- Email: [email protected]
Necdin acts as a transcriptional repressor that interacts with multiple guanosine clusters
Gene 272 (2001) 173±179
www.elsevier.com/locate/gene
Necdin acts as a transcriptional repressor that interacts with multiple guanosine clusters Kuniharu Matsumoto, Hideo Taniura, Taichi Uetsuki, Kazuaki Yoshikawa* Division of Regulation of Macromolecular Functions, Institute for Protein Research, Osaka University, Yamadaoka 3-2, Suita, Osaka 565-0871, Japan Received 26 March 2000; received in revised form 15 May 2001; accepted 29 May 2001 Received by T. Sekiya
Abstract Necdin is a growth suppressor expressed predominantly in postmitotic neurons, and ectopic expression of this protein suppresses cell growth. Here we report that Necdin directly binds to speci®c DNA sequences and serves as a transcriptional repressor. Polyhistidine-tagged Necdin was used for selection of random-sequence oligonucleotides by polymerase chain reaction-based ampli®cation. Necdin recognized guanosine (G)-rich sequences that encompass multiple G clusters and intervening mono- or di-nucleotides of A, T and C. These sequences, termed GN boxes, resemble multiply aligned forms of the canonical GC box which is recognized by Sp family members. Necdin directly bound to a GN box consisting of contiguous two GC boxes with four G clusters, but not to a single GC box with two G clusters, whereas Sp1 bound to both. In a reporter system using Drosophila Schneider Line 2 cells, Necdin repressed Sp1-dependent activity of mouse c-myc P1 promoter that contains a typical GN box. Deletion of the GN box from the c-myc P1 promoter or its conversion to the single GC box abolished the Necdin-dependent repression. These results suggest that Necdin modulates gene transcription via the GN box that is potentially recognized by GC box-targeting Sp family members. q 2001 Elsevier Science B.V. All rights reserved. Keywords: DNA-binding protein; Cell growth suppressor; Cis-acting elements; Postmitotic neurons
1. Introduction Necdin is a 325 amino acid residue protein encoded in a cDNA sequence isolated from neurally differentiated murine embryonal carcinoma P19 cells (Maruyama et al., 1991). In mouse brain, the necdin gene is constitutively expressed in postmitotic neurons, whereas necdin mRNA is undetectable in neuroepithelial stem cells in the neural tube (Aizawa et al., 1992; Uetsuki et al., 1996). Necdin is also expressed in skeletal muscles at early stages of their differentiation (Taniguchi et al., 2000). These expression patterns suggest that the necdin gene is expressed in postmitotic cells that complete terminal differentiation. The human necdin gene NDN is located on chromosome 15q11.2-q12, which is deleted in the Prader±Willi syndrome (PWS) (Jay et al., 1997; MacDonald and Wevrick, 1997; Nakada et al., 1998), a neurogenetic disorder related to genomic imprinting. NDN is maternally imprinted and transcribed only from the paternal allele (Jay et al., 1997; Abbreviations: G, guanosine; His, polyhistidine; PCR, polymerase chain reaction; PWS, Prader±Willi syndrome; SL2, Schneider Line 2 * Corresponding author. Tel.: 181-6-6879-8621; fax: 181-6-6879-8623. E-mail address: [email protected] (K. Yoshikawa).
MacDonald and Wevrick, 1997; Sutcliffe et al., 1997). Necdin is not expressed in the cells prepared from PWS patients whose chromosome 15q11.2-q12 region in the paternal allele is deleted. Necdin is most abundant in postmitotic neurons in the brain stem and hypothalamus of mice (Aizawa et al., 1992; Uetsuki et al., 1996). Disruption of the mouse necdin gene results in early postnatal lethality (Gerard et al., 1999), a reduction in the number of speci®c hypothalamic neurons, and behavioral alternations (Muscatelli et al., 2000), which are reminiscent of human PWS. These observations suggest that Necdin plays a key role in differentiation and development of neurons in the brain, especially in the hypothalamus. Necdin is localized to the nucleus of postmitotic neurons (Maruyama et al., 1991; Aizawa et al., 1992). Ectopic expression of Necdin suppresses proliferation of NIH3T3 cells (Hayashi et al., 1995) and colony formation of SAOS-2 osteosarcoma cells (Taniura et al., 1998). Necdin binds to viral oncoproteins such as simian virus 40 large T antigen and adenovirus E1A (Taniura et al., 1998). Moreover, Necdin interacts with the transactivation domain of E2F1 and p53, major transcription factors that regulate the cell cycle, and suppresses their transcriptional activities (Taniura et al., 1998). These observations led us to speculate
0378-1119/01/$ - see front matter q 2001 Elsevier Science B.V. All rights reserved. PII: S 0378-111 9(01)00544-3
174
K. Matsumoto et al. / Gene 272 (2001) 173±179
that Necdin, like other cell cycle regulatory factors such as E2F1 and p53, is a transcription factor that recognizes speci®c DNA sequences. In this study we have attempted to elucidate a speci®c Necdin-binding cis-element by polymerase chain reaction (PCR)-based selection of random-sequence oligonucleotides using a recombinant Necdin protein. Necdin speci®cally binds to a G-rich motif, designated GN box, which is also recognized by Sp1. We analyzed the promoter activity using GN box-carrying c-myc P1 promoter as a model system and found that Necdin represses the Sp1-dependent promoter activity. 2. Materials and methods 2.1. Synthesis of recombinant Necdin protein in baculovirus expression system Full-length mouse necdin cDNA (Maruyama et al., 1991) was inserted into the baculovirus transfer vector pBlueBac4 (Invitrogen). For construction of a vector expressing polyhistidine (His)-tagged Necdin protein, necdin cDNA was inserted into the baculovirus transfer vector pBlueBacHis2/B (Invitrogen). The transfer vectors and Bac-NBlue AcMNPV were introduced to Sf21 insect cells to obtain recombinant baculoviruses. His-tagged Necdin expressed in virus-infected Sf9 insect cells was puri®ed with Probond metal-binding resin (Invitrogen) and dialyzed against a buffer containing 20 mM Tris±HCl (pH 8.0), 300 mM NaCl, 1 mM EDTA, 1 mM dithiothreitol, 1% Triton X100, 0.5% Tween 20 and 0.1% CHAPS. Proper folding of the Necdin protein was con®rmed by af®nity to E2F1 and p53 as described previously (Taniura et al., 1998, 1999). 2.2. DNA cellulose chromatography Sf9 cells were infected with baculovirus vectors for expression of full-length Necdin, harvested 72 h later, and lysed in a buffer containing 25 mM Tris±HCl (pH 8.0), 10 mM NaCl, 1 mM dithiothreitol, 0.1 mM EDTA, 100 mg/ml phenylmethylsulfonyl ¯uoride, 0.5 mg/ml leupeptin, 0.5 mg/ml aprotinin, 1 mg/ml pepstatin A, and 0.1% Tween 20. The cell lysate or His-tagged Necdin was mixed with 1 ml of DNA cellulose or cellulose (50% slurry) (Sigma) and incubated at 48C for 1 h. The resin was then washed with the same buffer and packed into a column. Bound proteins were eluted with a stepwise gradient of NaCl from 50 mM to 1.0 M and analyzed by Western blotting using an antibody against Necdin (C2) (Maruyama et al., 1991). 2.3. Determination of Necdin-binding sequences Selection of speci®c Necdin-binding sequences was carried out by PCR-based ampli®cation of randomsequence oligonucleotides (Pollock, 1996). A 76 base oligo-
nucleotide 5 0 -CAGGTCAGTTCAGCGGATCCTGTCG(A/ T/G/C)26GAGGCGAATTCAGTGCAACTGCAGC-3 0 was double-stranded by annealing with primer F (5 0 -GCTGCAGTTGCACTGAATTCGCCTC-3 0 ) and extending with the Ex-Taq polymerase (TAKARA, Tokyo) and dNTPs at 948C (1 min), 628C (3 min), and 728C (9 min). The product was puri®ed by phenol-chloroform extraction followed by ethanol precipitation, and dissolved in a buffer containing 10 mM Tris±HCl (pH 8.0) and 1 mM EDTA. Puri®ed Histagged Necdin was immobilized onto Ni-chelating resin, and incubated with double-stranded degenerated oligonucleotides at 48C for 30 min in a buffer (50 ml) containing 15 mM HEPES-KOH (pH 7.5), 4% Ficoll 400, 50 mM KCl, and 4 mg/ml bovine serum albumin. Protein±DNA complexes were precipitated by centrifugation and washed three times with the buffer. Bound oligonucleotides were recovered by boiling the resin and used as a template of PCR with primer R (5 0 -CAGGTCAGTTCAGCGGATCCTGTCG-3 0 ) and primer F (160 ng each) for 15±20 cycles (948C for 1 min, 628C for 1 min, 728C for 1 min). PCR was carried out in the above buffer containing 10 mg/ml poly(dIdC)(dI-dC). The PCR product (,50 ng of double-stranded DNA) was used for the successive round of selection. The oligonucleotide pool was cloned into Bluescript II SK 1 , and the sequences of 14 randomly selected clones were determined with an automatic DNA sequencer (Model 4000, LI-COR, Lincoln, NE). 2.4. Electrophoretic mobility shift assay Oligonucleotides selected at each cycle of enrichment were labeled with [a- 32P]dCTP by PCR. The PCR products were extracted from an 8% polyacrylamide gel and used as probes. Puri®ed His-tagged Necdin (80 ng protein per assay) was preincubated with 32P-labeled oligonucleotides at 48C for 15 min in a buffer (20 ml) containing 15 mM HEPESKOH (pH 7.9), 4% Ficoll 400, 4 mg/ml bovine serum albumin, 50 mM KCl, 1 mM EDTA, and 100 ng of poly(dIdC)(dI-dC). For detection of protein±GC box complexes, double-stranded oligonucleotides 5 0 -ATTCGATCGGGGCGGGGCGAGC-3 0 , 5 0 -ATTCGATCGGTTCGGGGCGAGC-3 0 , and 5 0 -ATTCGATCGGGGCGGGTGGGCGGGGCGAGC-3 0 were synthesized and used as probes of a single GC box (1 £ GC), mutated GC box, and contiguous two GC boxes (2 £ GC box), respectively. A doublestranded oligonucleotide 5 0 -GGCTAATCCCCGCCCACCCGCCCTTTATATTCCGGGGGTC-3 0 corresponding to a c-myc P1 GN box was also used. The oligonucleotide probes were labeled with [g- 32P]ATP using T4 polynucleotide kinase. Nuclear extracts were prepared from undifferentiated mouse embryonal carcinoma P19 cells (McBurney et al., 1982) and Drosophila Schneider Line 2 (SL2) transfected with Sp1 expression plasmid pPacSp1 (Schreiber et al., 1989). The nuclear extracts (2 mg protein each) and Histagged Necdin (80 ng protein) were preincubated for 10 min in a reaction mixture (20 ml) containing 15 mM HEPES-
K. Matsumoto et al. / Gene 272 (2001) 173±179
175
galactosidase (p97b) were kindly provided by Dr Guntram Suske (Erasmus University, The Netherlands). Necdin cDNA was inserted into the Drosophila expression vector pPac to make pPacNecdin. Expression of Necdin in SL2 cells was con®rmed by immunoblotting with anti-Necdin antibody C2 (Maruyama et al., 1991). Sets of plasmids were co-transfected into ,70% con¯uent SL2 cells in a 35 mm dish by the calcium phosphate method (Graham and van der Eb, 1973). Luciferase activities were measured with a luminometer (Lumat LB9501, Berthold) using a reagent kit (Toyo Ink, Tokyo, Japan). The transfection ef®ciency was normalized with co-transfected b-galactosidase reporter plasmid (p97b).
3. Results and discussion Fig. 1. DNA-binding properties of recombinant Necdin proteins synthesized in the baculovirus system. (A) DNA binding of Necdin and Histagged Necdin proteins. The lysate of Sf9 cells expressing Necdin (Necdin) or His-tagged Necdin (His-Necdin) was mixed with cellulose or DNA cellulose in a binding buffer containing 10 mM NaCl. Bound proteins were eluted with stepwise gradient of NaCl at the concentrations indicated (in M), and analyzed by Western blotting with antibody C2 against Necdin. (B) Electrophoretic mobility shift assay for Necdin-binding sequences selected by the PCR-based selection method. A 76 bp oligonucleotide containing 26 random-sequence nucleotides at the central region was used as a probe for initial PCR. Selected DNA sequences were radiolabeled, incubated with His-tagged Necdin, and analyzed by the electrophoretic mobility shift assay. Note the progressive enrichment of Necdin-binding sequences after successive rounds of selection. CX, DNA±Necdin complexes; F, free probe.
KOH (pH 7.9), 1 mM EDTA, 4% Ficoll 400, 4 mg/ml bovine serum albumin, 50 mM KCl, 6.25 mM MgCl2, 25 mM ZnCl2, and 50 mg/ml poly(dA-dT)(dA-dT). The radiolabeled probes were added to the mixture, which was incubated at 48C for 30 min. DNA-binding activities were analyzed by electrophoresis in an 8% non-denaturing polyacrylamide gel (acrylamide/bisacrylamide, 29:1) run in a buffer containing 45 mM Tris±borate (pH 8.2), 1 mM EDTA at 48C. Signals were detected by autoradiography. 2.5. Reporter assay Mouse c-myc P1 core promoter region (positions 268 to 122) was prepared by PCR using mouse genomic c-myc DNA (a gift from Dr Hisato Kondo, Osaka University) as a template. PCR products were cloned into the luciferase reporter plasmid PGV-B to make pMycP1. Necdin-binding sequence (GN box) was eliminated from the P1 promoter to make pMycP1DGN by self-ligation of the PCR product synthesized with the GN box-¯anking sequences using pMycP1 as a template. The myc GN box was converted to 1 £ GC box to make pMycP1SGC by self-ligation of PCR product synthesized with primers 5 0 -GATTAGCCAGAGAACCTCTCTTTCTCCCC-3 0 and 5 0 -ACCCGCCCTTTATATTCCGGGGGTC-3 0 using the pMycP1 template. Drosophila expression plasmids for Sp1 (pPacSp1) and b-
3.1. DNA-binding properties of the Necdin protein synthesized in the baculovirus system To test whether Necdin directly binds to DNA, we synthesized the Necdin protein in the baculovirus system. The extract of Necdin-expressing Sf9 cells was applied to DNA cellulose column chromatography. Necdin bound to DNA cellulose and was eluted with 300 mM NaCl, but it failed to interact with the cellulose matrix (Fig. 1A, upper), suggesting that Necdin expressed in Sf9 cells is competent for DNA binding, although it remained unclear whether Necdin indirectly bound to DNA via other proteins present in Sf9 cells. We then synthesized the Necdin protein as a His-tagged fusion protein which was puri®ed by Ni-chelating af®nity chromatography. Puri®ed His-tagged Necdin also bound to DNA cellulose (Fig. 1A, lower). Using this recombinant protein, we determined the speci®c Necdin-binding sequences by a repetitive PCR-assisted DNA-binding site selection. Necdin-binding sequences were radiolabeled and monitored by electrophoretic mobility shift assay at each round of selection (Fig. 1B). At the ®fth round, a distinct protein±DNA complex became apparent. Several complexes with different sizes were clearly detected at rounds 7, 8 and 9. These complexes may correspond to multiple Necdin molecules that bind to the selected oligonucleotides. The oligonucleotide pool selected at the seventh round was cloned into a plasmid, and 14 independent clones were randomly selected (Fig. 2). All of the clones contained inserts carrying highly G-rich sequences, which are heterogeneous and contain no typical consensus motifs. The heterogeneity of the selected oligonucleotides may have different af®nities to the Necdin protein oligomers, which may give rise to the multiple bands in the gel shift assay shown in Fig. 1B. Although 26 random-sequence oligonucleotides were utilized in this system, the selected sequences contained 23±26 nucleotides (mean ^ SEM 25 ^ 0.25). The statistical data of these oligonucleotides are as follows: the numbers of Gs are 19±21 (19.6 ^ 0.17) and %G values are 76±83 (78.6 ^ 0.71). Each oligonucleotide contains four
176
K. Matsumoto et al. / Gene 272 (2001) 173±179
to ®ve G clusters (4.8 ^ 0.25), and the number of Gs per cluster varies from two to ten (4 ^ 0.3). Mono- or di-nucleotides separate the G clusters (mono-nucleotides, 57; dinucleotides, 9; A 34, T 25, C 16 in 14 sequences). Since the mean number of Gs per cluster was four, large G clusters consisting of greater than or equal to seven Gs, which are present in 12 sequences, are assumed to be divided into two smaller clusters consisting of greater than or equal to three Gs by `intervening G'. Thus, a typical Necdin-binding sequence is summarized as a highly Grich sequence (%G ,80) that consists of four to ®ve clusters of contiguous ,four Gs and intervening A, T, C, and G. We termed this speci®c G-rich motif `GN boxes'. 3.2. Necdin binds to the GN box but not to the canonical GC box
Fig. 2. Necdin-binding sequences. Enriched oligonucleotides after seven rounds of selection were ampli®ed by PCR and cloned into Bluescript II. Inserts of 14 randomly selected clones were sequenced. Contiguous G clusters (n $ 2) are underlined, and intervening nucleotides A, T, and C are dotted. Statistical data are in Section 3.
The GN boxes resemble multiply aligned GC boxes, which are potentially recognized by speci®c transcription factors such as Sp family members (Philipsen and Suske, 1999). We then tested whether Necdin recognizes the canonical GC box (GGGGCGGGG) using an oligonucleotide probe in the electrophoretic mobility shift assay. Necdin failed to form a complex with the GC box oligonucleotide (Fig. 3A, lane 3). As a positive control for GC box-binding activity, we prepared the nuclear extract from undifferentiated P19 cells. GC box-speci®c complexes were formed in the nuclear extract only with the probe containing canonical GC box
Fig. 3. Necdin binds to a GN box consisting of contiguous two GC boxes. (A) Inability of Necdin to bind to the canonical GC box. Puri®ed His-tagged Necdin (Necdin) and P19 nuclear extract (P19 NE) were incubated with oligonucleotide probes containing the GC box (GC) (C) and its binding-defective mutant (M). Note that Necdin neither forms complexes with the canonical GC box (lane 3) nor competes for formation of the GC box±protein complex (lane 7). CX, DNA± protein complexes; F, free probe. (B) Complex formation between Necdin and contiguous two GC boxes. Nuclear extracts from Drosophila SL2 cells transfected with Sp1 expression vector (Sp1) and untransfected SL2 cells (SL2) were incubated with radiolabeled oligonucleotide probes, and formation of the complexes was analyzed by electrophoretic mobility shift assay. His-tagged Necdin (Necdin) binds to contiguous two GC boxes (2 £ GC) (lane 2), but not to a single GC box (1 £ GC) (lane 6). Arrows point to the complexes speci®c to Sp1 and Necdin. F, free probe.
K. Matsumoto et al. / Gene 272 (2001) 173±179
but not with the mutated probe (Fig. 3A, lanes 5 and 6). Histagged Necdin protein added to the reaction mixture failed to compete with complexes between the GC box and its binding proteins (Fig. 3A, lane 7). We then synthesized an oligonucleotide containing contiguous two GC boxes (2 £ GC, GGGGCGGGTGGGCGGGG), which ful®lls the criteria for GN boxes (%G 82, four clusters of contiguous three to four G, intervening C and T). His-tagged Necdin interacted with 2 £ GC box but not with 1 £ GC box (Fig. 3B, lanes 2 and 6). On the other hand, the nuclear extract from Sp1expressing Drosophila SL2 cells formed complexes with both 1 £ GC and 2 £ GC (Fig. 3B, lanes 3 and 7). The Necdin
177
protein failed to compete against the complex formation between Sp1 and GN box (2 £ GC) in this gel shift assay (data not shown), presumably because the GN box-binding potency of puri®ed His-tagged Necdin protein is weaker than that of the Sp1 protein produced in SL2 cells. 3.3. Necdin binds to a GN box in the c-myc P1 promoter We constructed a luciferase reporter plasmid carrying 4 £ GC boxes (in normal orientation) upstream of TATA box, and found that the basal activity of this reporter was signi®cantly reduced by cotransfected necdin cDNA in
Fig. 4. Characterization of the GN box-like motif in c-myc P1 core promoter as a Necdin-binding sequence. (A) Diagram illustrating the reporter constructs carrying mouse c-myc P1 core promoter (pMycP1) and a GN box-deleted mutant (pMycP1DGN). GN, GN box (-like motif); TATA, TATA box; Luc, luciferase reporter gene; P1, P1 transcription initiation site. pMycP1 contains a GN box-like motif (in inverted orientation), which was eliminated to make pMycP1DGN. Positions are numbered, beginning with the P1 site (11). (B) Complex formation between Necdin and c-myc GN box. A complex between Histagged Necdin (Necdin) and radiolabeled c-myc P1 GN box probe was analyzed in the presence of 100- and 200-fold molar excesses ( £ 100, £ 200) of the oligonucleotide competitors (Comp) of c-myc P1 GN box (P1GN) and GC box (GC). CX, GN box±necdin complex; F, free probe. (C) Activation of c-myc P1 promoter by Sp1 through the GN box. Drosophila SL2 cells were transfected with pMycP1 and pMycP1DGN reporter plasmids (0.8 mg each) in combination with pPacSp1 (in ng per assay). The total amount of plasmid DNA was adjusted to 3.2 mg per assay by adding empty pPac vector. Each value represents the mean 1 SD of two independent experiments.
178
K. Matsumoto et al. / Gene 272 (2001) 173±179
NIH3T3 cells (data not shown). We then searched natural promoters for GN box-like motifs to examine whether interactions between Necdin and GN box are of physiological signi®cance. We found a GN box-like motif (5 0 -GGGCGGGTGGGCGGGG-3 0 in inverted orientation) adjacent to the TATA box of the c-myc P1 promoter (Fig. 4A). This motif represents a typical GN box (%G 81, four clusters of contiguous three to four G, intervening C and T). The gel shift assay revealed that Necdin bound to this proximal promoter region at positions 268 to 122 (Fig. 4B). The shifted signal was markedly reduced by addition of the GN box sequence in 200-fold molar excess (Fig. 4B, lanes 3 and 4), but the same molar excess of the GC box sequence, which was used as a negative control, showed only a slight reduction (Fig. 4B, lanes 5 and 6), suggesting that the length of the G cluster in the single GC box is insuf®cient to compete with P1 GN box. We then examined whether the myc P1 core promoter shows responsiveness to Sp1 through the GN box. We were unable to demonstrate speci®c repression by Necdin in Sp1-dependent transactivation in several mammalian cell lines which normally contain high levels of endogenous Sp family proteins (data not shown). We alternatively used Drosophila SL2 cells, which contain no Sp family proteins and are commonly used for studies on Sp-related transcription factors (Majello et al., 1995). c-Myc core promoter was strongly activated by Sp1 in SL2 cells (Fig. 4C). The promoter defective in the GN box (pMycP1DGN) showed much less responsiveness to Sp1, suggesting that the GN box mediates the Sp1-dependent activation of the c-myc P1 promoter.
the consensus sequence CCCTCCCC known as the CT element (DesJardins and Hay, 1993) between 2117 and 2138 upstream of the P1 initiation site (Majello et al., 1995). We noted that this region represents a typical GN box motif (5 0 -GGGGAGGGTGGGGAGGGTGGGG-3 0 in inverted orientation) (%G 82, G 18, ®ve G clusters). A GN box-like motif is also found in the promoter region of the human telomerase catalytic subunit hTERT (5 0 -GGGAGGGGTCGGGACGGGGCGGGG-3 0 between 2144 and 2167 in inverted orientation) (%G 78, G 18, ®ve G clusters) (Cong et al., 1999). We infer that Necdin potentially modulates the transcription of these genes via the GN boxes in their promoters. In this study, we used the Sp1 system as a model to elucidate the physiological signi®cance of interactions between Necdin and the GN box. Sp1 may be one of many transcription factors acting on the GN box. For example, Sp makes a family of at least ®ve members Sp1±Sp5. Some of them are thought to act as negative regulators. Furthermore, it is currently known that Sp-related transcrip-
3.4. Necdin represses Sp1-dependent activities of a GN boxcontaining natural promoter Necdin repressed the Sp1-dependent transactivation of the myc P1 core promoter in a dose-dependent manner (Fig. 5A). Necdin slightly activated the GN box-containing promoter in the absence of Sp1, suggesting that Necdin acts as a partial transcriptional activator for GN box-carrying promoter. In Drosophila SL2 cells, the higher concentration of Necdin was required to compete with Sp1. Subtraction of basal activities (Sp12) from Sp1-activated values revealed that Necdin maximally suppresses the Sp1-dependent activity below the control value (DSp1^). A mutant defective in the GN box (pMycP1DGN) was slightly activated by Sp1 but showed no response to Necdin. Another mutant carrying the GC box in place of the GN box (pMcyP1SGC) was moderately activated by Sp1, but not repressed by Necdin (Fig. 5B). Necdin exerted no effects on the basal activities (Sp12) of pMycP1DGN and pMcyP1SGC (data not shown). These results indicate that Necdin represses Sp1-dependent myc P1 promoter activity via its GN box, and that the repression of Sp1-dependent transactivation is observed only when the Sp1-binding sequence overlaps with the GN box. Human c-myc promoter contains ®ve tandem repeats of
Fig. 5. GN box-dependent transcriptional repression by Necdin. (A) c-myc P1 promoter assay. Drosophila SL2 cells were transfected with the reporter plasmid pMycP1 (0.8 mg) in combination with the expression vector pPacNecdin (in mg per assay) in the absence (Sp1 2 ) or presence (Sp11) of 0.04 mg pPacSp1 (0.04). The total amount of plasmid DNA was adjusted to 3.2 mg per assay by adding empty pPac. Each value represents the mean 1 SD of two independent experiments. DSp1^, Sp12 subtracted from Sp11. (B) Requirement of the GN box for Necdin-induced repression of c-myc P1 core promoter. SL2 cells were transfected with the GN box-deleted mutant (pMycP1DGN) and a mutant carrying the GC box in place of the GN boxes (pMycP1SGC) in combination with pPacSp1 and pPacNecdin as above.
K. Matsumoto et al. / Gene 272 (2001) 173±179
tion factors have at least 16 different mammalian members (Philipsen and Suske, 1999). Therefore, we infer that Necdin exerts different modulating effects depending on transcription factors, promoter con®gurations, and cellular contexts. Further studies on the interactions between Necdin and the GN boxes may provide valuable information about physiological functions of Necdin in the regulation of gene expression in terminally differentiated cells. 4. Conclusion Necdin binds to the novel motifs containing multiple G clusters and intervening mono- or di-nucleotides of A, T, and C. These motifs resemble those of multiple GC boxes which are potentially recognized by the transcription factor Sp1. Necdin binds to contiguous two GC boxes but not to a single GC box, suggesting that the numbers of Gs and G clusters are critical for the recognition by Necdin. Necdin represses Sp1-dependent transcriptional activity of mouse cmyc P1 proximal core promoter through the GN box. Necdin may regulate the cell proliferation, at least in part, as a transcriptional repressor of cell cycle-related genes containing the GN box in their promoters. Acknowledgements We thank Dr Hisato Kondo for the mouse c-myc gene, and Dr Guntram Suske for Drosophila expression vectors. This work was supported by grants-in-aid for scienti®c research from the Ministry of Education, Culture, Sports, Science and Technology of Japan, by Health Sciences Research Grants (Research on Brain Science) from the Ministry of Health, Labour and Welfare of Japan, and by the Program for Promotion of Fundamental Studies in Health Sciences of the Organization for Pharmaceutical Safety and Research of Japan. References Aizawa, T., Maruyama, K., Kondo, H., Yoshikawa, K., 1992. Expression of necdin, an embryonal carcinoma-derived nuclear protein, in developing mouse brain. Brain Res. Dev. Brain Res. 68, 265±274. Cong, Y.S., Wen, J., Bacchetti, S., 1999. The human telomerase catalytic subunit hTERT: organization of the gene and characterization of the promoter. Hum. Mol. Genet. 8, 137±142. DesJardins, E., Hay, N., 1993. Repeated CT elements bound by zinc ®nger proteins control the absolute and relative activities of the two principal human c-myc promoters. Mol. Cell. Biol. 13, 5710±5724. Gerard, M., Hernandez, L., Wevrick, R., Stewart, C.L., 1999. Disruption of
179
the mouse necdin gene results in early post-natal lethality. Nat. Genet. 23, 199±202. Graham, F.L., van der Eb, A.J., 1973. A new technique for the assay of infectivity of human adenovirus 5 DNA. Virology 52, 456±467. Hayashi, Y., Matsuyama, K., Takagi, K., Sugiura, H., Yoshikawa, K., 1995. Arrest of cell growth by necdin, a nuclear protein expressed in postmitotic neurons. Biochem. Biophys. Res. Commun. 213, 317±324. Jay, P., Rougeulle, C., Massacrier, A., Moncla, A., Mattei, M.G., Malzac, P., Roeckel, N., Taviaux, S., Lefranc, J.L., Cau, P., Berta, P., Lalande, M., Muscatelli, F., 1997. The human necdin gene, NDN, is maternally imprinted and located in the Prader-Willi syndrome chromosomal region. Nat. Genet. 17, 357±361. MacDonald, H.R., Wevrick, R., 1997. The necdin gene is deleted in PraderWilli syndrome and is imprinted in human and mouse. Hum. Mol. Genet. 6, 1873±1878. Majello, B., De Luca, P., Suske, G., Lania, L., 1995. Differential transcriptional regulation of c-myc promoter through the same DNA binding sites targeted by Sp1-like proteins. Oncogene 10, 1841±1848. Maruyama, K., Usami, M., Aizawa, T., Yoshikawa, K., 1991. A novel brain-speci®c mRNA encoding nuclear protein (necdin) expressed in neurally differentiated embryonal carcinoma cells. Biochem. Biophys. Res. Commun. 178, 291±296. McBurney, M.W., Jones-Villeneuve, E.M., Edwards, M.K., Anderson, P.J., 1982. Control of muscle and neuronal differentiation in a cultured embryonal carcinoma cell line. Nature 299, 165±167. Muscatelli, F., Abrous, D.N., Massacrier, A., Boccaccio, I., Moal, M.L., Cau, P., Cremer, H., 2000. Disruption of the mouse Necdin gene results in hypothalamic and behavioral alterations reminiscent of the human Prader-Willi syndrome. Hum. Mol. Genet. 9, 3101±3110. Nakada, Y., Taniura, H., Uetsuki, T., Inazawa, J., Yoshikawa, K., 1998. The human chromosomal gene for necdin, a neuronal growth suppressor, in the Prader-Willi syndrome deletion region. Gene 213, 65±72. Philipsen, S., Suske, G., 1999. A tale of three ®ngers: the family of mammalian Sp/XKLF transcription factors. Nucleic Acids Res. 27, 2991±3000. Pollock, R.M., 1996. Determination of protein-DNA sequence speci®city by PCR-assisted binding-site selection. In: Ausubel, F.M., Brent, R., Kingston, R.E., Moore, D.D., Seidman, J.G., Smith, J.A., Struhl, K. (Eds.), Current Protocols in Molecular Biology. Wiley, New York, pp. 12.11.1±12.11.11. Schreiber, E., Matthias, P., Muller, M.M., Schaffner, W., 1989. Rapid detection of octamer binding proteins with `mini-extracts', prepared from a small number of cells. Nucleic Acids Res. 17, 6419. Sutcliffe, J.S., Han, M., Christian, S.L., Ledbetter, D.H., 1997. Neuronallyexpressed necdin gene: an imprinted candidate gene in Prader-Willi syndrome. Lancet 350, 1520±1521. Taniguchi, N., Taniura, H., Niinobe, M., Takayama, C., TominagaYoshino, K., Ogura, A., Yoshikawa, K., 2000. The postmitotic growth suppressor necdin interacts with a calcium-binding protein (NEFA) in neuronal cytoplasm. J. Biol. Chem. 275, 31674±31681. Taniura, H., Taniguchi, N., Hara, M., Yoshikawa, K., 1998. Necdin, a postmitotic neuron-speci®c growth suppressor, interacts with viral transforming proteins and cellular transcription factor E2F1. J. Biol. Chem. 273, 720±728. Taniura, H., Matsumoto, K., Yoshikawa, K., 1999. Physical and functional interactions of neuronal growth suppressor necdin with p53. J. Biol. Chem. 274, 16242±16248. Uetsuki, T., Takagi, K., Sugiura, H., Yoshikawa, K., 1996. Structure and expression of the mouse necdin gene. Identi®cation of a postmitotic neuron-restrictive core promoter. J. Biol. Chem. 271, 918±924.
www.elsevier.com/locate/gene
Necdin acts as a transcriptional repressor that interacts with multiple guanosine clusters Kuniharu Matsumoto, Hideo Taniura, Taichi Uetsuki, Kazuaki Yoshikawa* Division of Regulation of Macromolecular Functions, Institute for Protein Research, Osaka University, Yamadaoka 3-2, Suita, Osaka 565-0871, Japan Received 26 March 2000; received in revised form 15 May 2001; accepted 29 May 2001 Received by T. Sekiya
Abstract Necdin is a growth suppressor expressed predominantly in postmitotic neurons, and ectopic expression of this protein suppresses cell growth. Here we report that Necdin directly binds to speci®c DNA sequences and serves as a transcriptional repressor. Polyhistidine-tagged Necdin was used for selection of random-sequence oligonucleotides by polymerase chain reaction-based ampli®cation. Necdin recognized guanosine (G)-rich sequences that encompass multiple G clusters and intervening mono- or di-nucleotides of A, T and C. These sequences, termed GN boxes, resemble multiply aligned forms of the canonical GC box which is recognized by Sp family members. Necdin directly bound to a GN box consisting of contiguous two GC boxes with four G clusters, but not to a single GC box with two G clusters, whereas Sp1 bound to both. In a reporter system using Drosophila Schneider Line 2 cells, Necdin repressed Sp1-dependent activity of mouse c-myc P1 promoter that contains a typical GN box. Deletion of the GN box from the c-myc P1 promoter or its conversion to the single GC box abolished the Necdin-dependent repression. These results suggest that Necdin modulates gene transcription via the GN box that is potentially recognized by GC box-targeting Sp family members. q 2001 Elsevier Science B.V. All rights reserved. Keywords: DNA-binding protein; Cell growth suppressor; Cis-acting elements; Postmitotic neurons
1. Introduction Necdin is a 325 amino acid residue protein encoded in a cDNA sequence isolated from neurally differentiated murine embryonal carcinoma P19 cells (Maruyama et al., 1991). In mouse brain, the necdin gene is constitutively expressed in postmitotic neurons, whereas necdin mRNA is undetectable in neuroepithelial stem cells in the neural tube (Aizawa et al., 1992; Uetsuki et al., 1996). Necdin is also expressed in skeletal muscles at early stages of their differentiation (Taniguchi et al., 2000). These expression patterns suggest that the necdin gene is expressed in postmitotic cells that complete terminal differentiation. The human necdin gene NDN is located on chromosome 15q11.2-q12, which is deleted in the Prader±Willi syndrome (PWS) (Jay et al., 1997; MacDonald and Wevrick, 1997; Nakada et al., 1998), a neurogenetic disorder related to genomic imprinting. NDN is maternally imprinted and transcribed only from the paternal allele (Jay et al., 1997; Abbreviations: G, guanosine; His, polyhistidine; PCR, polymerase chain reaction; PWS, Prader±Willi syndrome; SL2, Schneider Line 2 * Corresponding author. Tel.: 181-6-6879-8621; fax: 181-6-6879-8623. E-mail address: [email protected] (K. Yoshikawa).
MacDonald and Wevrick, 1997; Sutcliffe et al., 1997). Necdin is not expressed in the cells prepared from PWS patients whose chromosome 15q11.2-q12 region in the paternal allele is deleted. Necdin is most abundant in postmitotic neurons in the brain stem and hypothalamus of mice (Aizawa et al., 1992; Uetsuki et al., 1996). Disruption of the mouse necdin gene results in early postnatal lethality (Gerard et al., 1999), a reduction in the number of speci®c hypothalamic neurons, and behavioral alternations (Muscatelli et al., 2000), which are reminiscent of human PWS. These observations suggest that Necdin plays a key role in differentiation and development of neurons in the brain, especially in the hypothalamus. Necdin is localized to the nucleus of postmitotic neurons (Maruyama et al., 1991; Aizawa et al., 1992). Ectopic expression of Necdin suppresses proliferation of NIH3T3 cells (Hayashi et al., 1995) and colony formation of SAOS-2 osteosarcoma cells (Taniura et al., 1998). Necdin binds to viral oncoproteins such as simian virus 40 large T antigen and adenovirus E1A (Taniura et al., 1998). Moreover, Necdin interacts with the transactivation domain of E2F1 and p53, major transcription factors that regulate the cell cycle, and suppresses their transcriptional activities (Taniura et al., 1998). These observations led us to speculate
0378-1119/01/$ - see front matter q 2001 Elsevier Science B.V. All rights reserved. PII: S 0378-111 9(01)00544-3
174
K. Matsumoto et al. / Gene 272 (2001) 173±179
that Necdin, like other cell cycle regulatory factors such as E2F1 and p53, is a transcription factor that recognizes speci®c DNA sequences. In this study we have attempted to elucidate a speci®c Necdin-binding cis-element by polymerase chain reaction (PCR)-based selection of random-sequence oligonucleotides using a recombinant Necdin protein. Necdin speci®cally binds to a G-rich motif, designated GN box, which is also recognized by Sp1. We analyzed the promoter activity using GN box-carrying c-myc P1 promoter as a model system and found that Necdin represses the Sp1-dependent promoter activity. 2. Materials and methods 2.1. Synthesis of recombinant Necdin protein in baculovirus expression system Full-length mouse necdin cDNA (Maruyama et al., 1991) was inserted into the baculovirus transfer vector pBlueBac4 (Invitrogen). For construction of a vector expressing polyhistidine (His)-tagged Necdin protein, necdin cDNA was inserted into the baculovirus transfer vector pBlueBacHis2/B (Invitrogen). The transfer vectors and Bac-NBlue AcMNPV were introduced to Sf21 insect cells to obtain recombinant baculoviruses. His-tagged Necdin expressed in virus-infected Sf9 insect cells was puri®ed with Probond metal-binding resin (Invitrogen) and dialyzed against a buffer containing 20 mM Tris±HCl (pH 8.0), 300 mM NaCl, 1 mM EDTA, 1 mM dithiothreitol, 1% Triton X100, 0.5% Tween 20 and 0.1% CHAPS. Proper folding of the Necdin protein was con®rmed by af®nity to E2F1 and p53 as described previously (Taniura et al., 1998, 1999). 2.2. DNA cellulose chromatography Sf9 cells were infected with baculovirus vectors for expression of full-length Necdin, harvested 72 h later, and lysed in a buffer containing 25 mM Tris±HCl (pH 8.0), 10 mM NaCl, 1 mM dithiothreitol, 0.1 mM EDTA, 100 mg/ml phenylmethylsulfonyl ¯uoride, 0.5 mg/ml leupeptin, 0.5 mg/ml aprotinin, 1 mg/ml pepstatin A, and 0.1% Tween 20. The cell lysate or His-tagged Necdin was mixed with 1 ml of DNA cellulose or cellulose (50% slurry) (Sigma) and incubated at 48C for 1 h. The resin was then washed with the same buffer and packed into a column. Bound proteins were eluted with a stepwise gradient of NaCl from 50 mM to 1.0 M and analyzed by Western blotting using an antibody against Necdin (C2) (Maruyama et al., 1991). 2.3. Determination of Necdin-binding sequences Selection of speci®c Necdin-binding sequences was carried out by PCR-based ampli®cation of randomsequence oligonucleotides (Pollock, 1996). A 76 base oligo-
nucleotide 5 0 -CAGGTCAGTTCAGCGGATCCTGTCG(A/ T/G/C)26GAGGCGAATTCAGTGCAACTGCAGC-3 0 was double-stranded by annealing with primer F (5 0 -GCTGCAGTTGCACTGAATTCGCCTC-3 0 ) and extending with the Ex-Taq polymerase (TAKARA, Tokyo) and dNTPs at 948C (1 min), 628C (3 min), and 728C (9 min). The product was puri®ed by phenol-chloroform extraction followed by ethanol precipitation, and dissolved in a buffer containing 10 mM Tris±HCl (pH 8.0) and 1 mM EDTA. Puri®ed Histagged Necdin was immobilized onto Ni-chelating resin, and incubated with double-stranded degenerated oligonucleotides at 48C for 30 min in a buffer (50 ml) containing 15 mM HEPES-KOH (pH 7.5), 4% Ficoll 400, 50 mM KCl, and 4 mg/ml bovine serum albumin. Protein±DNA complexes were precipitated by centrifugation and washed three times with the buffer. Bound oligonucleotides were recovered by boiling the resin and used as a template of PCR with primer R (5 0 -CAGGTCAGTTCAGCGGATCCTGTCG-3 0 ) and primer F (160 ng each) for 15±20 cycles (948C for 1 min, 628C for 1 min, 728C for 1 min). PCR was carried out in the above buffer containing 10 mg/ml poly(dIdC)(dI-dC). The PCR product (,50 ng of double-stranded DNA) was used for the successive round of selection. The oligonucleotide pool was cloned into Bluescript II SK 1 , and the sequences of 14 randomly selected clones were determined with an automatic DNA sequencer (Model 4000, LI-COR, Lincoln, NE). 2.4. Electrophoretic mobility shift assay Oligonucleotides selected at each cycle of enrichment were labeled with [a- 32P]dCTP by PCR. The PCR products were extracted from an 8% polyacrylamide gel and used as probes. Puri®ed His-tagged Necdin (80 ng protein per assay) was preincubated with 32P-labeled oligonucleotides at 48C for 15 min in a buffer (20 ml) containing 15 mM HEPESKOH (pH 7.9), 4% Ficoll 400, 4 mg/ml bovine serum albumin, 50 mM KCl, 1 mM EDTA, and 100 ng of poly(dIdC)(dI-dC). For detection of protein±GC box complexes, double-stranded oligonucleotides 5 0 -ATTCGATCGGGGCGGGGCGAGC-3 0 , 5 0 -ATTCGATCGGTTCGGGGCGAGC-3 0 , and 5 0 -ATTCGATCGGGGCGGGTGGGCGGGGCGAGC-3 0 were synthesized and used as probes of a single GC box (1 £ GC), mutated GC box, and contiguous two GC boxes (2 £ GC box), respectively. A doublestranded oligonucleotide 5 0 -GGCTAATCCCCGCCCACCCGCCCTTTATATTCCGGGGGTC-3 0 corresponding to a c-myc P1 GN box was also used. The oligonucleotide probes were labeled with [g- 32P]ATP using T4 polynucleotide kinase. Nuclear extracts were prepared from undifferentiated mouse embryonal carcinoma P19 cells (McBurney et al., 1982) and Drosophila Schneider Line 2 (SL2) transfected with Sp1 expression plasmid pPacSp1 (Schreiber et al., 1989). The nuclear extracts (2 mg protein each) and Histagged Necdin (80 ng protein) were preincubated for 10 min in a reaction mixture (20 ml) containing 15 mM HEPES-
K. Matsumoto et al. / Gene 272 (2001) 173±179
175
galactosidase (p97b) were kindly provided by Dr Guntram Suske (Erasmus University, The Netherlands). Necdin cDNA was inserted into the Drosophila expression vector pPac to make pPacNecdin. Expression of Necdin in SL2 cells was con®rmed by immunoblotting with anti-Necdin antibody C2 (Maruyama et al., 1991). Sets of plasmids were co-transfected into ,70% con¯uent SL2 cells in a 35 mm dish by the calcium phosphate method (Graham and van der Eb, 1973). Luciferase activities were measured with a luminometer (Lumat LB9501, Berthold) using a reagent kit (Toyo Ink, Tokyo, Japan). The transfection ef®ciency was normalized with co-transfected b-galactosidase reporter plasmid (p97b).
3. Results and discussion Fig. 1. DNA-binding properties of recombinant Necdin proteins synthesized in the baculovirus system. (A) DNA binding of Necdin and Histagged Necdin proteins. The lysate of Sf9 cells expressing Necdin (Necdin) or His-tagged Necdin (His-Necdin) was mixed with cellulose or DNA cellulose in a binding buffer containing 10 mM NaCl. Bound proteins were eluted with stepwise gradient of NaCl at the concentrations indicated (in M), and analyzed by Western blotting with antibody C2 against Necdin. (B) Electrophoretic mobility shift assay for Necdin-binding sequences selected by the PCR-based selection method. A 76 bp oligonucleotide containing 26 random-sequence nucleotides at the central region was used as a probe for initial PCR. Selected DNA sequences were radiolabeled, incubated with His-tagged Necdin, and analyzed by the electrophoretic mobility shift assay. Note the progressive enrichment of Necdin-binding sequences after successive rounds of selection. CX, DNA±Necdin complexes; F, free probe.
KOH (pH 7.9), 1 mM EDTA, 4% Ficoll 400, 4 mg/ml bovine serum albumin, 50 mM KCl, 6.25 mM MgCl2, 25 mM ZnCl2, and 50 mg/ml poly(dA-dT)(dA-dT). The radiolabeled probes were added to the mixture, which was incubated at 48C for 30 min. DNA-binding activities were analyzed by electrophoresis in an 8% non-denaturing polyacrylamide gel (acrylamide/bisacrylamide, 29:1) run in a buffer containing 45 mM Tris±borate (pH 8.2), 1 mM EDTA at 48C. Signals were detected by autoradiography. 2.5. Reporter assay Mouse c-myc P1 core promoter region (positions 268 to 122) was prepared by PCR using mouse genomic c-myc DNA (a gift from Dr Hisato Kondo, Osaka University) as a template. PCR products were cloned into the luciferase reporter plasmid PGV-B to make pMycP1. Necdin-binding sequence (GN box) was eliminated from the P1 promoter to make pMycP1DGN by self-ligation of the PCR product synthesized with the GN box-¯anking sequences using pMycP1 as a template. The myc GN box was converted to 1 £ GC box to make pMycP1SGC by self-ligation of PCR product synthesized with primers 5 0 -GATTAGCCAGAGAACCTCTCTTTCTCCCC-3 0 and 5 0 -ACCCGCCCTTTATATTCCGGGGGTC-3 0 using the pMycP1 template. Drosophila expression plasmids for Sp1 (pPacSp1) and b-
3.1. DNA-binding properties of the Necdin protein synthesized in the baculovirus system To test whether Necdin directly binds to DNA, we synthesized the Necdin protein in the baculovirus system. The extract of Necdin-expressing Sf9 cells was applied to DNA cellulose column chromatography. Necdin bound to DNA cellulose and was eluted with 300 mM NaCl, but it failed to interact with the cellulose matrix (Fig. 1A, upper), suggesting that Necdin expressed in Sf9 cells is competent for DNA binding, although it remained unclear whether Necdin indirectly bound to DNA via other proteins present in Sf9 cells. We then synthesized the Necdin protein as a His-tagged fusion protein which was puri®ed by Ni-chelating af®nity chromatography. Puri®ed His-tagged Necdin also bound to DNA cellulose (Fig. 1A, lower). Using this recombinant protein, we determined the speci®c Necdin-binding sequences by a repetitive PCR-assisted DNA-binding site selection. Necdin-binding sequences were radiolabeled and monitored by electrophoretic mobility shift assay at each round of selection (Fig. 1B). At the ®fth round, a distinct protein±DNA complex became apparent. Several complexes with different sizes were clearly detected at rounds 7, 8 and 9. These complexes may correspond to multiple Necdin molecules that bind to the selected oligonucleotides. The oligonucleotide pool selected at the seventh round was cloned into a plasmid, and 14 independent clones were randomly selected (Fig. 2). All of the clones contained inserts carrying highly G-rich sequences, which are heterogeneous and contain no typical consensus motifs. The heterogeneity of the selected oligonucleotides may have different af®nities to the Necdin protein oligomers, which may give rise to the multiple bands in the gel shift assay shown in Fig. 1B. Although 26 random-sequence oligonucleotides were utilized in this system, the selected sequences contained 23±26 nucleotides (mean ^ SEM 25 ^ 0.25). The statistical data of these oligonucleotides are as follows: the numbers of Gs are 19±21 (19.6 ^ 0.17) and %G values are 76±83 (78.6 ^ 0.71). Each oligonucleotide contains four
176
K. Matsumoto et al. / Gene 272 (2001) 173±179
to ®ve G clusters (4.8 ^ 0.25), and the number of Gs per cluster varies from two to ten (4 ^ 0.3). Mono- or di-nucleotides separate the G clusters (mono-nucleotides, 57; dinucleotides, 9; A 34, T 25, C 16 in 14 sequences). Since the mean number of Gs per cluster was four, large G clusters consisting of greater than or equal to seven Gs, which are present in 12 sequences, are assumed to be divided into two smaller clusters consisting of greater than or equal to three Gs by `intervening G'. Thus, a typical Necdin-binding sequence is summarized as a highly Grich sequence (%G ,80) that consists of four to ®ve clusters of contiguous ,four Gs and intervening A, T, C, and G. We termed this speci®c G-rich motif `GN boxes'. 3.2. Necdin binds to the GN box but not to the canonical GC box
Fig. 2. Necdin-binding sequences. Enriched oligonucleotides after seven rounds of selection were ampli®ed by PCR and cloned into Bluescript II. Inserts of 14 randomly selected clones were sequenced. Contiguous G clusters (n $ 2) are underlined, and intervening nucleotides A, T, and C are dotted. Statistical data are in Section 3.
The GN boxes resemble multiply aligned GC boxes, which are potentially recognized by speci®c transcription factors such as Sp family members (Philipsen and Suske, 1999). We then tested whether Necdin recognizes the canonical GC box (GGGGCGGGG) using an oligonucleotide probe in the electrophoretic mobility shift assay. Necdin failed to form a complex with the GC box oligonucleotide (Fig. 3A, lane 3). As a positive control for GC box-binding activity, we prepared the nuclear extract from undifferentiated P19 cells. GC box-speci®c complexes were formed in the nuclear extract only with the probe containing canonical GC box
Fig. 3. Necdin binds to a GN box consisting of contiguous two GC boxes. (A) Inability of Necdin to bind to the canonical GC box. Puri®ed His-tagged Necdin (Necdin) and P19 nuclear extract (P19 NE) were incubated with oligonucleotide probes containing the GC box (GC) (C) and its binding-defective mutant (M). Note that Necdin neither forms complexes with the canonical GC box (lane 3) nor competes for formation of the GC box±protein complex (lane 7). CX, DNA± protein complexes; F, free probe. (B) Complex formation between Necdin and contiguous two GC boxes. Nuclear extracts from Drosophila SL2 cells transfected with Sp1 expression vector (Sp1) and untransfected SL2 cells (SL2) were incubated with radiolabeled oligonucleotide probes, and formation of the complexes was analyzed by electrophoretic mobility shift assay. His-tagged Necdin (Necdin) binds to contiguous two GC boxes (2 £ GC) (lane 2), but not to a single GC box (1 £ GC) (lane 6). Arrows point to the complexes speci®c to Sp1 and Necdin. F, free probe.
K. Matsumoto et al. / Gene 272 (2001) 173±179
but not with the mutated probe (Fig. 3A, lanes 5 and 6). Histagged Necdin protein added to the reaction mixture failed to compete with complexes between the GC box and its binding proteins (Fig. 3A, lane 7). We then synthesized an oligonucleotide containing contiguous two GC boxes (2 £ GC, GGGGCGGGTGGGCGGGG), which ful®lls the criteria for GN boxes (%G 82, four clusters of contiguous three to four G, intervening C and T). His-tagged Necdin interacted with 2 £ GC box but not with 1 £ GC box (Fig. 3B, lanes 2 and 6). On the other hand, the nuclear extract from Sp1expressing Drosophila SL2 cells formed complexes with both 1 £ GC and 2 £ GC (Fig. 3B, lanes 3 and 7). The Necdin
177
protein failed to compete against the complex formation between Sp1 and GN box (2 £ GC) in this gel shift assay (data not shown), presumably because the GN box-binding potency of puri®ed His-tagged Necdin protein is weaker than that of the Sp1 protein produced in SL2 cells. 3.3. Necdin binds to a GN box in the c-myc P1 promoter We constructed a luciferase reporter plasmid carrying 4 £ GC boxes (in normal orientation) upstream of TATA box, and found that the basal activity of this reporter was signi®cantly reduced by cotransfected necdin cDNA in
Fig. 4. Characterization of the GN box-like motif in c-myc P1 core promoter as a Necdin-binding sequence. (A) Diagram illustrating the reporter constructs carrying mouse c-myc P1 core promoter (pMycP1) and a GN box-deleted mutant (pMycP1DGN). GN, GN box (-like motif); TATA, TATA box; Luc, luciferase reporter gene; P1, P1 transcription initiation site. pMycP1 contains a GN box-like motif (in inverted orientation), which was eliminated to make pMycP1DGN. Positions are numbered, beginning with the P1 site (11). (B) Complex formation between Necdin and c-myc GN box. A complex between Histagged Necdin (Necdin) and radiolabeled c-myc P1 GN box probe was analyzed in the presence of 100- and 200-fold molar excesses ( £ 100, £ 200) of the oligonucleotide competitors (Comp) of c-myc P1 GN box (P1GN) and GC box (GC). CX, GN box±necdin complex; F, free probe. (C) Activation of c-myc P1 promoter by Sp1 through the GN box. Drosophila SL2 cells were transfected with pMycP1 and pMycP1DGN reporter plasmids (0.8 mg each) in combination with pPacSp1 (in ng per assay). The total amount of plasmid DNA was adjusted to 3.2 mg per assay by adding empty pPac vector. Each value represents the mean 1 SD of two independent experiments.
178
K. Matsumoto et al. / Gene 272 (2001) 173±179
NIH3T3 cells (data not shown). We then searched natural promoters for GN box-like motifs to examine whether interactions between Necdin and GN box are of physiological signi®cance. We found a GN box-like motif (5 0 -GGGCGGGTGGGCGGGG-3 0 in inverted orientation) adjacent to the TATA box of the c-myc P1 promoter (Fig. 4A). This motif represents a typical GN box (%G 81, four clusters of contiguous three to four G, intervening C and T). The gel shift assay revealed that Necdin bound to this proximal promoter region at positions 268 to 122 (Fig. 4B). The shifted signal was markedly reduced by addition of the GN box sequence in 200-fold molar excess (Fig. 4B, lanes 3 and 4), but the same molar excess of the GC box sequence, which was used as a negative control, showed only a slight reduction (Fig. 4B, lanes 5 and 6), suggesting that the length of the G cluster in the single GC box is insuf®cient to compete with P1 GN box. We then examined whether the myc P1 core promoter shows responsiveness to Sp1 through the GN box. We were unable to demonstrate speci®c repression by Necdin in Sp1-dependent transactivation in several mammalian cell lines which normally contain high levels of endogenous Sp family proteins (data not shown). We alternatively used Drosophila SL2 cells, which contain no Sp family proteins and are commonly used for studies on Sp-related transcription factors (Majello et al., 1995). c-Myc core promoter was strongly activated by Sp1 in SL2 cells (Fig. 4C). The promoter defective in the GN box (pMycP1DGN) showed much less responsiveness to Sp1, suggesting that the GN box mediates the Sp1-dependent activation of the c-myc P1 promoter.
the consensus sequence CCCTCCCC known as the CT element (DesJardins and Hay, 1993) between 2117 and 2138 upstream of the P1 initiation site (Majello et al., 1995). We noted that this region represents a typical GN box motif (5 0 -GGGGAGGGTGGGGAGGGTGGGG-3 0 in inverted orientation) (%G 82, G 18, ®ve G clusters). A GN box-like motif is also found in the promoter region of the human telomerase catalytic subunit hTERT (5 0 -GGGAGGGGTCGGGACGGGGCGGGG-3 0 between 2144 and 2167 in inverted orientation) (%G 78, G 18, ®ve G clusters) (Cong et al., 1999). We infer that Necdin potentially modulates the transcription of these genes via the GN boxes in their promoters. In this study, we used the Sp1 system as a model to elucidate the physiological signi®cance of interactions between Necdin and the GN box. Sp1 may be one of many transcription factors acting on the GN box. For example, Sp makes a family of at least ®ve members Sp1±Sp5. Some of them are thought to act as negative regulators. Furthermore, it is currently known that Sp-related transcrip-
3.4. Necdin represses Sp1-dependent activities of a GN boxcontaining natural promoter Necdin repressed the Sp1-dependent transactivation of the myc P1 core promoter in a dose-dependent manner (Fig. 5A). Necdin slightly activated the GN box-containing promoter in the absence of Sp1, suggesting that Necdin acts as a partial transcriptional activator for GN box-carrying promoter. In Drosophila SL2 cells, the higher concentration of Necdin was required to compete with Sp1. Subtraction of basal activities (Sp12) from Sp1-activated values revealed that Necdin maximally suppresses the Sp1-dependent activity below the control value (DSp1^). A mutant defective in the GN box (pMycP1DGN) was slightly activated by Sp1 but showed no response to Necdin. Another mutant carrying the GC box in place of the GN box (pMcyP1SGC) was moderately activated by Sp1, but not repressed by Necdin (Fig. 5B). Necdin exerted no effects on the basal activities (Sp12) of pMycP1DGN and pMcyP1SGC (data not shown). These results indicate that Necdin represses Sp1-dependent myc P1 promoter activity via its GN box, and that the repression of Sp1-dependent transactivation is observed only when the Sp1-binding sequence overlaps with the GN box. Human c-myc promoter contains ®ve tandem repeats of
Fig. 5. GN box-dependent transcriptional repression by Necdin. (A) c-myc P1 promoter assay. Drosophila SL2 cells were transfected with the reporter plasmid pMycP1 (0.8 mg) in combination with the expression vector pPacNecdin (in mg per assay) in the absence (Sp1 2 ) or presence (Sp11) of 0.04 mg pPacSp1 (0.04). The total amount of plasmid DNA was adjusted to 3.2 mg per assay by adding empty pPac. Each value represents the mean 1 SD of two independent experiments. DSp1^, Sp12 subtracted from Sp11. (B) Requirement of the GN box for Necdin-induced repression of c-myc P1 core promoter. SL2 cells were transfected with the GN box-deleted mutant (pMycP1DGN) and a mutant carrying the GC box in place of the GN boxes (pMycP1SGC) in combination with pPacSp1 and pPacNecdin as above.
K. Matsumoto et al. / Gene 272 (2001) 173±179
tion factors have at least 16 different mammalian members (Philipsen and Suske, 1999). Therefore, we infer that Necdin exerts different modulating effects depending on transcription factors, promoter con®gurations, and cellular contexts. Further studies on the interactions between Necdin and the GN boxes may provide valuable information about physiological functions of Necdin in the regulation of gene expression in terminally differentiated cells. 4. Conclusion Necdin binds to the novel motifs containing multiple G clusters and intervening mono- or di-nucleotides of A, T, and C. These motifs resemble those of multiple GC boxes which are potentially recognized by the transcription factor Sp1. Necdin binds to contiguous two GC boxes but not to a single GC box, suggesting that the numbers of Gs and G clusters are critical for the recognition by Necdin. Necdin represses Sp1-dependent transcriptional activity of mouse cmyc P1 proximal core promoter through the GN box. Necdin may regulate the cell proliferation, at least in part, as a transcriptional repressor of cell cycle-related genes containing the GN box in their promoters. Acknowledgements We thank Dr Hisato Kondo for the mouse c-myc gene, and Dr Guntram Suske for Drosophila expression vectors. This work was supported by grants-in-aid for scienti®c research from the Ministry of Education, Culture, Sports, Science and Technology of Japan, by Health Sciences Research Grants (Research on Brain Science) from the Ministry of Health, Labour and Welfare of Japan, and by the Program for Promotion of Fundamental Studies in Health Sciences of the Organization for Pharmaceutical Safety and Research of Japan. References Aizawa, T., Maruyama, K., Kondo, H., Yoshikawa, K., 1992. Expression of necdin, an embryonal carcinoma-derived nuclear protein, in developing mouse brain. Brain Res. Dev. Brain Res. 68, 265±274. Cong, Y.S., Wen, J., Bacchetti, S., 1999. The human telomerase catalytic subunit hTERT: organization of the gene and characterization of the promoter. Hum. Mol. Genet. 8, 137±142. DesJardins, E., Hay, N., 1993. Repeated CT elements bound by zinc ®nger proteins control the absolute and relative activities of the two principal human c-myc promoters. Mol. Cell. Biol. 13, 5710±5724. Gerard, M., Hernandez, L., Wevrick, R., Stewart, C.L., 1999. Disruption of
179
the mouse necdin gene results in early post-natal lethality. Nat. Genet. 23, 199±202. Graham, F.L., van der Eb, A.J., 1973. A new technique for the assay of infectivity of human adenovirus 5 DNA. Virology 52, 456±467. Hayashi, Y., Matsuyama, K., Takagi, K., Sugiura, H., Yoshikawa, K., 1995. Arrest of cell growth by necdin, a nuclear protein expressed in postmitotic neurons. Biochem. Biophys. Res. Commun. 213, 317±324. Jay, P., Rougeulle, C., Massacrier, A., Moncla, A., Mattei, M.G., Malzac, P., Roeckel, N., Taviaux, S., Lefranc, J.L., Cau, P., Berta, P., Lalande, M., Muscatelli, F., 1997. The human necdin gene, NDN, is maternally imprinted and located in the Prader-Willi syndrome chromosomal region. Nat. Genet. 17, 357±361. MacDonald, H.R., Wevrick, R., 1997. The necdin gene is deleted in PraderWilli syndrome and is imprinted in human and mouse. Hum. Mol. Genet. 6, 1873±1878. Majello, B., De Luca, P., Suske, G., Lania, L., 1995. Differential transcriptional regulation of c-myc promoter through the same DNA binding sites targeted by Sp1-like proteins. Oncogene 10, 1841±1848. Maruyama, K., Usami, M., Aizawa, T., Yoshikawa, K., 1991. A novel brain-speci®c mRNA encoding nuclear protein (necdin) expressed in neurally differentiated embryonal carcinoma cells. Biochem. Biophys. Res. Commun. 178, 291±296. McBurney, M.W., Jones-Villeneuve, E.M., Edwards, M.K., Anderson, P.J., 1982. Control of muscle and neuronal differentiation in a cultured embryonal carcinoma cell line. Nature 299, 165±167. Muscatelli, F., Abrous, D.N., Massacrier, A., Boccaccio, I., Moal, M.L., Cau, P., Cremer, H., 2000. Disruption of the mouse Necdin gene results in hypothalamic and behavioral alterations reminiscent of the human Prader-Willi syndrome. Hum. Mol. Genet. 9, 3101±3110. Nakada, Y., Taniura, H., Uetsuki, T., Inazawa, J., Yoshikawa, K., 1998. The human chromosomal gene for necdin, a neuronal growth suppressor, in the Prader-Willi syndrome deletion region. Gene 213, 65±72. Philipsen, S., Suske, G., 1999. A tale of three ®ngers: the family of mammalian Sp/XKLF transcription factors. Nucleic Acids Res. 27, 2991±3000. Pollock, R.M., 1996. Determination of protein-DNA sequence speci®city by PCR-assisted binding-site selection. In: Ausubel, F.M., Brent, R., Kingston, R.E., Moore, D.D., Seidman, J.G., Smith, J.A., Struhl, K. (Eds.), Current Protocols in Molecular Biology. Wiley, New York, pp. 12.11.1±12.11.11. Schreiber, E., Matthias, P., Muller, M.M., Schaffner, W., 1989. Rapid detection of octamer binding proteins with `mini-extracts', prepared from a small number of cells. Nucleic Acids Res. 17, 6419. Sutcliffe, J.S., Han, M., Christian, S.L., Ledbetter, D.H., 1997. Neuronallyexpressed necdin gene: an imprinted candidate gene in Prader-Willi syndrome. Lancet 350, 1520±1521. Taniguchi, N., Taniura, H., Niinobe, M., Takayama, C., TominagaYoshino, K., Ogura, A., Yoshikawa, K., 2000. The postmitotic growth suppressor necdin interacts with a calcium-binding protein (NEFA) in neuronal cytoplasm. J. Biol. Chem. 275, 31674±31681. Taniura, H., Taniguchi, N., Hara, M., Yoshikawa, K., 1998. Necdin, a postmitotic neuron-speci®c growth suppressor, interacts with viral transforming proteins and cellular transcription factor E2F1. J. Biol. Chem. 273, 720±728. Taniura, H., Matsumoto, K., Yoshikawa, K., 1999. Physical and functional interactions of neuronal growth suppressor necdin with p53. J. Biol. Chem. 274, 16242±16248. Uetsuki, T., Takagi, K., Sugiura, H., Yoshikawa, K., 1996. Structure and expression of the mouse necdin gene. Identi®cation of a postmitotic neuron-restrictive core promoter. J. Biol. Chem. 271, 918±924.