Cell cycle regulation of the murine cdc25B promoter: essential role for nuclear factor-Y and a proximal repressor element.

Expression of the cdc25B gene is up-regulated late during cell cycle progression (S/G(2)). We have cloned the murine cdc25B promoter to identify elements involved in transcriptional regulation. A detailed structure-function analysis led to the identification of several elements that are located upstream of a canonical Inr motif at the site of transcription initiation and are involved in transcriptional activation and regulation. Activation of the promoter is largely mediated by NF-Y and Sp1/3 interacting with one and four proximal binding sites, respectively. In addition, NF-Y plays an essential role in cell cycle regulation in conjunction with a repressor element (cell cycle-regulated repressor) located approximately 30 nucleotides upstream of the putative Inr element and overlapping a consensus TATA motif. The cell cycle-regulated repressor is unrelated to the previously described cell cycle-regulated repressor elements. Taken together, our observations suggest that expression of the cdc25B gene is controlled through a novel mechanism of cell cycle-regulated transcription.

B-myb promoter identified an E2F binding site close to the transcription start sites, which is necessary but not sufficient for cell cycle regulation (15,16). Mutational analyses showed that an adjacent element, termed Bmyb-CHR, 1 is indispensable for repression and acts as a corepressor element together with the E2F-binding site.
cdc25C exemplifies a group of cell cycle genes whose transcription is up-regulated later than that of B-myb, i.e. in S/G 2 . cdc25 was originally discovered in Schizosaccharomyces pombe as a regulator of the G 2 to M progression (17,18). Higher eukaryotes contain at least three genes with a high degree of similarity to cdc25, encoding the Cdc25A, Cdc25B, and Cdc25C protein phosphatases (19 -28). The Cdc25C phosphatase activates the Cdc2/cyclin B complex and thereby enables the entry into mitosis (20, 24, 28 -30). Cdc25A appears to play a role in regulating entry into S phase (13,26,31), whereas Cdc25B is required for the G 2 to M progression (32)(33)(34)(35)(36).
For the cdc25C promoter, repression of upstream activators via a bipartite site, consisting of the "cell cycle-dependent element" and the "cell cycle genes homology region" (CHR), has been established as the major regulatory mechanism (37,38). As shown by genomic footprinting, both elements are cooperatively bound in a periodic fashion by a repressor that has been designated CDF-1 (37,39). A similar mechanism seems to be of global relevance, because a number of other similarly regulated cell cycle genes, such as cyclin A (37,40), cdc2 (37,41), CENP-A (42), polo-like kinase (43), and survivin (44), have been identified. Recently, a factor (CHF) interacting with the CHR in the cyclin A promoter has been described (45).
Cell cycle regulation of cdc25B resembles that of cdc25C, which is in agreement with its function at the final stages of the cell cycle (32)(33)(34)(35)(36). The cdc25B gene is of interest also in view of its possible involvement in human cancer (19, 46 -48), and its oncogenic potential in transgenic mice (49,50). However, to date, the promoter of the cdc25B gene has not been analyzed, and consequently the mechanism controlling the cell cycleregulated expression is unknown. In the present study, we have addressed this question. We have cloned the murine cdc25B promoter and have identified regulatory elements and interacting transcription factors required for cdc25B transcription and contributing to its regulation of expression during the cell cycle.

EXPERIMENTAL PROCEDURES
Cell Culture-The murine cell line NIH3T3 (kindly provided by R. Treisman, ICRF, London) was maintained at 37°C in 5% CO 2 in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine * This work was supported by grants from the Deutsche Forschungsgemeinschaft (SFB397/C1, Mu601/9-2). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM  Transfections and Luciferase Assays-Cells were plated on 35-mm (diameter) tissue culture plates at a density producing 60 -80% confluence at the time of the transfection and transfected using the cationic lipid DOTAP as described by the manufacturer (Roche Molecular Biochemicals). For synchronization in G 0 , cells were maintained in serumfree medium for 3 days. Stimulation was carried out for the indicated times with 10% fetal calf serum. Luciferase activity was determined as published elsewhere (38,51).
Primer Extension Analysis-32 P-labeled primer (10 pmol) and total cellular RNA, isolated from normal cycling NIH3T3 cells, were denatured for 10 min at 65°C and then incubated for 30 min at 37°C. Primer extension was carried out in a total volume of 50 l containing 50 mM Tris, pH 8.3, 75 mM KCl, 10 mM dithiothreitol, 3 mM MgCl 2 , 400 M dNTPs, 2 units of RNasin, and 400 units of Moloney murine leukemia virus reverse transcriptase (Life Technologies, Inc.). After incubation for 1 h at 37°C, the reaction was stopped with EDTA followed by RNase treatment. The DNA was precipitated, redissolved, and separated by electrophoresis on a 6% acrylamide, 7 M urea gel.
Reverse Transcriptase PCR-For cDNA synthesis (53), 4 g of total RNA were annealed to 1 g of oligo(dT) and incubated with 200 units of Moloney murine leukemia virus reverse transcriptase for 1 h at 37°C in a final volume of 20 l. One-tenth of the reaction mixture was amplified by 25 cycles of PCR in the presence of 0.5 Ci of [␣-32 P]dCTP (38,54). The experimental strategy included the following precautions. (i) The number of PCR cycles was kept low to obtain a linear amplification of the PCR products, which was possible by the incorporation of radioactive precursor nucleotides and evaluation by autoradiography and ␤-radiation scanning. (ii) All results were standardized using the signal obtained with glyceraldehyde-3-phosphate dehydrogenase, whose expression is independent of cell proliferation. (iii) All experiments were performed with at least two independent cDNA preparations.
cdc25B Promoter Constructs-Primers carrying restriction sites were used for PCR with pBIISKcdc25B as the template to generate a series of 5Ј terminal deletions with compatible ends for cloning as KpnI/NheI fragments into the multiple cloning region of the promoterless luciferase vector pGL3-basic (Promega, Madison, WI). All PCR-amplified fragments were verified by DNA sequencing. 1-7-base pair mutations were introduced into the regions of the cdc25B promoter spanning Ϫ950 to ϩ167 or Ϫ223 to ϩ167 using PCR-directed mutagenesis (37). Primers carrying the mutations (see below) and a second set of primers for subcloning (5Јcdc25B, 5Јcdc25B223, and 3Јcdc25B) were designed. The first PCR reaction (54) was performed with the oligonucleotides (i) 5Јcdc25B and 3Ј-primer carrying the mutation and (ii) 3Јcdc25B and 5Ј-primer carrying the mutation. The resulting products were purified (QIAquick Spin PCR purification; Qiagen) and amplified in a second PCR reaction using 5Јcdc25B or 5Јcdc25B223 and 3Јcdc25B as primers. Site-directed mutagenesis of the first E-box (Ϫ947) (mutated bases underlined) was generated by PCR with the primer 5ЈmE1 (5Ј-AGCT-GGTACCTTCTCAAGCTTTCCCACTAGGTCCTTCCCAG-3Ј) and the primer cdc25B NheI (see below). The resulting fragments carrying the mutations were cloned into the KpnI/NheI sites of the promoterless luciferase vector pGL3-basic (Promega) and verified by DNA sequencing.
The following oligonucleotides were used as primers: cdc25B KpnI, Electrophoretic Mobility Shift Assays-Preparation of nuclear extracts and electrophoretic mobility shift assays (EMSAs) were performed as described (55,56) using poly(dI⅐dC) or poly(dA⅐dT) for CCRR gel shifts or poly(dI⅐dC) for Sp1 and NF-Y gel shifts as nonspecific competitors. 1-2 l of HeLa or 4 to 6 l of NIH3T3 nuclear extracts were incubated with ϳ0.5 pmol of radiolabeled probe in the suitable binding buffer (NF-Y EMSA: 10 mM Hepes (pH 7.8), 50 mM K-glutamate, 5 mM MgCl 2 , 1 mM dithiothreitol, 5% (v/v) glycerol, 1 mM EDTA (pH 8.0), 0.5 g/l poly(dI⅐dC); Sp1 EMSA: 20 mM Tris⅐Cl (pH 7.5), 0.1 mM EDTA, 0.5 mM MgCl 2 , 10 mM KCl, 0.2 mM ZnSO 4 , 10% glycerol, 0.4 g/l poly(dI⅐dC); CCRR EMSA: 100 mM Tris⅐Cl (pH 7.9), 30% glycerol, 0.4 mM EDTA (pH 8.0), 2 mM dithiothreitol, 0.5 g/l poly(dA⅐dT)). EMSA reactions for Sp1/3 and NF-Y binding were performed at room temperature for 15 min followed by gel electrophoresis at 4°C using 4% polyacrylamide gels. Supershifts were carried out by pre-incubating EMSA reactions on ice for 20 min with 1 l of the indicated antibodies prior to addition of the radiolabeled probe. For detection of other protein-DNA complexes, EMSA reactions were carried out on ice for 15 min. Sp1 and Sp3 antibodies were obtained from G. Suske (IMT, Mar-burg, Germany). The NF-Y antibody was obtained from R. Mantovani (Milan). The following oligonucleotides were used as probes and/or competitors: cdc25B NF-Y, 5Ј-GGAACCGGCGCCCCCATTGGTCG-3Ј; bona fide NF-Y, 5Ј-GATTTTTTCCTGATTGGTTAAAAGT-3Ј; mcdc25B NF-Y (MY), 5Ј-GGAACCGGCGCCCCCATTAATGG-3Ј; GT box, 5Ј-AG- Genomic Footprinting-For genomic footprinting (38,57), NIH3T3 cells were maintained in serum-free medium for 3 days for synchronization in G 0 , and stimulation was carried out for the indicated times with 10% fetal calf serum. The cells were then treated with 0.2% DMS for 2 min. After DMS treatment, cells were washed three times with cold phosphate-buffered saline, and the DNA was isolated. As reference, NIH3T3 genomic DNA was methylated in vitro with 0.2% DMS for 10 -30 s. Piperidine cleavage was performed as described. Genomic DNA (3 g) was used for ligation-mediated PCR as described. The Stoffel fragment of Taq polymerase (PerkinElmer Life Sciences) was used instead of the native enzyme. Samples were phenol-extracted and ethanol-precipitated before primer extension with 32 P-labeled primers.

RESULTS
Cloning of the Mouse cdc25B Promoter-A mouse embryo genomic DNA library was screened with an oligonucleotide representing the mouse cdc25B coding region. Several recombinant phage spanning ϳ30 kilobases of genomic DNA were isolated and mapped (Fig. 1A). One phage clone (designated III in Fig. 1A) was used to subclone a 1.1-kilobase fragment representing the sequence 5Ј to the translation start codon. This fragment (B950) was linked to the firefly luciferase gene and transfected into NIH3T3 cells to test whether the isolated promoter fragment was functional in a transient expression assay. As shown in Fig. 2A, B950 was cell cycle-regulated after serum stimulation of cells that had been synchronized in G 0 . Thus, hardly any luciferase activity was detectable in G 0 cells and at early stages after serum stimulation, but there was an ϳ4-fold induction at 18 h after serum stimulation, peaking at 22 h (8-fold induction). At this stage, most cells had entered or passed through G 2 (data not shown). In addition, we determined the expression profile of the endogenous cdc25B gene in the same cell system and found a similar time course (Fig. 2B; cdc2 induction shown for comparison). These data indicate that the isolated promoter fragment is sufficient to confer on a luciferase reporter gene a pattern of cell cycle regulation that mirrors its physiological regulation.
Structure of the Mouse cdc25B Promoter-The nucleotide sequence of B950 was determined for both strands (Gen-Bank TM accession number AJ296019). The most relevant part of the sequence, as determined below, is shown in Fig. 1B. Inspection of the sequence revealed a match with a canonical TATA box motif 190 nucleotides 5Ј to the ATG (Fig. 1C). A single transcription start site cluster was identified by primer extension analysis ϳ30 nucleotides downstream of this motif and overlapping with an Initiator (Inr) consensus sequence (Figs. 1C and 3). Although we cannot formally rule out the formal possibility that the cdc25B gene contains additional initiation sites outside the region analyzed, these observations strongly suggest that a TATA box and/or an Inr element direct the initiation of transcription and define the transcriptional start site. The A within the Inr motif was therefore designated position ϩ1 (see Fig. 1C). A search for potential regulatory sites revealed the presence of additional putative transcription factor binding sites: two E boxes (Ϫ947 and Ϫ800), three E2F sites (Ϫ232, Ϫ58, and Ϫ50), five Sp1 sites (Ϫ570, Ϫ217, Ϫ200, Ϫ105, and Ϫ95), and an NF-Y binding site (Ϫ70).
Delineation of Functional Regions in the Mouse cdc25B Promoter by Truncation Analysis-To identify functionally relevant regions in cdc25B promoter, a series of terminal truncations was generated from the B950 construct (Ϫ950/ϩ167) and analyzed for expression in G 0 versus normally cycling cells (N) (Fig. 4). This analysis led to the following conclusions. (i) The terminal deletion of 10 nucleotides, which removes a potential E box led to an increase in transcriptional activity of ϳ40% but had no effect on cell cycle regulation. Truncation of the adjacent fragment spanning positions Ϫ980 to Ϫ768, which harbors another potential E box, had no detectable effect on transcriptional activity or cell cycle regulation. (ii) The region from Ϫ340 to Ϫ250 seems to have a negative effect on transcriptional activity. However, because no putative binding sites could be identified in this region, and there was no effect on cell cycle regulation, we did not pursue this finding. (iii) Further deletion of a fragment spanning nucleotides Ϫ250 to Ϫ223 and harboring a potential E2F site had no detectable effect. (iv) Truncation of a fragment spanning positions Ϫ223 to Ϫ180, which contains two potential Sp1 sites, led to a clear reduction in transcriptional activity. This was further decreased by truncation of the adjacent region spanning nucleotides Ϫ180 to Ϫ87, which harbors two more potential Sp1 sites. The loss of these FIG. 7. Binding of Sp1 and Sp3 to four elements of the murine cdc25B promoter. Fragments encompassing positions Ϫ103 to Ϫ80, Ϫ120 to Ϫ97, Ϫ209 to Ϫ187, and Ϫ226 to Ϫ206 were used as probes in EMSAs using NIH3T3 nuclear extract. The assay was performed in the presence and absence of antibodies specific for Sp1 (␣Sp1) or Sp3 (␣Sp3). The respective pre-immune sera did not show any effect (data not shown). Competitors (comp.) were identical to the respective probes (s, self-competition) or represented a nonspecific sequence (ns). four potential Sp1 sites led to a total decrease in transcriptional activity of 60%, with only a marginal effect on cell cycle regulation. (v) The terminal deletion of an additional 20 nucleotides resulted in a further drop in transcriptional activity but also led to a clear decrease in cell cycle regulation, indicating that this promoter region, which harbors a potential NF-Y site, is of particular functional relevance. (vi) Further truncations had no additional effect on cell cycle regulation, presumably because these constructs all lacked the NF-Y site.
Identification of Functional Upstream Elements in the Mouse cdc25B Promoter-To confirm and extend the findings obtained by promoter truncation, the putative E boxes and NF-Y binding site were altered by point mutations, and the functional consequences were analyzed in transient transfection assays. The proximal potential E2F sites were not included in this analysis because no binding of E2F-1, E2F-3, or E2F-4 to the cdc25B promoter could be detected in EMSA using either normal NIH3T3 cells or retrovirally transduced cells overexpressing the respective E2F protein (kindly provided by R. Bernards, Amsterdam), although clear binding was seen in the same assay with a bona fide E2F site from the B-myb promoter (5, 15) (data not shown).
The mutation analyses yielded the following results (Fig. 5). (i) Mutation of the most distal E box led to a slight increase in promoter activity of ϳ36% but did not show any influence on cell cycle regulation. Mutation of the second E box had only a very weak effect, and mutation of both E boxes had the same effect as mutation of the most distal one alone. These data are in line with the truncation analysis described above and indicate that the E boxes are not crucial with respect to cell cycle regulation. This promoter region was therefore not further investigated. (ii) Point mutations in the NF-Y binding site led to a drastic loss of both transcriptional activity (ϳ73%) and cell cycle regulation (57%). This result is in perfect agreement with the deletion analysis and confirms the importance of the NF-Y site both for transcriptional activity and cell cycle regulation.
Interaction of NF-Y and Sp1/Sp3 with the Mouse cdc25B Upstream Activating Sequence-To investigate protein interactions at the potential NF-Y site in the cdc25B promoter, we performed EMSAs with nuclear extracts from normally cycling NIH3T3 cells. A synthetic oligonucleotide encompassing this element was used as a probe, and competitors representing either the same site (self-competition), a bona fide NF-Y site from the MHC class II promoter (E␣-Y) (58), an Sp1 binding site (GT box), or a mutated cdc25B element (MY) were also used. As shown in Fig. 6, only the former two oligonucleotides were able to prevent the formation of a DNA-protein complex. Neither the GT box nor the mutated cdc25B element showed any competition. Likewise, no effect on complex formation was seen when binding sites for other CAAT box-binding factors, i.e. C/EBP or NF-I/CTF (59), were used (data not shown). To obtain further evidence that NF-Y interacts with the cdc25B promoter, we analyzed the effect of a monoclonal antibody (␣NF-Y A) against the A subunit of NF-Y (kindly provided by D. Mathis) (58). This antibody led to the expected supershift of the observed complex (58,59). Taken together, these data clearly suggest that the protein complex interacting with the cdc25B site is NF-Y.
Similar experiments were performed to analyze protein binding to the four functionally relevant Sp1 sites at positions Ϫ217, Ϫ200, Ϫ105, and Ϫ95. EMSAs were performed using four different probes representing these sites in conjunction with a specific (self) or nonspecific competitor (unrelated sequence) and antibodies specific for Sp1 or Sp3 (kindly provided by G. Suske, IMT, Marburg, Germany) (60,61). The data in Fig. 7 clearly show that all four sites specifically interact with Sp1 and Sp3, leading to the formation of the expected com- FIG. 9. Identification of a CCRR binding activity. A fragment encompassing positions Ϫ64 to Ϫ20 of the murine cdc25B promoter was used as a probe for EMSA using NIH3T3 nuclear extract. Four different competitors were used: s, identical to the probe; Ϫ64/Ϫ20 3mCCRR, same as probe but with three mutations in the region of the CCRE (at Ϫ32, Ϫ33, and Ϫ34); Ϫ64/Ϫ29, same as probe but lacking nine nucleotides at the 5Ј end; ns, nonspecific sequence. The uppermost band represents a specific CCRR-protein complex. The nature of the other complexes is unclear, but on the basis of the competition data these appear to be nonspecific. plexes (60).
Identification of a Proximal Repressor Element-Finally, we scanned the proximal promoter for the presence of additional sites that might play a role in cell cycle regulation. Toward this end, we introduced point mutations into this region in the context of an otherwise intact promoter fragment (Ϫ223/ϩ167 construct). Construct 2mCCRR harbors two mutations at positions Ϫ32 and Ϫ33, whereas construct m30G is mutated at position Ϫ30, i.e. the first nucleotide of the TATA motif. As shown in Fig. 8, both these mutations led to a 3-to 4-fold increased activity in G 0 cells, resulting in a 50 -60% loss in cell cycle regulation. These results indicate that this region of the promoter functions as a cell cycle-regulated repressor. Previous studies have shown that other S/G 2 genes are regulated by two contiguous repressor elements, the cell cycle-dependent element and CHR, whose function is dependent on an exact spacing relative to each other (37,39). Because the sequence surrounding the repressor element in the cdc25B promoter (TGTTATTTTTCGAATATAT; the approximate position of the repressor element is underlined) only bears a vague resemblance to a cell cycle-dependent element-CHR module (cdc25C: CT GGCGGAAGGTTTGAA; the cell cycle-dependent element and CHR are underlined), it can be concluded that these sequences are functionally unrelated. We refer to this element in the cdc25B promoter as "cell cycle-regulated repressor" (CCRR).
Protein Interaction with the CCRR-Finally, we sought to obtain direct evidence for the existence of a protein complex interacting with the CCRR. For this purpose, we performed EMSAs with a fragment containing nucleotides Ϫ64 to Ϫ20 of the murine cdc25B promoter as a probe and NIH3T3 nuclear extract. As shown in Fig. 9, the most slowly migrating complex specifically interacted with the CCRR. Whereas self-competition was highly efficient, no competition was seen with the same oligonucleotide harboring a mutation in the region of the CCRR or a 5Ј truncation of nine nucleotides. Likewise, no competition was observed with an unrelated sequence. In addition, the binding activity was not competed by B-myb or cdc25C CHR sequences (data not shown), which confirms the conclusion that the CCRR represents a functionally unrelated repressor element.
In Vivo Protection of the CCRR Region-To obtain further evidence that the CCRR represents a protein binding site, we performed genomic DMS footprinting of the region surrounding the transcriptional start site in NIH3T3 cells. Fig. 10 shows a typical in vivo footprint of the bottom strand. It is obvious that in the region of the CCRE multiple residues were protected: A at Ϫ28, A at Ϫ30, and G at Ϫ34, the two former nucleotides being part of the TATA motif. DISCUSSION The data reported in the present study suggest that the cdc25B promoter is controlled by a novel mechanism of cell cycle-regulated transcription, which involves both an NF-Y binding site and the CCRR repressor element. Neither of the two binding sites is sufficient to confer cell cycle regulation on its own, pointing to a functional interplay between the putative repressor interacting with NF-Y. Although NF-Y has been shown for a number of other promoters to play a crucial role in cell cycle-regulated transcription (37,59,(62)(63)(64)(65), its precise role has not been determined. The data presented in the present study point to a dual function of NF-Y in the context of the cdc25B promoter. NF-Y is crucial not only for promoter activation, which might be related to its described ability to recruit other transcription factors to a promoter (66), but also for cell cycle regulation. This is reminiscent of the situation described for the cdc25C promoter, where NF-Y cooperates with the cell cycle-regulated repressor CDF-1 (59). In this case, CDF-1 presumably functions by specifically repressing NF-Y-mediated activation, because the repressor function of CDF-1 is dependent on an active promoter and is specific for a small subset of transcriptional activators (67). It is possible that an analogous situation exists in the case of the cdc25B promoter, but the precise underlying mechanism remains to be investigated.
Another interesting aspect relates to the fact that the CCRE apparently overlaps the TATA motif. Although there is no formal proof at present that the putative TATA element is functional in the cdc25B promoter, its sequence (TATATAA) exactly fits that of a canonical TATA box, and its spacing relative to the transcriptional start site and the putative Inr element is within the expected range. This raises the intriguing possibility that a CCRE-interacting repressor functions by interfering with the basal transcriptional machinery, e.g. by inhibiting the assembly of a functional initiation complex. Future analyses will have to address these mechanistic questions in detail. The present study provides the basis for such studies.