteolytic systems, the effects of specific protease inhibitors were tested against the degradation of the two BIR family proteins as well as apoptotic p150 Glued cleavage. Although the pan-caspase inhibitor Z-VAD-FMK at 100 lm effectively blocked the apoptotic cleavage of p150 Glued and nuclear fragmentation, the initial decrease in the two BIR family proteins within 2 h was not inhibited (Fig. 5B,C). However, the amounts of MBP-xEIAP and MBP-xXIAP did not change sig- nificantly during the next 2 h, indicating that cleavage of these proteins during this period was blocked by the inhibitor. These results suggest that the cleavage of these proteins proceeded in two steps: a caspase-inde- pendent first step and caspase-dependent second step. In contrast, a proteasome ⁄ calpain inhibitor, MG-132 (carbobenzoxy-l-leucyl-l-leucyl-l-leucinal), at 100 lm could not prevent, or even accelerated, the apoptotic cleavage of p150 Glued , the decreases in MBP-xEIAP ⁄ XLX and MBP-xXIAP, and nuclear fragmentation (Fig. 5B,C). With the same inhibitor concentration, proteolysis of endogenous cyclin B2 (a typical sub- strate of the ubiquitin–proteasome pathway) during the calcium-induced metaphase exit was effectively blocked (data not shown). UCF-101, a recently devel- oped specific inhibitor of HtrA family proteases [44], had no effect at 100 lm (data not shown) and acceler- ated the cleavage of these three proteins and nuclear fragmentation at 1 mm. It should be noted that B A Fig. 4. Acceleration of apoptotic onset in xXIAP-depleted egg extracts. (A) Time course of apoptotic nuclear fragmentation in interphase extracts depleted of each BIR family protein. The mor- phological changes in the nuclei were counted as in Fig. 3B. Data are represented as mean ± SEM from three independent experi- ments. (B) Apoptotic cleavage of p150 Glued in interphase extracts depleted of each BIR family protein. The extracts at 3 h incubation in one experiment in (A) were analyzed as in Fig. 3C. Depletion of each protein is denoted as follows: s (lane C), control-depleted; n (lane S1), xSvv1-depleted; m (lane S2), xSvv2-depleted; h (lane E), xEIAP ⁄ XLX-depleted; (lane X), xXIAP-depleted. Xenopus egg apoptosis inhibitors Y. Tsuchiya et al. 2242 FEBS Journal 272 (2005) 2237–2250 ª 2005 FEBS inhibitors of the lysosomal proteases (leupeptin, chy- mostatin, and pepstatin at 10 lgÆmL )1 ) were also included in the extracts, excluding the possibility of nonspecific lysosomal proteolysis. These data suggest that both proteins are degraded at least in two steps, initially by an unknown proteolytic system and subse- quently by activated caspases. To define the domains that regulate the function and stability of xEIAP ⁄ XLX and xXIAP, we further analyzed the stabilities of three C-terminal deletion mutants, ED1, ED2, and XD1, as in Fig. 5. ED1 lacks the second PEST sequence (PEST2) and RING finger domain of xEIAP ⁄ XLX, and the first PEST sequence (PEST1) is further deleted in ED2 (Fig. 6A). Similarly, the XD1 mutant lacks the RING finger domain and C-terminal half of the BIR3 domain (Fig. 6A). Each of these three mutants was fused with N-terminal MBP, expressed in Escherichia coli, and purified for biochemical characterization (Fig. 6B). Western blot revealed that, in contrast with the full-length version, both MBP-ED1 and MBP-ED2 were stable during the first 2 h in interphase egg extracts without any detect- able decreases, indicating that the C-terminal region confers susceptibility on the initial phase of degrada- tion (Fig. 6C). During the next 2 h, however, MBP- fused ED1 and ED2 both showed remarkable decreases. Most of the MBP-ED2 was cleaved into smaller pep- tides during this period, whereas only about a half of the MBP-ED1 was degraded. These data indicate that MBP-ED1 is more stable than MBP-ED2 because of the existence of residues 219–269. Moreover, cleaved fragments from MBP-ED1 and MBP-ED2 generated after 4 h incubation appeared more stable than those derived from full-length MBP-xEIAP (compare Figs 5A and 6C). MBP-XD1 was also resistant to ini- tial degradation within 2 h but was cleaved during the next 2 h (Fig. 6C). Similar to the cases of MBP- ED1 ⁄ ED2, cleavage products from MBP-XD1 were also more stable than those from full-length MBP-xXIAP (compare Figs 5A and 6C). Thus, in both xEIAP ⁄ XLX and xXIAP, the C-terminal region containing the RING finger domain is required for the initial phase of proteolysis within 2 h and also for the final degra- dation of caspase-generated fragments. A B C Fig. 5. Different stabilities of BIR family proteins in Xenopus inter- phase egg extracts. (A) Stability of BIR family proteins in interphase egg extracts. Purified MBP-tagged recombinants at 1 lgÆmL )1 (upper) or rabbit reticulocyte lysates containing 35 S-labeled His 6 - tagged recombinants at 0.1 volume (lower) were supplied to inter- phase egg extracts. The remaining proteins after 0, 2, and 4 h incubation were detected by western blot using anti-MBP serum (aMBP) and image analyzer ( 35 S), respectively. (B) Effects of prote- ase inhibitors on the degradation of p150 Glued , MBP-xEIAP, and MBP-xXIAP. Interphase egg extracts containing 1 lgÆmL )1 MBP- xEIAP or MBP-xXIAP were incubated for 0, 2, and 4 h in the pres- ence of dimethylsulfoxide, 100 l M Z-VAD-FMK, 100 lM MG-132, and 1 m M UCF-101, respectively. p150 Glued and the remaining MBP fusion proteins were detected as in Figs 3C and 5A, respectively. (C) Effects of protease inhibitors on apoptotic nuclear fragmenta- tion. The morphological changes in nuclei were counted as in Fig. 3B. s, control dimethylsulfoxide; d, 100 l M MG-132; n, 100 l M Z-VAD-FMK; m,1mM UCF-101. Data are represented as mean ± SEM from three independent experiments. Y. Tsuchiya et al. Xenopus egg apoptosis inhibitors FEBS Journal 272 (2005) 2237–2250 ª 2005 FEBS 2243 Nuclear fragmentation and p150 Glued cleavage assays indicate that, although MBP-ED1 did not inhibit apop- tosis despite its stability, MBP-ED2 significantly pre- vented apoptosis (Fig. 6D,E). These data suggest that the apoptosis-inhibiting potential of the two N-ter- minal BIR domains in xEIAP ⁄ XLX is self-inhibited by residues 219–269. In contrast, MBP-XD1 retained anti- apoptotic function as strong as full-length MBP- xXIAP despite the lack of functional BIR3 and RING finger domains (Fig. 6D,E). A recent report indicated that Xenopus CSF-arres- ted metaphase egg extracts are more resistant to apoptosis than interphase extracts by a mitogen-acti- vated protein kinase (MAPK)-dependent, post-cyto- chrome c release mechanism [45]. To check whether this resistance involves any of the BIR family pro- teins, we first immunodepleted endogenous molecules and studied the timing of apoptotic onset in CSF- arrested egg extracts containing cycloheximide. Because nuclear morphology could not be observed because of mitotic chromatin condensation in meta- phase extracts, only the apoptotic cleavage of p150 Glued was examined. In control-depleted CSF- arrested extract, apoptotic p150 Glued cleavage became evident after 6 h incubation, significantly later than in interphase extract (Fig. 7A). Immunodepletion of xSvv1, xSvv2, or xEIAP ⁄ XLX showed no effect on the timing of apoptotic onset (data not shown). In contrast, p150 Glued cleavage started after only 2 h incubation in xXIAP-depleted CSF-arrested extract, as quickly as in xXIAP-depleted interphase extract (compare Figs 4B and 7A). Therefore, in the absence of xXIAP, both CSF-arrested and interphase egg extracts executed apoptosis with similar timing. To examine whether this difference between CSF- arrested and interphase egg extracts can be attributed to the activities of MAPK or other kinases, we used several protein kinase inhibitors and compared the apoptotic cleavage of p150 Glued as well as the stabil- ity of MBP-xXIAP. Apoptotic p150 Glued cleavage did AB C E D Fig. 6. Apoptosis-inhibiting activities of MBP-xEIAP and MBP-xXIAP C-terminal deletion mutants. (A) Structures of deletion mutants. BIR domain (gray ovals), RING finger domain (black boxes) and PEST sequence (gray boxes) are schematically presented. (B) Purity of recombin- ant proteins. Purified MBP-tagged mutants were detected as in Fig. 2B. Positions of size markers are indicated on the left. (C) Stability of deletion mutants in interphase egg extracts. Purified MBP-tagged mutants at 1 lgÆmL )1 were supplied to interphase egg extracts. The remaining proteins after 0, 2, and 4 h incubation were detected as in Fig. 5A. In XD1, a smaller truncated product was also observed, which was poorly detected by protein staining in B. (D) Time course of apoptotic nuclear fragmentation in the presence of recombinant mutants. The morphological changes in nuclei were counted as in Fig. 3B. Data are represented as mean ± SEM from three independent experi- ments. (E) Apoptotic cleavage of p150 Glued in the presence of recombinant mutants. The extracts at 3 h incubation in one experiment in (D) were analyzed as in Fig. 3C. (C, D, E) The addition of each protein is denoted as follows: s (lane C), MBP; d (lane ED1), MBP-ED1; n (lane ED2), MBP-ED2; m (lane XD1), MBP- XD1. Xenopus egg apoptosis inhibitors Y. Tsuchiya et al. 2244 FEBS Journal 272 (2005) 2237–2250 ª 2005 FEBS not occur until 6 h in dimethylsulfoxide-treated control CSF-arrested egg extracts (Fig. 7B, left). Whereas a Cdk inhibitor, roscovitine, at 100 lm had little effect, an MEK inhibitor, U0126, at 100 lm significantly accelerated caspase-dependent cleavage of p150 Glued to interphase level (4 h). The degrada- tion of MBP-xXIAP was also delayed in CSF-arres- ted egg extracts (compare right, upper panels of Figs 5B and 7B; dimethylsulfoxide). This delay was mainly due to the suppression of the second phase of the cleavage because the initial degradation within 2 h in CSF-arrested egg extracts was similar to that in interphase egg extracts. The presence of dimethyl- sulfoxide or roscovitine did not affect these results (Fig. 7B, right). However, the second caspase-depend- ent apoptotic cleavage of MBP-xXIAP during 2–4 h incubation, which had been almost completely sup- pressed in CSF-arrested egg extracts, was recovered in the presence of U0126 (Fig. 7B, right). In con- trast, the initial degradation of MBP-xXIAP was not affected by U0126. However, MAPK did not phos- phorylate MBP-xXIAP in vitro (data not shown). Thus, in CSF-arrested egg extracts, MAPK appears to suppress apoptosis by delaying caspase activation rather than stabilizing xXIAP by direct phosphoryla- tion. Discussion xXIAP is the central apoptosis inhibitor in Xenopus egg extracts In this study, we identified four BIR family proteins from Xenopus eggs and evaluated their biochemical and physiological roles as apoptosis inhibitors. Our recombinant protein addition (Fig. 3) and immunode- pletion (Figs 4 and 7A) analyses indicated that xXIAP, a putative ortholog of mammalian XIAP, functions as a genuine maternal apoptosis inhibitor in egg extracts, and probably also in eggs in vivo. Numerous studies have indicated that XIAP is the most potent IAP in mammals (reviewed in [20–25]). Thus, the central role of XIAP as a physiological apoptosis inhibitor may be conserved among animal species. BIR3 and RING fin- ger domains of xXIAP were dispensable for apoptosis inhibition in our assay (Fig. 6). This result is consistent with the structural data for mammalian XIAP showing that effector caspases (caspases-3 ⁄ 7) are inhibited by the linker peptide between BIR1 and BIR2 domains, which is also conserved in xXIAP. xEIAP/XLX is a potential apoptosis inhibitor in Xenopus egg extracts xEIAP ⁄ XLX was less effective in apoptotic inhibition (Fig. 3), and its immunodepletion did not affect the timing of apoptotic onset in egg extracts (Fig. 4). Although Holley et al. [33] did not examine the func- tion of xEIAP ⁄ XLX, our assay using deletion mutants indicates that residues 1–218 of xEIAP ⁄ XLX contain- ing two BIR domains could inhibit apoptosis (Fig. 6). However, the presence of the residue 219–269-contain- ing PEST1 blocked its apoptosis-inhibiting function, and the further addition of the residue 270–401-con- taining PEST2 and RING finger domain resulted in rapid proteolysis. Therefore, the apoptosis-inhibiting potential of xXIAP ⁄ XLX may be normally hindered by both self-inhibition and rapid degradation in Xen- opus eggs. However, we do not exclude possible func- tions of xEIAP ⁄ XLX in other cell death pathways. We are currently searching physiological conditions in which xEIAP ⁄ XLX becomes functional. Multiple proteolytic pathways regulate the concentrations of BIR family proteins Both xEIAP ⁄ XLX and xXIAP were degraded in a multistep process in egg extracts. During the first step operating within 2 h of incubation, xEIAP ⁄ XLX was degraded rapidly, whereas xXIAP was degraded rather A B Fig. 7. Critical role of xXIAP in the increased resistance against apoptosis in CSF-arrested egg extracts. (A) Delay of apoptotic onset in CSF-arrested extracts was compromised by xXIAP immunodeple- tion. CSF-arrested extracts were depleted with control (left) or anti- xXIAP (right) antibody, supplied with cycloheximide and incubated. The time course of apoptotic p150 Glued cleavage every 2 h was analyzed by western blot as in Fig. 3C. (B) MAPK-dependent indi- rect stabilization of xXIAP in CSF-arrested egg extracts. MBP-xXIAP was supplied at 1 lgÆmL )1 to CSF-arrested metaphase egg extracts containing cycloheximide and incubated for 0, 2, and 4 h in the presence of dimethylsulfoxide (DMSO), 100 l M roscovitine, and 100 l M U0126. p150 Glued and the remaining MBP fusion proteins were similarly detected as in Figs 3C and 5A, respectively. Y. Tsuchiya et al. Xenopus egg apoptosis inhibitors FEBS Journal 272 (2005) 2237–2250 ª 2005 FEBS 2245 slowly. For both molecules, this initial degradation step required the RING finger domain-containing C-terminal region, and could not be blocked by inhibi- tors of caspases, proteasome ⁄ calpain, and HtrA family proteases. After longer incubation, the second caspase- dependent step further cleaved the remaining molecules (Fig. 5) and apoptosis was triggered. In addition, the final digestion of cleavage fragments also partially depended on the RING finger domain because caspase-generated fragments were detected more clearly in deletion mutants than in full-length versions (compare Figs 5 and 6). This final step was also resist- ant to MG-132 and UCF-101 (our unpublished result). The involvement of these multiple proteolytic systems can explain why recombinant MBP-xXIAP at 10 lgÆmL )1 , but not at 1 lgÆmL )1 , effectively delayed apoptosis onset (compare Figs 3 and 5). Although the second step was clearly caspase-dependent, we were not able to determine which proteolytic mechanisms were responsible for the first and the last degradation steps. Whereas previous reports suggested the involve- ment of the ubiquitin–proteasome system in the RING finger-dependent and N-end rule pathways [26–28,43], our data indicate that neither the initial nor the final degradation step was blocked by proteasome inhibi- tion. Holley et al. [33] also reported the reduction of xEIAP ⁄ XLX by caspase-dependent and translational inhibition-dependent dual mechanisms. Cycloheximide, a protein synthesis inhibitor, is required for Xenopus egg extracts to be used for in vitro apoptosis study. The main purpose of this drug is to prevent cyclin B reaccumulation and to arrest the cell cycle in inter- phase, but general protein translation is also inhibited. Holley et al. found that, in Xenopus egg extracts, the cell death induced by reaper involves general trans- lation inhibition, probably in a similar manner to cycloheximide-induced apoptosis. Otherwise, the stabil- ities of these proteins may be regulated by conforma- tional changes or stoichiometric interaction with other molecules. Further studies are being performed to address these issues. As the number of mitochondria greatly increases during Xenopus oogenesis, unwanted cytochrome c release from mitochondria may occur spontaneously even in healthy eggs. The level of endogenous xXIAP may be sufficient to prevent caspase amplification induced by a low level of cytochrome c release. Caspase inhibition can further prevent the caspase- dependent degradation of xXIAP, thereby forming a positive feedback loop. Once translation inhibition or aging ⁄ damage-induced enhanced cytochrome c release leads to consumption of all endogenous xXIAP, caspase amplification may occur rapidly. This mechanism can convert gradual cytochrome c release into all-or-none apoptotic execution. Therefore, unper- turbed synthesis of xXIAP may be at least in part necessary for apoptosis suppression in egg extracts in vitro and for egg survival in vivo. xXIAP stability and MAPK-dependent apoptosis inhibition A current model of CSF function is to inhibit the ana- phase-promoting complex to protect several cell cycle regulators (cyclin B and securin, for example) from ubiquitylation and proteasome-dependent proteolysis, leading to cell cycle arrest at metaphase. In CSF-arres- ted egg extracts, apoptotic onset was delayed in a MAPK-dependent manner (Fig. 7B and [45]). How- ever, in the absence of endogenous xXIAP, both CSF- arrested and interphase egg extracts executed apoptosis with similar timing (2 h; compare Figs 4B and 7A), indicating that MAPK-induced delay of apoptosis depends on the presence of xXIAP. Moreover, as the degradation of xXIAP is accelerated by MAPK kinase (MEK) inhibition, this also indicates that MAPK increases the stability of xXIAP. Therefore, it is rea- sonable to consider that MAPK delays apoptosis by stabilizing xXIAP. This MAPK-induced delay of xXIAP degradation is mainly due to the inhibition of the caspase-dependent second phase because the initial caspase-independent degradation of xXIAP is not sig- nificantly altered in CSF-arrested egg extracts. At first, we speculated that MAPK stabilizes xXIAP by direct phosphorylation of the protein. However, recombinant MBP-xXIAP was not significantly phosphorylated by MAPK or CSF-arrested egg extracts (our unpublished result). MAPK may stabilize xXIAP indirectly by sup- pressing caspase activation or by inhibiting caspase activities themselves. Further studies will be necessary to identify the molecular determinant and signal-trans- duction pathway for MAPK-dependent egg survival. Survivin orthologs do not inhibit apoptosis in Xenopus egg extracts We also isolated two survivin-related molecules, xSvv1 and xSvv2 ⁄ SIX, but failed to show their anti-apoptosis activities. Whereas survivin was originally identified as an apoptosis inhibitor, recent studies have suggested that survivin-related proteins of yeasts, insects (except Drosophila deterin) and vertebrates play important roles in cell cycle progression rather than caspase inhi- bition (reviewed in [20–25]). Survivin forms a ‘chromo- somal passenger complex’ with INCENP and Aurora B kinase to regulate sister chromatid segregation and Xenopus egg apoptosis inhibitors Y. Tsuchiya et al. 2246 FEBS Journal 272 (2005) 2237–2250 ª 2005 FEBS [...]... signal transduction in Xenopus egg extracts [46] Experimental procedures Chemical reagents Unless otherwise stated, all the reagents were purchased from Wako Pure Chemicals (Osaka, Japan) cDNA cloning of Xenopus BIR family proteins During the course of another unrelated study, we unexpectedly obtained a Xenopus egg cDNA fragment coding for a RING finger motif highly homologous to those of vertebrate IAPs... His6 -BIR family proteins, 1 mm isopropyl b-d-thiogalactoside and 1 mm ZnSO4 were added, and protein expression was induced for 3 h at 37 °C with vigorous shaking As all the His6 -BIR family proteins were insoluble, inclusion body fractions were prepared according to the manufacturer’s instructions (Novagen) Each recombinant was solubilized with binding buffer (0.1 m Tris ⁄ HCl, pH 8.0, 0.5 m NaCl) containing... screening of a Xenopus oocyte kZapII cDNA library (kindly provided by Professor Haruhiko Takisawa and Dr Satoru Mimura, Osaka University, Osaka, Japan) We identified an ORF encoding two N-terminal BIR domains and a C-terminal RING finger motif, which was designated xEIAP for Xenopus egg or embryonic IAP While the initial phase of this study was in progress, many Xenopus EST clones homologous to mammalian BIR. .. many Xenopus EST clones homologous to mammalian BIR family proteins were deposited in the GenBank database Homology searches revealed the presence of other BIR family proteins expressed in Xenopus oocytes or eggs By exploiting this information, we were also able to isolate the clones encoding survivin1 (xSvv1), survivin2 (xSvv2), and an apparent ortholog of mammalian XIAP (xXIAP) from the same oocyte cDNA... al cytokinesis Although we could not detect xSvv1 by our western blot, Bolton et al [29] and Losada et al [30] identified xSvv1 in egg extracts as a stoichiometric binding partner of INCENP and Aurora B kinase The reason for this discrepancy is unknown and under investigation Instead, our western blot detected xSvv2 ⁄ SIX in egg extracts [32] As Xenopus egg extracts are also useful for studying the... corresponding to 0.4 lL egg extracts were subjected to western blot using monoclonal anti-p150Glued Ig (1 : 1000 dilution; BD Transduction Laboratory, San Diego, CA, USA) as described previously [35] Stability analyses of BIR proteins in CSF-arrested and interphase egg extracts MBP-fused BIR family proteins were supplied at 1 lgÆmL)1 to CSF-arrested or interphase extracts, both containing 0.1 mgÆmL)1 cycloheximide... added to extracts (1% dimethylsulfoxide) and similarly analyzed after 2 and 4 h of incubation All egg extracts also contained 10 lgÆmL)1 leupeptin, pepstatin, and chymostatin (Peptide Institute) to inhibit nonspecific lysosomal proteolysis [13,16,47] For the stability assay of radiolabeled proteins, in vitro translated rabbit reticulocyte lysates and egg extracts were mixed at 1 : 9 and similarly incubated... GS & Duckett CS (2002) IAP proteins: blocking the road to death’s door Nat Rev Mol Cell Biol 3, 401–410 Silke J & Vaux DL (2001) Two kinds of BIR- containing protein: inhibitors of apoptosis, or required for mitosis J Cell Sci 114, 1821–1827 Yang Y, Fang S, Jensen JP, Weissman AM & Ashwell JD (2000) Ubiquitin protein ligase activity of IAPs and 2249 Xenopus egg apoptosis inhibitors 27 28 29 30 31 32... using StuI and HindIII, and MBP-XD1 (1–301 residues of xXIAP and an extra 9 residues derived from an alternative reading frame, lacking RING finger and a part of the BIR3 domain) using SphI For large-scale preparation of recombinant proteins, the vectors were transformed into E coli BL21 (DE3) strain, and bacterial overnight culture was inoculated into Luria– Bertani medium and grown at 37 °C with agitation... 0.5 m NaCl) containing 6 m guanidine hydrochloride, and manually injected using a syringe into a nickel-loaded HiTrap chelating column (1 mL; Amersham Biosciences, Tokyo, Japan) equilibrated with binding buffer containing 6 m urea The column was washed with 2· Hepes buffer (20 mm Hepes ⁄ KOH, pH 7.8, 200 mm KCl, 2 mm MgCl2) containing 6 m urea and 20 mm imidazole, and recombinant protein was eluted with . Different stabilities of BIR family proteins in Xenopus inter- phase egg extracts. (A) Stability of BIR family proteins in interphase egg extracts. Purified MBP-tagged recombinants at 1 lgÆmL )1 (upper). repeat, in which IAP stands for inhibitor of apoptosis protein) domain is an evolutionarily conserved zinc-binding motif. Many, but not all, proteins containing the BIR domain (BIR fam- ily proteins) . Apoptosis-inhibiting activities of BIR family proteins in Xenopus egg extracts Yuichi Tsuchiya, Shin Murai and Shigeru Yamashita Department of Biochemistry, Toho University School of Medicine,