OncJuly3 6..6 - Perelman School of Medicine at the University of ...

Oncogene (1999) 18, 4153 ± 4159 ã 1999 Stockton Press All rights reserved 0950 ± 9232/99 $12.00 http://www.stockton-press.co.uk/onc

SHORT REPORT

The antiapoptotic decoy receptor TRID/TRAIL-R3 is a p53-regulated DNA damage-inducible gene that is overexpressed in primary tumors of the gastrointestinal tract M Saeed Sheikh*,1, Ying Huang2, Ester A Fernandez-Salas3, Wa®k S El-Deiry4, Helmut Friess5, Sally Amundson1, Jing Yin6, Stephen J Meltzer6, Nikki J Holbrook2 and Albert J Fornace Jr1 1

Division of Basic Sciences, National Cancer Institute, National Institutes of Health, Bethesda, Maryland MD 20892, USA; 2Gene Expression and Aging Section, Laboratory of Biological Chemistry, National Institute on Aging, National Institutes of Health, Baltimore, Maryland MD 21224, USA; 3Laboratory of Carcinogenesis, National Cancer Institute, National Institutes of Health, Bethesda, Maryland MD 20892, USA; 4Laboratory of Molecular Oncology, Howard Hughes Medical Institute, University of Pennsylvania School of Medicine, Philadelphia, Pennsylvania PA 19104, USA; 5Department of Visceral and Transplantation Surgery, University of Bern, Inselspital CH-3010 Bern, Switzerland; 6Department of Medicine, GI Division, University of Maryland at Baltimore and Baltimore VA Hospital, Maryland MD 21201, USA

Both DR4 and DR5 have recently been identi®ed as membrane death receptors that are activated by their ligand TRAIL to engage the intracellular apoptotic machinery. TRID (also named as TRAIL-R3) is an antagonist decoy receptor and lacks the cytoplasmic death domain. TRID protects from TRAIL-induced apoptosis by competing with DR4 and DR5 for binding to TRAIL. TRID has been shown to be overexpressed in normal human tissues but not in malignantly transformed cell lines. DR5 is a p53-regulated gene and we have recently reported that DR5 expression is induced in response to genotoxic stress in both a p53-dependent and independent manner (Sheikh et al., 1998). In the current study, we demonstrate that TRID gene expression is also induced by the genotoxic agents ionizing radiation and methyl methanesulfonate (MMS) in predominantly p53 wild-type cells, whereas UV-irradiation does not induce TRID gene expression. Consistent with these results, exogenous wild-type p53 also upregulates the expression of endogenous TRID in p53-null cells. Thus, TRID appears to be a p53 target gene that is regulated by genotoxic stress in a p53-dependent manner. Using primary gastrointestinal tract (GIT) tumors and their matching normal tissue, we also demonstrate for the ®rst time that TRID expression is enhanced in primary tumors of the GIT. It is, therefore, possible that TRID overexpressing GIT tumors may gain a selective growth advantage by escaping from TRAIL-induced apoptosis. Keywords: p53; DNA damage; apoptosis Apoptosis is a genetically controlled program of cell death (reviewed by Issac, 1994; Cleveland and Ihle, 1995). Apoptotic elimination of cells during physiological and pathological processes is important for normal tissue homeostasis (reviewed by Issac, 1994; Cleveland and Ihle, 1995). Recent studies have

*Correspondence: MS Sheikh, Rm 5 C09, Bldg. 37, Division of Basic Sciences, National Cancer Institute, National Institutes of Health, Bethesda, Maryland 20892, USA Received 9 October 1998; revised 7 January 1999; accepted 16 February 1999

demonstrated that the program of cell death is controlled via a number of signaling events (reviewed by Nagata, 1997). These signals, some of which are mediated via membrane death receptors, are either protective or inductive of cell death (reviewed by Cleveland and Ihle, 1995). Tumor necrosis factor receptor 1 (TNFR1) and Fas are protoypes of cell surface receptors that are thought to mediate death signals (reviewed by Cleveland and Ihle, 1995; Nagata, 1997). Other cell surface receptors that belong to this family of membrane death receptors (DRs) include DR3, DR4 and KILLER/DR5 (hereafter referred to as DR5) (reviewed by Cleveland and Ihle, 1995; Nagata, 1997). Although the ligand for DR3 remains unknown, both DR4 and DR5 are activated by their ligand, named TRAIL or APO2L (Pan et al., 1997; Sheridan et al., 1997; MacFarlane et al., 1997). These receptors contain an intracellular region of homology designated as the death domain (Pan et al., 1997; Sheridan et al., 1997; MacFarlane et al., 1997). Ligand-dependent and -independent activation of these receptors involves multimerization with subsequent recruitment of intracellular adaptor molecules to relay signals inside the cell (reviewed by Nagata, 1997). The apoptotic signals originating from these death receptors converge on the proximal caspases, such as caspase 8 and caspase 10, and subsequently engage the downstream caspase cascade leading to cleavage of death substrates and ®nally cell death (reviewed by Alnemri, 1997; Salvesen and Dixit, 1997). Two additional membrane antagonist decoy receptors that belong to the above family of death receptors have also been identi®ed. These decoy receptors known as TRID/TRAIL-R3 (hereafter refer to as TRID) (Pan et al., 1997; Sheridan et al., 1997; MacFarlane et al., 1997) and TRUNDD (Pan et al., 1998; Marsters et al., 1997) either lack the death domain or contain a mutated version of the death domain and consequently protect cells from the cytotoxic eects of TRAIL. Interestingly, the antagonist decoy receptor TRID was found to be constitutively overexpressed in a majority of normal human tissues including peripheral blood leukocytes, spleen, heart, placenta, liver, lung, skeletal muscle, kidney and pancreas. Eight dierent cell lines representing various tumor types, by

DNA damage inducible TRID overexpression in primary tumors MS Sheikh et al

4154

contrast, displayed very low levels of TRID message (Pan et al., 1997; Sheridan et al., 1997; MacFarlane et al., 1997). DR4 and DR5 mRNAs, on the other hand, have been found to be expressed in both normal human tissues and cancer cell lines (Pan et al., 1997; Sheridan et al., 1997; MacFarlane et al., 1997). Since TRID can protect from TRAIL-induced apoptosis, the selective overexpression TRID in normal tissues suggests the therapeutic potential of TRAIL as an anti-cancer agent to preferentially target tumors and not normal tissues (Pan et al., 1997; Sheridan et al., 1997; MacFarlane et al., 1997). The above mentioned studies have, however, assessed the expression pro®le of TRID in cancer cells lines and in the unmatched normal tissues. Further studies are, therefore, needed to evaluate the expression pro®le of TRID in primary tumors and their matching normal tissues before a therapeutic use of TRAIL as an anti-cancer agent can be interrogated. The death receptor DR5 has been identi®ed as p53regulated gene (Wu et al., 1997). We have recently demonstrated that DR5 is regulated both in a p53dependent and independent manner in response to genotoxic stress and TNFa (Sheikh et al., 1998). Ionizing radiation (IR) induction of DR5 expression occurred in a p53-dependent manner, whereas methyl methanesulfonate (MMS) induced DR5 expression in p53 wild-type and mutant cells; ultraviolet (UV)irradiation by contrast did not induce DR5 expression (Sheikh et al., 1998). Our earlier ®ndings demonstrated that the TRID mRNA levels were also enhanced by genotoxic agent IR in one p53 wild-type cell line tested (Sheikh et al., 1998). More detailed studies are, however, needed to investigate the genotoxic stress regulation of TRID gene expression. The aims of this study were twofold: (i) to investigate in more detail the genotoxic stress regulation of TRID and to ascertain whether TRID is also a p53-regulated gene, (ii) to investigate the expression pro®le of TRID in primary tumors and their matching normal tissues. We used a panel of p53 wild-type and p53-negative cell lines to assess the eects of genotoxic stresses IR, MMS and UV-irradiation on TRID gene expression. One hallmark of p53-regulated genes is that they are frequently induced following IR exposure in p53 wildtype cells. Consistent with this notion, TRID induction in response to IR occurred predominantly in p53 wildtype cells (Figure 1a). MMS also strongly enhanced the TRID mRNA levels only in p53 wild-type cell lines whereas its eect was either absent or weak in p53negative cells (Figure 1b). UV-irradiation, on the other hand, did not alter TRID mRNA levels in the majority of cell lines tested and minimally upregulated TRID mRNA levels in RKO human colon carcinoma cells (Figure 1b). Figure 1c shows a representative Northern blot illustrating the eects of IR, MMS and UV on TRID expression in p53 wild-type A549 cells. Taken together, the above results demonstrate that the induction of TRID gene expression in response to DNA damage appears to predominantly occur in p53 wild-type cells. We next investigated whether TRID is indeed a p53regulated gene. To address this issue, we stably transfected the p53-null H1299 cells with an expression vector carrying a human temperature sensitive p53 (p53Ala143). At 378C, p53Ala143 is in a mutant conforma-

tion whereas it has been shown to acquire wild-type conformation at 328C (Friedlander et al., 1996). Several isolated clones as well as pooled clones were characterized for the expression of exogenous p53. Two isolated clones expressing the exogenous p53 were pooled and selected for further studies. Figure 2a shows the expression and nuclear localization of the exogenously introduced p53 in the stable transfectants, whereas vector-transfected cells do not exhibit such a staining pattern. To ascertain the functionality of the exogenous p53, both vector-transfected and p53Ala143transfected cells were subjected to permissive temperature (328) for 48 h and processed for immunofluorescent staining to detect the levels of endogenous p21Waf1. As shown in Figure 2a, vector-transfected cells exhibited very weak p21Waf1 staining, whereas the p53Ala143 transfected cells growing at 328C displayed strong p21Waf1 speci®c nuclear staining, suggesting that the p53 target p21Waf1 was upregulated in these cells and thus substantiating the functionality of exogenous p53. Northern blot analysis demonstrated that the endogenous p21WAF1 was clearly induced at the mRNA level when cells were shifted to permissive temperature (328C), further con®rming the functionality of exogenous temperature-sensitive p53 in these cells. TRID gene expression was next monitored in these cells grown at non-permissive (378C) and permissive (328) temperatures. Northern blot analyses were performed on RNAs obtained from the p53Ala143transfected cells grown at permissive temperature (328C) for 6, 24 and 48 h, whereas vector-transfected cells were used as controls to monitor the eect of temperature shift on the levels of endogenous TRID mRNA. A representative Northern blot shown in Figure 2b demonstrates that TRID mRNA levels were clearly increased when p53Ala143 containing cells were shifted to 328C, a temperature at which p53Ala143 acquires the wild-type con®guration. An increase in TRID mRNA levels was noted within 6 h of temperature shift to permissive temperature (328C); levels were maximal by 24 h and declined thereafter (Figure 2b). Vector-transfected cells, by contrast, did not reveal such alterations in TRID mRNA levels (Figure 2b), thereby con®rming that the eects were speci®c to p53Ala143 and not due to changes in temperature. The comparison of the kinetics of TRID and p21WAF1 induction by p53 (compare Figure 2a lower panel with Figure 2b) demonstrates that both TRID and p21WAF1 are induced by p53 as early as 6 h. Endogenous TRID expression was also induced following adenovirus vector-mediated transfer of exogenous wild-type p53 in two dierent cells lines (El-Deiry and colleagues, manuscript submitted for publication). Cumulatively, the above results demonstrate that TRID expression is regulated by p53. It is not clear why TRID mRNA levels decline at 48 h when p53 is still in its wild-type conformation. Recently several p53 inducible genes (PIGs) have been identi®ed by Polyak et al. (1997). Those PIGs also exhibited variable kinetics of induction by p53. For example, p53-mediated induction of PIG1 was noted as early as 12 h and the maximal induction was achieved by 48 h. p53-dependent PIG5 induction, on the other hand, was noted by 3 h, the maximal induction was achieved by 16 h and the levels declined by 48 h when wild-type p53 was still expressed at higher levels. It is


4155

c

a

b

Figure 1 The eect of genotoxic agents on TRID gene expression in human cancer cell lines. (a) The eect of ionizing radiation (IR) on the mRNA levels of TRID in p53-negative and p53 wild-type cells. Logarithmically growing cells were exposed to 20 Gy IR and harvested 4 h following IR exposure. RNA extraction and quantitative dot blot or Northern hybridizations were performed as described previously (Sheikh et al., 1998). Values are means+s.e.m. of two to three experiments and expressed as relative to untreated respective controls all of which were given a value of 1. TRID expression was also induced by IR in the p53 wild-type ML-1 cell but these cells exhibited considerable variations in the fold-induction of TRID mRNA levels. (b) The eect of MMS and UV-irradiation on the TRID mRNA levels in p53-negative and p53 wild-type cells. Logarithmically growing cells were treated with MMS (100 mg/ml) or exposed to UV (14 J/m2) and harvested 4 h later. Values are means+s.e.m. of two to three independent experiments and expressed as relative to untreated respective controls all of which were given a value of 1. (c) A representative Northern blot showing the eect of MMS, IR and UV-irradiation on the TRID mRNA expression in p53 wild-type A549 human lung carcinoma cells. Cells were either left untreated or treated with indicated agents and Northern hybridizations were performed as described previously (Sheikh et al., 1998). Ethidium bromide staining of the same gel shows RNA integrity and comparable loading in each lane


4156

possible that dierences in the mRNA stability may account for the dierential kinetics of p53-dependent induction of these genes. We next investigated the TRID gene expression in primary tumors and their matching normal tissues. Total RNAs from matching tumor and normal tissue samples obtained from six colon carcinoma patients, one gastric carcinoma and seven esophageal carcinoma patients were subjected to Northern blot hybridization. The overall results presented in Table 1, demonstrate

that TRID is expressed at higher levels in tumors of the gastrointestinal tract (GIT). Four of six colon (67%), one of one gastric (100%), and ®ve of seven esophageal (71%) tumor specimens exhibited higher TRID expression levels when compared to their matching normal tissue samples. The normal tissue samples either did not express TRID message or exhibited very weak signals. Representative results illustrating TRID mRNA expression in tumors versus matching normal tissues are shown in Figure 3a. Thus,

a b

Figure 2 (a) Representative photomicrographs show immuno¯uorescent detection of exogenous p53Ala143 and endogenous p21Waf1 in p53null H1299 cells. Vector-transfected (Neo) and p53Ala143-transfected H1299 cells were grown at 328C for 48 h and subjected to indirect immuno¯uorescent staining using anti-p53 and anti-p21Waf1 antibodies (Sheikh et al., 1996). Lower panel shows a representative Northern blot performed on p53Ala143-transfected cells grown at 37 or 328C for indicated periods of time. (b) The eect of exogenous temperaturesensitive human p53Ala143 on the endogenous TRID gene expression. p53-null H1299 cells stably transfected with exogenous p53Ala143 were either kept at 37 or 328C for indicated periods of time. The vector-transfected H1299 cells were also similarly subjected to alterations in temperature. RNA extraction and Northern blot hybridizations were performed as described previously (Sheikh et al., 1998). Upper panel illustrates the representative Northern blot whereas the lower panel shows the bar graphs depicting the induction of TRID in response to exogenous p53. The values shown in bar graphs are expressed as fold induction in TRID mRNA levels at 328C relative to the values at 378C. The plotted data is the mean+s.e.m. of three independent experiments


TRID expression is enhanced in primary tumors of the GIT. Northern analyses performed on RNAs obtained from unmatched normal stomach and gastric tumor tissues also revealed higher TRID expression in three of ®ve gastric tumors (60%) (data not shown). Given that normal colon tissue samples either did not display or exhibited very low TRID mRNA expression (Figure 3b), the readily detectable levels of TRID mRNA in six of six colon cancer cell lines (Figure 3b) are also in line with these results. Depending on cell and tissue type, TRID gene exhibits either multiple transcripts or one major transcript of approximately 4 kb. The multiple TRID transcripts have been reported to result from alterations in the 3-untranslated region (Pan et al., 1997). Southern blot analyses performed on genomic DNAs obtained from twenty primary esophageal and colon tumor samples and their matching normal tissues did not reveal any gross rearrangements or gene ampli®cation. The distinct TRID speci®c band pattern is typical of a single copy gene (Figure 3c). Although, the tissue samples for RNA and DNA analyses were obtained from dierent patients, our Southern blot results would argue that the increase in TRID mRNA levels noted in primary tumors is due to altered gene expression and not due to structural aberrations and/or gene ampli®cation. In this study, we have investigated two issues: (i) TRID gene regulation by genotoxic stress and wildtype p53; (ii) TRID expression in primary tumors of the GIT and their matching normal tissue samples. We have demonstrated that expression of the antagonist decoy receptor TRID is induced by genotoxic agents IR and MMS in predominantly p53 wild-type cells, whereas UV-irradiation does not induce TRID gene expression. Consistent with these results, exogenous p53 also upregulates the expression of endogenous TRID. Whether TRID is a direct target of p53 or is post-transcriptionally regulated by p53 is an issue that remains unclear at the present time. IR induction of p53-regulated genes frequently occurs in p53 wild-type cells (Ko and Prives, 1996 and references therein). The fact that TRID was also regulated by IR only in p53

wild-type cells suggests that p53 appears to directly regulate the expression of TRID. p53 has recently been shown to regulate death receptor DR5 gene expression (Wu et al., 1997). We have also recently demonstrated that IR predominantly induced DR5 gene expression in p53 wild-type cells, whereas MMS induction of DR5 was noted in both p53 wild-type and mutant cells; UVirradiation, on the other hand, did not induce DR5 expression (Sheikh et al., 1998). TRID and DR5 thus appear to exhibit similarities and dierences in their induction in response to genotoxic stress. It is interesting that UV-irradiation which is known to activate p53 does not induce TRID expression. Given that post-translational modi®cations and alterations in the p53-binding proteins may also modulate p53 function (reviewed by Giaccia and Kastan, 1998), it is possible that UV-irradiation may also activate other as of yet unknown inhibitory factors that selectively down modulate the expression of TRID and DR5. Further studies are needed and are in progress to address this issue. TRID lacks the typical death domain and has been shown to act as an antiapoptotic molecule (Pan et al., 1997; Sheridan et al., 1997; MacFarlane et al., 1997). p53 regulation of antiapoptotic or growth-promoting genes has been well-documented. For example, p53 has been shown to upregulate EGFR (Sheikh et al., 1997), TGFa (Shin et al., 1995), cyclin G (Okamoto and Beach, 1994) and MDM2 (Barak et al., 1993) gene expression. Antiapoptotic BclXL has also been shown to be induced by p53 (Zhan et al., 1996). p53 is known to induce the expression of DR5 (Wu et al., 1997) and may mediate its apoptotic eects at least in part via DR5-mediated engagement of caspase cascade. TRID is believed to blunt the eects of TRAIL-induced apoptosis by competing with DR4 and DR5 for binding to TRAIL (Pan et al., 1997; Sheridan et al., 1997; MacFarlane et al., 1997). It is, therefore, possible that p53, by upregulating the levels of TRID, tapers the apoptotic eects of similarly regulated DR5 and thereby blunts its own apoptotic eects. In this context, it is important to note that the temperature-sensitive

Table 1 TRID expression in primary tumors of gastrointestinal tract Patient # 1 2 3 4 5 6 7 8 9 10 11 12 13 14 a

Age (years) 33 56 83 59 22 34 ± 80 77 63 ± 48 57 68

Sex

Primary site

Tumor type

TRID tumor signala

Male Male Male Male Male Male ± Male Male Male ± Male Male Female

Colon Colon Colon Colon Colon Colon Stomach Esophagus Esophagus Esophagus Esophagus Esophagus Esophagus Esophagus

Dierentiated Adenocarcinoma Poorly Dierentiated Adenocarcinoma Adenocarcinoma Adenocarcinoma Adenocarcinoma Adenocarcinoma Gastric Carcinoma Sq. Carcinoma Sq. Carcinoma Sq. Carcinoma Sq. Carcinoma Sq. Carcinoma Sq. Carcinoma Sq. Carcinoma

Increased Increased None Detected None Detected Increased Increased (7.5) Increased (4.5) Increased (3.2) Increased (8.3) Increased (6.3) Increased Increased No Dierence No Dierence

Relative to signal in matching normal tissue. The normal tissues either did not express or exhibited very low levels of TRID message. The fold increase in TRID mRNA levels (values in parentheses) is given for those tumors whose matching normal tissues exhibited detectable TRID mRNA signal. Sq. Carcinoma, Squamous Cell Carcinoma. Esophageal, gastric and colorectal tumors and homologous normal tissues were obtained from surgical resections at the Greenbaum Cancer Center at University of Maryland, Baltimore, Maryland, USA. Tumor was grossly dissected free of surrounding normal tissue and the specimens were frozen and stored in liquid nitrogen. Northern blots carrying the RNA from matching normal and tumor tissues of colorectal and eosphageal cancer patients were also obtained from Invitorgen (San Diego, CA, USA). The unmatched normal and gastric tumor samples were obtained at the surgical department of the University of Bern (Bern, Switzerland) as previously described (Guo et al., 1998)

4157


4158

mutant p53 used in this study did not induce apoptosis in its wild-type conformation but did induce growth arrest (data not shown). The p53-induction of TRID was not the consequence of p53-induced growth arrest, since the TRID induction occurred within 6 h of temperature shift to permissive temperature (when p53 acquires a wild-type conformation), while growth arrest was not noted until 48 ± 72 h at the permissive temperature (data not shown). Thus, based on our ®ndings, we propose that TRID appears to be a bona ®de target of p53. We have also demonstrated for the ®rst time that TRID is overexpressed in primary tumors of the GIT. In the earlier studies, TRID was found to be

overexpressed in normal tissues, whereas its expression was relatively low in cancer cell lines (Pan et al., 1997; Sheridan et al., 1997; MacFarlane et al., 1997). In those studies, however, the expression pattern of TRID in eight cancer cell lines representing dierent tumor types was compared with its expression in completely unmatched normal tissues. In our study, we have used primary tumors of the GIT and their matching normal tissues to demonstrate that TRID is overexpressed in tumors of the GIT. The p53-status of primary tumors analysed in this study remains unknown. The consequence of p53 inactivation on the constitutive expression of p53 regulated genes during malignant progression remains less well investigated. It is well

a

b

c

Figure 3 (a) TRID mRNA expression in primary colon, gastric and esophageal tumors and their matching normal tissues. (b) TRID mRNA expression in normal human adult tissues (upper panel) and human colon and breast cancer cell lines (lower panel). RNA extraction and Northern blot hybridizations were performed as described previously (Sheikh et al., 1998). (c) Representative Southern blots showing TRID gene structure in esophageal tumors and their matching normal tissues. DNA extraction and Southern blot hybridization were performed as previously described (Meltzer et al., 1991). In brief, genomic DNA was digested to completion with EcoRI restriction enzyme and electrophoresed onto a 0.8% agarose gel and transferred to a supported Nylon membrane (Schleicher and Schuell, Keene, NH, USA). Blots were hybridized with 32P-labeled TRID cDNA probe in a hybridization solution containing 50% formamide at 508C overnight. Blots were washed twice for 10 min in 26SSPE and 0.2% SDS at 508C and twice for 10 min each in 0.26SSPE and 0.2% SDS at 608C. Blots were also probed with b-actin and GAPDHcDNA probes to con®rm dierences in the amount of DNA present in each lane


established that a majority of the p53-regulated genes are also regulated in a p53-independent manner; thus the constitutive mRNA levels of p53-regulated genes may or may not correlate with the wild-type or mutant status of p53. Future studies will be directed to investigate a correlation between the p53-status and the constitutive TRID expression levels in primary tumors. The process of oncogenesis is believed to be aected by the dierences in rates of proliferation and apoptosis. TRID is an antagonist decoy receptor that protects from TRAIL-induced apoptosis (Pan et al., 1997; Sheridan et al., 1997; MacFarlane et al., 1997). Our present ®ndings demonstrating the overexpression of TRID by tumors of the GIT does not appear paradoxical. It is likely that GIT tumors may gain a

growth advantage by overexpressing TRID in order to protect themselves against TRAIL-mediated apoptosis. In future studies a larger pool of samples representing a variety of dierent tumor types and their matching normal tissues will be analysed to ®rmly establish the status of TRID as an important tumor marker and to also investigate the expression pattern of the other decoy receptor TRUNDD. Acknowledgements Technical assistance of Kia S Brooks is gratefully acknowledged. SJ Meltzer was supported by NIH grants CA78843, CA77057, DK47717, CA67497 and the oce of Medical Research, Department of Veterans Aairs.

References Alnemri ES. (1997). J. Cell. Biochem., 64, 33 ± 42. Barak Y, Juven T, Haner R and Oren M. (1993). EMBO J., 12, 461 ± 468. Cleveland JL and Ihle JN. (1995). Cell, 81, 479 ± 482. Friedlander P, Legros Y, Soussi T and Prives C. (1996). J. Biol. Chem., 271, 25468 ± 25478. Giaccia AJ and Kastan MB. (1998). Gen. Dev., 12, 2973 ± 2983. Guo X-Z, Friess H, Maurer C, Berberat P, Tang W-H, Zimmermann A, Naef M, Graber HU, Korc M and Buchler MW. (1998). Cancer Res., 58, 753 ± 758. Issac JT. (1994). Curr. Opin. Oncol., 6, 82 ± 89. Ko LJ and Prives C. (1996). Gen. Dev., 10, 1054 ± 1072. MacFarlane M, Ahmad M, Srinivasula SM, FernandesAlnemri T, Cohen GM and Alnemri E. (1997). J. Biol. Chem., 272, 25417 ± 25420. Marsters SA, Sheridan JP, Pitti RM, Huang A, Skubatach M, Baldwin D, Yuan J, Gurney A, Goddard AD, Godowski P and Ashkenazi A. (1997). Curr. Biol., 7, 1003 ± 1006. Meltzer SJ, Yin J, Huang Y, McDaniel TK, Newkrik C, Iseri O, Vogelstein B and Resau JH. (1991). Proc. Natl. Acad. Sci. USA, 88, 4996 ± 4980. Nagata S. (1997). Cell, 88, 355 ± 365. Okamoto K and Beach D. (1994). EMBO J., 13, 4816 ± 4822. Pan G, Ni J, Wei Y-F, Yu G-L, Gentz R and Dixit VM. (1997). Science (Washington DC), 277, 815 ± 818.

Pan G, Ni J, Yu G-L, Wei Y-F and Dixit VM. (1998). FEBS Lett., 424, 41 ± 45. Polyak K, Kia Y, Zweier JL, Kinzler KW and Vogelstein B. (1997). Nature, 18, 300 ± 305. Salvesen GS and Dixit VM. (1997). Cell, 91, 443 ± 446. Sheikh MS, Garcia M, Zhan Q, Liu Y and Fornace Jr AJ. (1996). Cell Growth Dier., 7, 1599 ± 1607. Sheikh MS, Carrier F, Johnson AC, Ogdon SE and Fornace Jr AJ. (1997). Oncogene, 15, 1095 ± 1101. Sheikh MS, Burns TF, Huang Y, Wu GS, Amundson S, Brooks KS, Fornace Jr AJ and El-Deiry WS. (1998). Cancer Res., 58, 1593 ± 1598. Sheridan JP, Marsters SA, Pitti RM, Gurney A, Skubatch M, Baldwin D, Ramakrishnan L, Gray CL, Baker K, Wood WI, Goddard AD, Godowski P and Ashkenazi A. (1997). Science (Washington DC), 277, 818 ± 821. Shin TH, Paterson AJ and Kudlow JE. (1995). Mol. Cell. Biol., 15, 4694 ± 4701. Wu GS, Burns TF McDonald III ER, Jiang W, Meng R, Krantz ID, Kao G, Gan DD, Zhou J-Y, Muschel R, Hamilton SR, Spinner NB, Markowitz S, Wu G and ElDeiry WS. (1997). Nature Gen., 17, 141 ± 143. Zhan Q, Alamo Jr I, Yu K, Boise LH, O'Connor PM and Fornace Jr AJ. (1996). Oncogene, 13, 2287 ± 2293.

4159