Portal de Periódicos da CAPES

Regulation of apoptosis and cell cycle arrest by Zac1, a novel zinc finger protein expressed in the pituitary gland and the brain

1997; Springer Nature; Volume: 16; Issue: 10 Linguagem: Inglês

10.1093/emboj/16.10.2814

1460-2075

Dietmar Spengler, Martín Villalba, Anke Hoffmann, C. Pantaloni, Souheir Houssami, Joël Bockaert, Laurent Journot,

Epigenetics and DNA Methylation

Article15 May 1997free access Regulation of apoptosis and cell cycle arrest by Zac1, a novel zinc finger protein expressed in the pituitary gland and the brain Dietmar Spengler Corresponding Author Dietmar Spengler CNRS UPR-9023, Mécanismes Moléculaires des Communications Cellulaires, CCIPE, 141 rue de la Cardonille, F-34094 Montpellier, Cedex 05, France Search for more papers by this author Martin Villalba Martin Villalba CNRS UPR-9023, Mécanismes Moléculaires des Communications Cellulaires, CCIPE, 141 rue de la Cardonille, F-34094 Montpellier, Cedex 05, France Search for more papers by this author Anke Hoffmann Anke Hoffmann Max-Planck Institute of Psychiatry, Molecular Neurobiology, Kraepelinstraße 2, D-80804 Munich, Germany Search for more papers by this author Colette Pantaloni Colette Pantaloni CNRS UPR-9023, Mécanismes Moléculaires des Communications Cellulaires, CCIPE, 141 rue de la Cardonille, F-34094 Montpellier, Cedex 05, France Search for more papers by this author Souheir Houssami Souheir Houssami CNRS UPR-9023, Mécanismes Moléculaires des Communications Cellulaires, CCIPE, 141 rue de la Cardonille, F-34094 Montpellier, Cedex 05, France Search for more papers by this author Joel Bockaert Joel Bockaert CNRS UPR-9023, Mécanismes Moléculaires des Communications Cellulaires, CCIPE, 141 rue de la Cardonille, F-34094 Montpellier, Cedex 05, France Search for more papers by this author Laurent Journot Laurent Journot CNRS UPR-9023, Mécanismes Moléculaires des Communications Cellulaires, CCIPE, 141 rue de la Cardonille, F-34094 Montpellier, Cedex 05, France Search for more papers by this author Dietmar Spengler Corresponding Author Dietmar Spengler CNRS UPR-9023, Mécanismes Moléculaires des Communications Cellulaires, CCIPE, 141 rue de la Cardonille, F-34094 Montpellier, Cedex 05, France Search for more papers by this author Martin Villalba Martin Villalba CNRS UPR-9023, Mécanismes Moléculaires des Communications Cellulaires, CCIPE, 141 rue de la Cardonille, F-34094 Montpellier, Cedex 05, France Search for more papers by this author Anke Hoffmann Anke Hoffmann Max-Planck Institute of Psychiatry, Molecular Neurobiology, Kraepelinstraße 2, D-80804 Munich, Germany Search for more papers by this author Colette Pantaloni Colette Pantaloni CNRS UPR-9023, Mécanismes Moléculaires des Communications Cellulaires, CCIPE, 141 rue de la Cardonille, F-34094 Montpellier, Cedex 05, France Search for more papers by this author Souheir Houssami Souheir Houssami CNRS UPR-9023, Mécanismes Moléculaires des Communications Cellulaires, CCIPE, 141 rue de la Cardonille, F-34094 Montpellier, Cedex 05, France Search for more papers by this author Joel Bockaert Joel Bockaert CNRS UPR-9023, Mécanismes Moléculaires des Communications Cellulaires, CCIPE, 141 rue de la Cardonille, F-34094 Montpellier, Cedex 05, France Search for more papers by this author Laurent Journot Laurent Journot CNRS UPR-9023, Mécanismes Moléculaires des Communications Cellulaires, CCIPE, 141 rue de la Cardonille, F-34094 Montpellier, Cedex 05, France Search for more papers by this author Author Information Dietmar Spengler 2, Martin Villalba3, Anke Hoffmann1, Colette Pantaloni3, Souheir Houssami3, Joel Bockaert3 and Laurent Journot3 1Max-Planck Institute of Psychiatry, Molecular Neurobiology, Kraepelinstraße 2, D-80804 Munich, Germany 2CNRS UPR-9023, Mécanismes Moléculaires des Communications Cellulaires, CCIPE, 141 rue de la Cardonille, F-34094 Montpellier, Cedex 05, France 3CNRS UPR-9023, Mécanismes Moléculaires des Communications Cellulaires, CCIPE, 141 rue de la Cardonille, F-34094 Montpellier, Cedex 05, France The EMBO Journal (1997)16:2814-2825https://doi.org/10.1093/emboj/16.10.2814 PDFDownload PDF of article text and main figures. ToolsAdd to favoritesDownload CitationsTrack CitationsPermissions ShareFacebookTwitterLinked InMendeleyWechatReddit Figures & Info The proliferation rate of a cell population reflects a balance between cell division, cell cycle arrest, differentiation and apoptosis. The regulation of these processes is central to development and tissue homeostasis, whereas dysregulation may lead to overt pathological outcomes, notably cancer and neurodegenerative disorders. We report here the cloning of a novel zinc finger protein which regulates apoptosis and cell cycle arrest and was accordingly named Zac1. In vitro Zac1 inhibited proliferation of tumor cells, as evidenced by measuring colony formation, growth rate and cloning in soft agar. In vivo Zac1 abrogated tumor formation in nude mice. The antiproliferative activity of Zac1 was due to induction of extensive apoptosis and of G1 arrest, which proceeded independently of retinoblastoma protein and of regulation of p21WAF1/Cip1, p27Kip1, p57Kip2 and p16INK4a expression. Zac1-mediated apoptosis was unrelated to cell cycle phase and G1 arrest was independent of apoptosis, indicating separate control of apoptosis and cell cycle arrest. Zac1 is thus the first gene besides p53 which concurrently induces apoptosis and cell cycle arrest. Introduction The proliferation of a cell population is regulated by a balance between cell division, growth arrest, differentiation and programed cell death. A network of genes, including cell cycle regulatory genes (Hunter and Pines, 1994; Grana and Reddy, 1995), protooncogenes (Hoffman and Liebermann, 1995) and tumor suppressor genes (Levine, 1993) have emerged which play major roles in normal physiological processes, such as development (Raff et al., 1993; Steller, 1995) and aging (Jazwinski, 1996), as well as various pathological states (Thompson, 1995), such as neoplasia (Hartwell and Kastan, 1994; Karp and Broder, 1995) and neurodegenerative disorders (Heintz, 1993; Ross, 1996). All eukaryotic cells possess similar mechanisms to regulate progression of the cell cycle by sequential formation, activation and subsequent inactivation of a series of cyclin–cyclin-dependent kinase (Cdk) complexes (Grana and Reddy, 1995). Control of the protein kinase activity of these complexes is critical to the orchestrated development of multicellular organisms, as well as to the response and adaptation to various physiological or pathological stimuli in mature organisms. In addition to positive regulation by active cyclin–Cdk complexes, negative regulation of the cell cycle occurs at checkpoints, which are the transitions where feedback mechanisms operate to prevent premature entry of the cell into the next phase of the cycle prior to completion of the necessary macromolecular events. Growth suppressor genes play an important role in checkpoint function and loss or mutation of genes associated with checkpoint functions seem to have important implications in the development of cancer (Hartwell and Kastan, 1994). Apoptosis, or programed cell death, is the process by which a cell will actively commit suicide under tightly controlled circumstances (Steller, 1995). Apoptosis is a morphologically distinct form of programmed cell death that plays a major role during development and in many pathological states, including cancer (Fisher, 1994; Karp and Broder, 1995), acquired immunodeficiency syndrome (Gougeon and Montagnier, 1993) and neurodegenerative disorders (Heintz, 1993; Ross, 1996). Apoptosis occurs through the activation of a cell-intrinsic suicide program. The basic machinery to carry out apoptosis appears to be present in essentially all mammalian cells at all times, but activation of the suicide program is under the control of a network of interrelated signals that originate both from the intracellular and the extracellular milieu. Therefore, genetic lesions in genes controlling the cell cycle and/or the apoptotic machinery are pivotal to the development of malignancy or neurodegenerative disorders. Among these genes, the tumor supressor gene p53 deserves particular attention, due to its frequent loss of function in various human cancers (Hollstein et al., 1991) and its impact on tumor progression, prognosis and treatment regimens (Lowe et al., 1993, 1994). The tumor suppressor activity of p53 resides in the so far unique ability to control concurrently two fundamental cellular mechanisms, namely cell cycle progression and induction of apoptosis (Bates and Vousden, 1996; Ko and Prives, 1996). We report here cloning of a novel gene, designated Zac1, which encodes a protein with seven zinc fingers of the C2H2 type which is only distantly related to previously characterized zinc finger proteins. Unexpectedly, we isolated Zac1 by a functional expression cloning technique, which resulted in addition in isolation of the wild-type form of p53. This technique is based on co-transfection of pools of an expression library with a cAMP-responsive reporter gene (Spengler et al., 1993). Zac1 and p53 were found to induce expression of the gene encoding the type I receptor (PACAP1-R) for the peptide PACAP (pituitary adenylate cyclase activating polypeptide) (Miyata et al., 1989) through mechanisms which remain not totally elucidated at present. Of note, though Zac1 and p53 are structurally unrelated, they caused an equivalent inhibition of tumor cell growth in vitro under constitutive and regulated expression through apparently different pathways. Interestingly again, Zac1-dependent growth inhibition relied on separate induction of apoptosis and G1 arrest. Results Functional expression cloning of Zac1 We used a recently described expression cloning method (Spengler et al., 1993) to screen simultaneously for different receptors positively coupled to adenylyl cyclase. This method is based on transcriptional induction of a cAMP-responsive luciferase reporter gene by stimulation of adenylyl cyclase through activated target receptors. Pools of clones from a corticotrophic tumor cell line (AtT-20) cDNA library and from a newborn rat colliculus cDNA library were co-transfected with the reporter construct pΔMC16LUC into the host cell line LLC-PK1. Separate aliquots of cells were incubated with various peptides, including PACAP, 12 h after electroporation. One pool of clones from the rat colliculus library consistently stimulated luciferase activity in the presence of PACAP, and a functional clone encoding PACAP1-R was isolated by successive subdivisions (Spengler et al., 1993). Several other pools displayed the same phenotype, namely PACAP-dependent stimulation of the reporter gene (data not shown), and the corresponding active clones were isolated by the same subdivision process. Sequencing of one of these PACAP-positive clones from the rat library revealed that this cDNA was identical to the wild-type form of the tumor suppressor gene p53. Partial sequencing of two PACAP-positive clones from the AtT-20 library also identified them as wild-type mouse p53. In addition, two clones (p2195, EMBL accession No. X95503, and p1270) inducing PACAP1-R expression turned out to encode the same protein, which we named Zac1 for reasons detailed in the following sections of this manuscript. Structural analysis and tissue distribution of Zac1 The isolated cDNA clones, p2195 and p1270, contained a 2.8 and a 4.7 kb insert respectively. Entire sequencing of clone p2195 revealed a 2790 bp cDNA encoding an open reading frame of 667 amino acids giving rise to a protein with a predicted molecular weight of 75 kDa (Figure 1A). The ATG of AGGCCATGG was assigned as initiation codon on the basis of its close match to the CC(A/G)CCATGG Kozak consensus sequence for favored initiation of translation and the presence of an in-frame TGA stop codon 12 nt upstream (data not shown). Examination of the protein sequence revealed the presence of seven zinc finger motifs of the C2H2 type (Klug and Schwabe, 1995) in the N-terminal region of Zac1. However, homology to other members of the zinc finger protein family was low (30% at best), with the closest group being the GLI-Krüppel family, whose members have been implicated in normal development and tumor formation (Ruppert et al., 1993). In particular, the first H/C link (HSRERPFKC) is in good agreement with the consensus motif for the GLI-Krüppel family (H(S/T)GEKP(F/Y)XC) (Schuh et al., 1986). On the other hand, the remaining 459 C-terminal amino acids displayed no significant homology to sequences in the Swissprot and NBRF-PIR databases. The central region of the protein (275–383) is characterized by 34 PLE, PMQ or PML repeats, suggestive of a structure known as a poly(proline) type II helix, which is considered to be critically involved in protein–protein interactions (Williamson, 1994). The C-terminal region is particularly P, Q and E rich, a feature often displayed by transactivation domains of transcription factors. In addition, the presence of a putative phosphorylation site (HSPQK) for Cdks located between the second and third zinc finger motifs (residues 56–60), as well as a putative protein kinase A phosphorylation site (KKWT) at the very C-terminus (residues 663–666), suggests possible regulation by protein kinases. Figure 1.(A) Sequence of Zac1 protein. Cysteine and histidine residues of the seven zinc finger motifs of the C2H2 type are boxed. A putative phosphorylation site for Cdks, corresponding to the consensus motif (b/p)(S/T)Pxb located at residues 56–60, is underlined. A putative phosphorylation site for protein kinase A (PKA) at residue 666 is indicated (*). (B) Schematic representation of the Zac1 clones. Clones p2195 and p1270 were derived from the AtT-20 corticotrophic tumor cell line. Clone B-16 was isolated from a BALB/c pituitary library and encodes the same protein identified in p2195 and p1270. The coding region of p1270 and B-16 is interrupted at residue 658 by a 630 bp insertion. The sequences at the boundaries of this insertion are displayed in the lower part of the figure, are in excellent agreement with consensus exon–intron junctions and preserve the reading frame. Restriction sites for EcoRI (R), BamHI (B) and NotI (N) are indicated. Download figure Download PowerPoint Since the cDNAs p2195/p1270 were isolated from the AtT-20 tumor cell line, there was a potential risk that they harbored mutations which may have resulted in loss or gain of functions not associated with the wild-type form. To rule out this possibility, we recloned Zac1 from a plasmid library constructed from whole pituitary tissue from BALB/c mice. Screening of ∼0.5×106 clones with the p2195 cDNA probe allowed isolation of one full-length cDNA clone designated B-16 (EMBL accession No. X95504), which contained a 3.7 kb insert. Transfection of B-16 into LLC-PK1 cells successfully substituted for p2195, p1270 or wild-type p53 with respect to regulation of PACAP1-R expression (data not shown). Entire sequencing of clone B-16 showed an 86 bp 5′-untranslated region and an extended 3′-untranslated region of 0.7 kb (Figure 1B). The coding region of B-16 was identical to p2195 except that the reading frame was interrupted at residue 658 by a 630 bp insertion. The sequences at the boundaries of this insertion are in excellent agreement with consensus exon–intron junction sequences and preserve the reading frame (Figure 1B). We observed this insertion at exactly the same position in clone p1270 derived from the AtT-20 library (Figure 1B). This finding argues against a cloning artefact in clone B-16 and suggests the presence of an unspliced intron region. In support of this hypothesis, a PCR fragment corresponding to the intron region failed to hybridize to a poly(A)+ blot from AtT-20 cells (data not shown). The distribution of Zac1 was assessed by Northern blot of total RNA prepared from different mouse tissues. Interestingly, in adult mice the anterior pituitary gland displayed by far the highest level of expression of Zac1 mRNA (Figure 2). Zac1 was expressed at lower levels in various brain areas, including olfactory bulb, cortex, hippocampus, hypothalamus–thalamus, brain stem and cerebellum, and faintly in peripheral tissues such as stomach, kidney, adrenal gland, heart and lung (Figure 2 and data not shown). Figure 2.Distribution of Zac1 mRNA in mouse tissues. Zac1 distribution was assessed by Northern blot analysis of total RNA prepared from different brain regions [olfactory bulb (Olf), frontal cortex (fCx), occipital cortex (oCx), hippocampus (Hip), hypothalamus–thalamus (HyT), brain stem (BSt), cerebellum (Crb)] and peripheral tissues [anterior pituitary gland (Pit), heart (Hea), liver (Liv), stomach (Sto), intestine (Int), kidney (Kid), adrenal gland (Adr), spleen (Spl), lung (Lun)]. Ethidium bromide staining of the gel is shown in the insert to document equal and intact amounts of each RNA preparation. Download figure Download PowerPoint Constitutive expression of Zac1 and p53 abates growth of tumor cells To test whether Zac1 shares with p53 additional properties besides regulation of the PACAP1-R gene, we examined cell proliferation in a colony formation assay (Baker et al., 1990; Diller et al., 1990). We transfected Zac1 and p53 cDNAs in sense and antisense orientation in a vector carrying the puromycin resistance gene and selected for puromycin-resistant clones in LLC-PK1 and SaOs-2 cells. The Zac1 sense cDNA caused a substantial reduction in the number of colonies, comparable with p53, whereas similar numbers were noted for the parent vector and the antisense constructs (Figure 3). Abrogation of cell growth by Zac1 and p53 was more prominent in the SaOs-2 cell line. In addition, the clones that did appear after transfection of Zac1 or p53 sense constructs into the LLC-PK1 cell line died when re-exposed to selection after passaging and grew at a slow rate if further selection was omitted (data not shown). In contrast, expression of the human glucocorticoid receptor, which encodes a strong transactivation domain and profoundly alters cell homeostasis, revealed a <10-fold difference in the number of colonies in the absence and presence of dexamethasone, ruling out the possibility that differences under Zac1 and p53 are solely due to protein overexpression or squelching (data not shown). Figure 3.Zac1 and p53 inhibit colony formation. LLC-PK1 (A) or SaOs-2 (B) cells were transfected with plasmids encoding puromycin resistance alone (pCMVPUR) or containing Zac1 or p53 cDNAs in sense or antisense orientation. Puromycin was added for 10 days and viable colonies were scored following staining with MTT. Download figure Download PowerPoint Regulated expression of Zac1 and p53 impairs proliferation rate, colony formation in soft agar and tumor formation in nude mice Zac1 and p53 suppress growth of tumor cells. For further analysis, we used a tetracycline-regulated expression system (Gossen and Bujard, 1992, 1993) to control expression by a transactivator (tTA) which is blocked in the presence of tetracycline (Tc) or anhydrotetracycline (ATc). We included p53 in the following experiments to serve as a positive control and as a reference to evaluate the responses to Zac1 expression under the specific experimental conditions used in this study. A novel tetracycline-sensitive expression vector with lower basal activity (Hoffmann et al., 1997; A.Hoffmann, M.Villalba, L.Journot and D.Spengler, submitted) was used to express Zac1 and p53 in individual LLC-PK1 and SaOs-2 cell clones. One third of the individual clones isolated from each transfection condition (L-Zac = 95, L-p53 = 92, S-Zac = 77, S-p53 = 72) revealed strong inhibition of proliferation in the absence of ATc, as deduced from light microscopic inspection. Analysis of 10 clones randomly chosen from each of these groups revealed a close correlation between growth inhibition and Zac1 or p53 expression, as shown by measurement of cell number and immunoblot analysis (data not shown) and allowed the assignment of one clone for further study. Importantly, in the presence of the repressor ATc, no major differences in growth behavior were observed between Zac1- and p53-expressing clones and the parent clones, L-tTA and S-tTA (Figure 4A and B). Therefore, the differences in cell number on day 6 were primarily due to suppression of growth in the absence of the repressor. In line with this view, results from cell counts on primary pools (n = 3) from each condition revealed a 4- and 7-fold difference in cell numbers on day 6 for both the LLC-PK1 and SaOs-2 cell lines (data not shown). A strong increase in protein levels of Zac1 was noted in the induced state (data not shown and Figure 7C). Similar results were also obtained for regulation of p53 in LLC-PK1 and SaOs-2 cells (data not shown and Figure 7C). Figure 4.Zac1 and p53 alter proliferation of LLC-PK1 and SaOs-2 cells. ATc-regulated expression of Zac1 and p53 was established in LLC-PK1 and SaOs-2 cells. (A and B) Cell numbers of the parent tTA clones (L-tTA and S-tTA) were compared with those obtained with Zac1- and p53-expressing LLC-PK1 (L-Zac and L-p53 respectively) and SaOs-2 (S-Zac and S-p53 respectively) clones in the presence (+) or absence (−) of ATc. (C and D) Zac1 and p53 inhibit DNA synthesis (BrdU) and cell viability (MTT). For each time point, BrdU incorporation or formazan blue formation were measured in the presence (+) or absence (−) of ATc. (E) Growth inhibition by Zac1 and p53 is serum independent. Cells were grown in the presence of the indicated amount of fetal bovine serum (10 or 0.1%) and in the presence (+) or absence (−) of ATc. (F) Growth inhibition by Zac1 and p53 is reversible. Cells were seeded in ATc-containing medium, grown in the absence of ATc for 2 days before medium was renewed (arrow) with medium containing (−/+) or lacking (−/−) ATc. Download figure Download PowerPoint To confirm these results obtained by direct quantification of cell number, we evaluated the effects of Zac1 and p53 expression by two complementary methods. First, we studied DNA synthesis with a non-radioactive immunoassay based on incorporation of 2–bromodeoxyuridine (BrdU) into nuclear DNA on each of 6 days with or without ATc (Figure 4C). Second, we measured conversion of the tetrazolium salt Thiazolyl blue (MTT) to formazan blue, which depends on the activity of mitochondrial and cytoplasmic dehydrogenases. This activity depends on cell viability and closely correlates with cell proliferation. These experiments confirmed the measurements of cell number (Figure 4D). Similar results were obtained for L-Zac and L-p53 (data not shown). To exclude the possibility that Zac1-induced alteration of proliferation is due to down-regulation of transduction pathways activated by mitogenic serum factors, we assessed proliferation in the presence of different serum concentrations. Cells from LLC-PK1 (data not shown) and SaOs-2 clones kept under low serum conditions (0.1% fetal calf serum) in the repressed state displayed a reduced growth rate from day 3 on, indicating serum dependence in maintaining logarithmic growth (Figure 4E). In contrast, cell number was unaffected by serum concentration with expression of Zac1 and p53 (Figure 4E). Therefore, inhibition of tumor cell growth by Zac1 and p53 proceeds through mechanisms unrelated to down-regulation of mitogenic pathways in these cellular models. The ability of Zac1 to suppress growth could be due to a non-specific lethal effect of protein overproduction resulting in cell death. Alternatively, it could be a manifestation of a more specific effect on cell proliferation. To further investigate these two possibilities, we tested the growth pattern following re-exposure to ATc of the surviving cells. Surviving LLC-PK1 and SaOs-2 cells resumed logarithmic growth after 48 h of ATc deprivation (data not shown and Figure 4F). Therefore, Zac1- and p53-induced changes in cell growth were not permanent and, at least in part, were reversible, arguing against a non-specific effect of protein overproduction. In support of this view, high levels of expression of the inert luciferase gene from the same tetracycline-dependent expression system in LLC-PK1 or SaOs-2 cells revealed no changes in cell proliferation. Zac1 and p53 inhibit soft agar colony formation. Anchorage-independent growth is often correlated with tumorigenesis and is a strong criterion for cultured cell transformation. To test the influence of Zac1 on anchorage-independent growth, we assayed LLC-PK1 and SaOs-2 cell clones for their ability to grow in soft agar. Colony formation by Zac1- or p53-expressing cells was dramatically reduced compared with the repressed state (Figure 5). Also, the few colonies formed under Zac1 or p53 expression were of smaller size. These results demonstrate that Zac1 and p53 can abate anchorage-independent growth of tumor cells, one of the hallmarks of tumorigenicity and transformed cell growth. Figure 5.Zac1 and p53 inhibit soft agar colony formation. Zac1 (L-Zac and S-Zac) and p53 (L-p53 and S-p53) clones were grown in the presence of ATc before plating into soft agar at densities of 1×105 (nos 1 and 4), 5×104 (nos 2 and 5) and 2.5×104 (nos 3 and 6) cells per well in 6-well plates. The repressor ATc was included in the upper rows (+) and was omitted in the lower rows (−). For photography on day 10, the soft agar was overlaid with MTT for 4 h. Pictures shown are representative of three to five independent experiments. Download figure Download PowerPoint Zac1 and p53 suppress tumor formation in nude mice. The most stringent experimental test of neoplastic behavior is the ability of injected cells to form tumors in nude mice. Yet not all of the altered cellular growth properties commonly associated with the transformed state in vitro are required for neoplastic growth in vivo and vice versa. Therefore, loss of tumorigenicity under expression of Zac1 in vivo would be a critical test to substantiate the growth suppressor function of Zac1. To achieve gene regulation by Tc in nude mice, we implanted half of the animals with Tc pellets, whereas the remainder was implanted with placebo pellets. Two days later, each animal was injected s.c. on each side with S-Zac or S-p53 cells grown in the continuous presence of ATc. Due to the clonal origin of S-Zac and S-p53, we observed differences in the tumorigenicity of each clone, as shown by the difference in the observed lag in tumor formation, which was assessed at 11 weeks after cell injection for S-Zac and at 16 weeks for S-p53. Table I presents results from three experiments with S-Zac and one experiment with S-p53. In agreement with previous reports (Chen et al., 1990), p53 expression impaired tumor formation by SaOs-2 cells in vivo. Interestingly, Zac1 also inhibited tumor formation, as deduced from tumor incidence (Table I) and tumor weight [193 ± 13 mg (n = 14) for Tc versus 18 ± 7 mg (n = 2) for placebo]. Thus, Zac1 and p53 are equivalent at inhibiting tumor formation in vivo in xenografted nude mice. Table 1. Zac1 and p53 inhibit tumor formation in vivo Clone Tumor incidence (No. of tumor-bearing injection sites/No. of injection sites) Placebo Tc S-Zac (exp. 1) 2/12 14/14 S-Zac (exp. 2) 1/12 12/12 S-Zac (exp. 3) 1/8 8/8 S-p53 1/12 10/12 Nude mice were implanted with placebo or Tc pellets s.c. Two days later, 5×106 cells from each clone were injected s.c. into each side of each animal and tumor formation was scored at 11 (Zac1) and 16 weeks (p53). Expression of Zac1 and p53 induces apoptosis An increasing number of cells with signs of lost cell viability was observed from day 2 onwards following Zac1 or p53 expression. These cells, which were abundant on phase contrast microscopy, failed to convert MTT, shrank, revealed membrane blebbing and further rounded up before detaching from the plates. For Zac1 these alterations were most evident in SaOs-2 cells (S-Zac) and for p53 in LLC-PK1 cells (L-p53) (data not shown) and appeared reminiscent of apoptotic cell death. This form of cell death is often accompanied by degradation of the DNA into a ladder of regular fragments. To address this issue, we isolated genomic DNA from the LLC-PK1 and SaOs-2 clones kept for 3 days without ATc. When the repressor was omitted, a clearly visible degradation into oligonucleosomal DNA fragments became evident (Figure 6A and B), which was most advanced following expression of Zac1 in SaOs-2 cells. Quantification of fragmented DNA after 48 h without ATc indicated a 17.3- and 3.9-fold increase following Zac1 expression and a 3.4- and 6.8-fold increase under p53 in SaOs-2 and LLC-PK1 cells respectively (data not shown). Figure 6.Zac1 and p53 induce apoptotic cell death. (A and B) DNA laddering. Genomic DNA was isolated from Zac1- (L-Zac and S-Zac) and p53-expressing (L-p53 and S-p53) clones grown in the presence (+) or absence (−) of ATc for 3 days, centrifuged and soluble DNA was subjected to agarose gel electrophoresis and stained with ethidium bromide. (C) Fluorescence microscopy of L-Zac1 and L-p53 displaying nuclear signs of apoptosis. Cells were grown in the absence of ATc for 3 days. Floating cells (L-Zac, upper panel; L-p53, lower panel) were collected, incubated with ethidium bromide and examined by fluorescence microscopy (510–550 nm, ×1000). (D) DNA end-labeling. S-Zac (Zac1) and S-p53 (p53) cells were grown for 3 days in the presence (green) or absence (red) of ATc. Permeabilized cells were subjected to terminal transferase nick end-labeling in the presence of digoxigenin-labeled dUTP (TUNEL). Cells were then incubated with fluorescein-conjugated anti-digoxigenin antiserum and subjected to flow cytometry. Download figure Download PowerPoint Figure 7.Zac1 and p53 regulate cell cycle progression. (A) Induction of G1 arrest by Zac1. S-Zac cells were grown in the presence (upper panel) or absence (lower pane