Identification of a human homologue of the DREF transcription factor with a potential role in regulation of the histone H1 gene.

A human homologue (hDREF/KIAA0785) of Drosophila DREF, a transcriptional regulatory factor required for expression of genes involved in DNA replication and cell proliferation, was identified by BLAST search. Amino acid sequences corresponding to three regions highly conserved between two Drosophila species also proved to be very similar in the hDREF/KIAA0785 polypeptide. A consensus binding sequence (5'-TGTCG(C/T)GA(C/T)A) for hDREF/KIAA0785, determined by the CASTing method, overlapped with that for the Drosophila DREF (5'-TGTCGATA). We found hDREF/KIAA0785 binding sequences in the promoter regions of human genes related to cell proliferation. Analyses using a specific antibody revealed that an hDREF/KIAA0785 binds to the promoter region of the histone H1 gene. Co-transfection experiments with an hDREF/KIAA0785-expressing plasmid and a histone H1 promoter-directed luciferase reporter plasmid in HeLa cells revealed possible activation of the histone H1 promoter. Immunohistochemical analysis demonstrated that hDREF/KIAA0785 is localized in the nuclei. Although the expression level of the factor was found to be low in serum-deprived human normal fibroblasts, the amount was increased by adding serum to cultures and reached a maximum during S phase. RNA interference experiments targeting hDREF/KIAA0785 resulted in inhibition of S phase entry and reduction of histone H1 mRNA in HeLa cells. These results suggest that expression of hDREF/KIAA0785 may have a role in regulation of human genes related to cell proliferation.

Promoters of Drosophila genes related to DNA replication, such as those for the 180-kDa catalytic subunit of DNA polymerase ␣ and proliferating cell nuclear antigen (PCNA), 1 contain a common 8-bp palindromic sequence (5Ј-TATCGATA), named the DRE (DNA replication-related element) (1), in addition to E2F recognition sites (2)(3)(4)(5). Our previous studies (6,7) performed in vitro and in vivo suggested that the DRE sequence and the E2F binding sites function synergistically in activation of PCNA, DNA polymerase ␣, and dE2F genes. We found a specific DREF (DRE-binding factor) consisting of an 80-kDa polypeptide homodimer, and molecular cloning of its cDNA has allowed confirmation that DREF is a trans-activator for DREcontaining genes (8). An important role of the DRE/DREF regulatory system has been indicated by the finding that DRE/ DREF is a target of some differentiation signals. The zen (zerknullt) gene encoding a homeodomain-containing protein, Zen, which is expressed in the dorsal region of the early embryo at the cellular blastoderm stage, is involved in differentiation of the amnioserosa and the optic lobe (9). Zen expression in cultured cells results in repression of DRE-containing genes by reducing the DREF activity (10). Thus, the DRE/DREF system may occupy a cross-road position in growth and differentiation signaling pathways.
Recently, Hart et al. (11,12) proposed a novel function of DREF as an antagonist of the boundary element-associated factor (BEAF), which is involved in the boundary activity of the scsЈ region of the Drosophila 87A7 hsp70 gene (5,11,12). Staining of polytene chromosomes with anti-DREF and anti-BEAF antibodies revealed about 50% of signals for the two proteins to overlap. Furthermore, using a chromatin precipitation method, they demonstrated that DREF could bind to the same sequences as BEAF. From these results, they hypothesized that competition of binding between DREF and BEAF is important for the regulation of activity at the chromatin boundary.
More recently, we have established transgenic flies, in which ectopic expression of DREF was targeted to the eye imaginal discs (13). Adult flies expressing DREF exhibited a severe rough eye phenotype. We found that this was significantly suppressed by a half-dose reduction of any of the trithoraxgroup genes, brahma (14), moira (15), and osa (16), involved in determining chromatin structure or chromatin remodeling, whereas reduction of Distal-less, a transcription factor involved in proximal/distal pattern formation (17), enhanced the DREFinduced rough eye phenotype (18). These results suggest a possibility that DREF activity is regulated by protein complexes that play a role in determining chromatin structure (for example, establishment, maintenance, or cancellation of the chromatin boundary) such as those involving BEAF (5,11,12) and Polycomb/trithorax group proteins (19).
Despite its obvious potential importance, identification of a mammalian DRE/DREF system has not yet been achieved. To obtain clues for cDNA cloning of mammalian DREF, we isolated a gene for DREF from Drosophila virilis (13) and determined highly conserved regions in DREFs of this species and Drosophila melanogaster (Dm) (8). Comparison of deduced amino acid sequences for the two species allowed us to identify three highly conserved regions, CR1, CR2, and CR3 (13). A BLAST search with the amino acid sequence of DmCR1, a domain required for DNA binding and homodimer formation (8), hit a candidate human DREF protein registered to data bases as KIAA0785 (20). Comparison of amino acid sequences of KIAA0785 protein and the two Drosophila DREFs revealed obvious conservation of CR1, CR2, and CR3. In the present study, we characterized KIAA0785 protein as a human homologue of DREF (hDREF) and demonstrated sequence-specific DNA binding activity and fluctuation of expression during the cell cycle. Furthermore, we found that hDREF/KIAA0785 binds to the promoter region of the histone H1 gene and stimulates its promoter activity.

EXPERIMENTAL PROCEDURES
Cell Culture-HeLa cells were grown in Dulbecco's modified Eagle's medium with 10% fetal calf serum. Primary human embryonic lung fibroblast (HEL) cells were maintained at 37°C in 5% CO 2 and Dulbecco's modified Eagle's medium with 10% fetal calf serum. For synchronization, exponentially growing HEL cells were starved of serum for 72 h prior to addition of fresh medium with 10% fetal calf serum.
A full-length cDNA for hDREF/KIAA0785 was generated by PCR using primers hDREF5Ј, 5Ј-GGAgctagcATGGAGAATAAAAGCCTG-GAG, and hDREF3Ј, 5Ј-GCTctcgagCTACAGGAAGCTGCTGTCCCT. Recognition sites for NheI in hDREF5Ј and XhoI in hDREF3Ј are shown by lowercase letters with underlining.
Plasmid Construction-KIAA0785 cDNA/pBluescript ks(Ϫ) was obtained from Kazusa DNA Research Institute (20) and used as a PCR template to generate a full-length cDNA for hDREF/KIAA0785. The obtained cDNA was blunt-ended with a DNA blunting kit (Takara) and inserted into the SmaI site of pGEX-2T (Amersham Biosciences) to create pGST-hDREF/KIAA0785. A plasmid expressing hDREF/ KIAA0785 was constructed by amplifying the full-length cDNA of hDREF/KIAA0785 by the PCR method using KIAA0785 cDNA/ pBluescript ks(Ϫ) as a template, cutting with NheI and XhoI and inserting into the NheI-XhoI site of pcDNA3-HA. To construct the expression plasmid CR1hDREF/KIAA0785/pcDNA3-HA, cDNA encoding the CR1 region (1-140 amino acid residue) of hDREF/KIAA0785 was amplified by PCR and inserted into the blunt-ended XhoI site of pcDNA3-HA.
The H1-p/pGL3 reporter plasmid was constructed as follows. The promoter region (Ϫ537 to Ϫ1) of the human histone H1 gene (21) was amplified by PCR using genomic DNA from HeLa cells as template, cut with NheI and XhoI, and inserted into the NheI-XhoI site of pGL3 promoter vector (Promega), which contains the SV40 promoter placed upstream of the firefly luciferase gene. Then, the SV40 promoter was removed by cutting with HindIII and BglII. The obtained DNA fragment was blunt-ended with a DNA blunting kit (Takara) and selfligated. All plasmids were propagated in Escherichia coli XL-1 Blue, isolated by standard procedures (23) and further purified using a Qiagen Plasmid Midi kit (Qiagen Inc.).
Expression of GST Fusion Proteins-Expression and purification of GST-hDREF/KIAA0785 fusion proteins in E. coli XL-1 Blue were carried out as described elsewhere (8). Lysates of cells were prepared by sonication in Buffer D containing 0.6 M KCl, 1 mM phenylmethylsulfonyl fluoride, 1 g/ml each of pepstatin, leupeptin, and aprotinin, and cleared by centrifugation at 12,000 ϫ g for 20 min at 4°C before application to a glutathione-Sepharose (Amersham Biosciences) column to purify the GST-hDREF/KIAA0785 fusion protein. Control GST protein was expressed and purified in the same way.
Antibody-The purified GST-hDREF/KIAA0785 fusion protein was used to elicit polyclonal antibody production in a rabbit. Polyclonal antibodies reacting with GST-hDREF/KIAA0785 were purified by sequential passage through affinity resin covalently conjugated with GST or GST-hDREF/KIAA0785.
Immunoblotting-Polypeptides in E. coli cell extracts or total cell extracts were separated by SDS-PAGE and transferred to polyvinylidene difluoride membranes in a solution containing 50 mM borate-NaOH buffer (pH 9.0) and 20% methanol at 4°C for 16 h. Membranes were blocked with TBS (50 mM Tris-HCl (pH 8.3) and 150 mM NaCl) containing 10% skim milk for 1 h at room temperature and then incubated with the rabbit anti-hDREF/KIAA0785 polyclonal antibody for 1 h at room temperature. After extensive washing with TBS, membranes were incubated with an alkaline phosphatase-conjugated goat anti-rabbit IgG (used at a dilution of 1:2,000) (Promega) for 1 h at room temperature. After further extensive washing with TBS, color was developed in a solution containing 100 mM Tris-HCl (pH 9.5), 100 mM NaCl, 5 mM MgCl 2 , 0.34 mg/ml nitro blue tetrazolium salt, and 0.175 mg/ml 5-bromo-4-chloro-3-indoryl phosphate toluidinium salt.
Cell Cycle Analysis with Flow Cytometry-Cells were harvested by trypsinization, and the nuclei were stained with propidium iodide using a CycleTest kit (BD Biosciences) according to the manufacturer's instructions. Analysis was with a FACScan (BD Biosciences), and the percentages in each phase were calculated with the SOBR model in the CELFIT program.
DNase I digestion of Triton X-100-treated nuclei was performed as follows. Triton X-100-extracted nucleus suspension was prepared as described above. The samples (100 l) were supplemented with 1 mM ATP and 1,000 units/ml DNase I, followed by incubation for 1 h at room temperature. To examine the extent of hDREF/KIAA0785 solubilization by salt, Triton X-100-treated nuclei (100 l) were suspended in CSK containing NaCl at concentrations of 0.15, 0.35, or 0.55 M, followed by incubation for 1 h on ice. The pellet fractions and supernatants were then separated by low speed centrifugation, the former being resuspended in 100 l of the buffer, and the latter was clarified by centrifugation at 100,000 rpm for 15 min using a Beckman TLA100.1 rotor. The samples were then added to the same volumes of 2ϫ SDS sample buffer (1ϫ solution (62.5 mM Tris-HCl (pH 6.8), 2% SDS, 5% ␤-mercaptoethanol, 10% glycerol, 0.01% bromphenol blue)), boiled, and analyzed by immunoblotting.
Immunofluorescence Analysis-Cells grown on 13-mm coverglasses were fixed for 20 min in 100% methanol at Ϫ20°C and then permeabi-lized with 0.2% Triton X-100 for 2 min. All staining procedures were carried out at room temperature. The samples were incubated with either anti-hDREF/KIAA0785 IgG or normal rabbit IgG at a 1:400 dilution in PBS with 10% normal goat serum for 1 h. After washing three times with PBS, the samples were reacted with Alexa 488-conjugated goat anti-rabbit IgG antibody (Molecular Probes) at a 1:400 dilution in PBS with 10% normal goat serum for 1 h. After washing a further three times with PBS, the samples were mounted and analyzed by conventional microscopy.
Selection of hDREF/KIAA0785 Binding Sites with CASTing Experiments-A 32 P-labeled double-stranded degenerate oligonucleotide was prepared by incubating 50 l of reaction mixture containing 10 mM Tris-HCl (pH 9.0), 50 mM KCl, 1.5 mM MgCl 2 , 0.1% Triton X-100, 50 M each of dATP, dGTP, and dTTP, 20 M [␣-32 P]dCTP (740 kBq), 100 M each of primers F and R76 oligonucleotide, and 2 units of the Klenow fragment of DNA polymerase I at 37°C for 1 h. Binding reactions were performed by adding GST-hDREF/KIAA0785 (0.4 g) to Buffer D (20 mM Hepes (pH 7.9), 120 mM KCl, 1 mM DTT, 0.1% Tween 80, and 12% glycerol) containing 400 ng of poly(dI-dC), 400 ng of sonicated salmon sperm DNA, 50 g of bovine serum albumin, and 32 P-labeled doublestranded degenerate oligonucleotide (1.8 ng) and incubating at 4°C for 30 min. Then, glutathione-Sepharose beads (10 l) were added with incubation at 4°C for an additional 1 h. The DNA-protein complexes were then precipitated by centrifugation, and the pellets were washed six times with Buffer D. After elution with 100 l of Buffer D containing 5 mM reduced glutathione, the oligonucleotides were recovered by phenol extraction and ethanol precipitation and amplified by PCR in 20 l of a mixture containing 300 ng of primers F and R, 20 M [␣-32 P]dCTP (740 kBq), 50 M each of dATP, dGTP, and dTTP, 10 mM Tris-HCl, 50 mM KCl, 1.5 mM MgCl 2 , 0.1% Triton X-100, and 0.5 unit of Taq polymerase (Roche Molecular Biochemicals). DNAs were amplified by 20 cycles of 1 min at 94°C, 30 s at 60°C, and 1 min at 72°C, purified by passage through a Sephacryl-200 spin column, precipitated with ethanol and dissolved in 50 l of a buffer containing 10 mM Tris-HCl (pH 8.0) and 1 mM EDTA. The amount of amplified oligonucleotide was quantified, and a portion corresponding to 2 ng was used in subsequent CASTing cycles. After five cycles of CASTing, the radiolabeled oligonucleotides (1.5 ϫ 10 5 cpm) were used as probes in an electrophoretic mobility shift assay (EMSA). Oligonucleotides binding to GST-hDREF/ KIAA0785 were excised from gels, eluted overnight in 0.2 ml of a solution containing 0.2 M NaCl, 20 mM EDTA, and 0.1% SDS at 37°C, extracted once with phenol, and then precipitated with ethanol.
DNAs were amplified by PCR as described above, digested with EcoRI and BamHI, and then subcloned into EcoRI and BamHI sites of the pBluescript vector. Nucleotide sequences of 50 independent clones were determined.
Preparation of HeLa Cell Nuclear Extract-A nuclear extract was prepared by a modified Dignam method (25). Cultured HeLa cells in suspension were collected by centrifugation and washed with PBS. Two packed volumes of Buffer A containing 10 mM Hepes (pH 7.9), 1.5 mM MgCl 2 , 10 mM KCl, 0.5 mM DTT, and proteinase inhibitor mixture (1 mM phenylmethylsulfonyl fluoride, 5 g/ml leupeptin, 2 g/ml pepstatin, 0.75 g/ml aprotinin) were added to swell the cells on ice for 10 min. A Dounce homogenizer (size B) was then used to rupture the cell membranes, and nuclei were collected at 1,600 ϫ g at 4°C and resuspended in high salt Buffer B containing 20 mM Hepes (pH 7.9), 25% glycerol, 0.42 M NaCl, 1.5 mM MgCl 2 , 0.2 mM EDTA, 0.5 mM DTT, and the proteinase inhibitor mixture to extract nuclear proteins by rocking at 4°C for 1 h before centrifugation at 100,000 ϫ g for 5 min. The supernatant was dialyzed in Buffer C (20 mM Hepes (pH 7.9), 50 mM NaCl, 0.5 mM DTT, 0.5 mM phenylmethylsulfonyl fluoride, 20% glycerol), clarified by centrifugation, and stored at Ϫ80°C in small aliquots. The protein concentration of the extract was 4.8 mg/ml.
Electrophoretic Mobility Shift Assay-EMSA were performed as described earlier (1), with minor modifications. A reaction mixture containing 20 mM Hepes (pH 7.6), 150 mM KCl, 0.1 mM EDTA, 0.5 mM DTT, 10% glycerol, and 1 g of salmon sperm DNA was used with purified GST-hDREF/KIAA0785. Another reaction mixture containing 15 mM Hepes (pH 7.6), 120 mM KCl, 0.1 mM EDTA, 1 mM DTT, 2% glycerol, and 1 g of poly(dI-dC) DNA was employed for EMSA with the HeLa cell nuclear extracts. 32 P-Labeled probes (10,000 cpm) were incubated in 15 l of reaction mixture. When necessary, unlabeled DNA fragments were added as competitors at this step. Then, GST-hDREF/KIAA0785 fusion proteins (20 ng) or aliquots of the HeLa cell nuclear extract (5 g protein) were added, and the reaction mixture was incubated for 15 min on ice. In experiments with antibodies, the HeLa cell nuclear extract was preincubated with the antibody for 1 h on ice. DNA-protein complexes were electrophoretically resolved on 4% polyacrylamide gels in 100 mM Tris-borate (pH 8.3), 2 mM EDTA containing 2.5% glycerol at 25°C. The gels were dried and then autoradiographed.
DNA Transfection into Cells and Luciferase Assays-HeLa cells were plated at about 6 ϫ 10 4 cells per well in 24-well culture plates for 16 h. 100 ng of firefly luciferase reporter plasmid and 10 ng of pRL-TK plasmid (Promega), carrying sea pansy luciferase under the control of the herpes simplex virus tk gene promoter as an internal control, were co-transfected into cells using LipofectAMINE Plus reagent (Invitrogen) (26). In the co-transfection experiment with the expression plasmid, reporter plasmids were co-transfected with hDREF/KIAA0785/ pcDNA3-HA or CR1hDREF/KIAA0785/pcDNA3-HA with pRL-TK as an internal control. The total amount of expression plasmid was adjusted to 410 ng by addition of pcDNA3-HA. After DNA transfection, cells were cultured for 48 h, and luciferase activity was measured with a dualluciferase reporter assay system (Promega). Firefly luciferase activity was normalized to sea pansy luciferase activity. Transfections were performed several times with at least three independent plasmid preparations.
RT-PCR Analysis-Total RNA was isolated using RNeasy Mini Kit (Qiagen Inc.), and 250 ng of total RNA was subjected to RT-PCR analyses using a high fidelity RNA PCR kit (Takara). The cDNAs were generated using random 9-mers as primers. The PCR condition was 45-50 cycles at 94°C for 1 min, 53°C for 1 min, and 72°C for 1 min using 5Ј-H1 and 3Ј-H1 as primers. All the PCR were performed within the range of linear amplification. The PCR products were electrophoretically separated in a 2% agarose gel, stained with ethidium bromide, and quantified using ImageMaster VDS-CL (Amersham Biosciences).
siRNA Transfection-siRNAs (short interfering double-stranded RNAs) against hDREF/KIAA0785 were chemically synthesized by Dharmacon Research, Inc. The sequences for hDREFsiRNA1 and hDREFsiRNA2 correspond to regions ϩ221 to ϩ241 and ϩ132 to ϩ152 with respect to the translational start codon, respectively. Scramble siRNA (Dharmacon Research, Inc.) was used as a negative control. siRNAs were transfected into 30% confluent HeLa cells using Oligofectamine (Invitrogen) (27), and cells were incubated for 72 h at 37°C.

5-Bromo-2Ј-deoxyuridine (BrdUrd)
Labeling-HeLa cells were incubated in the presence of 20 g/ml BrdUrd (Roche Molecular Biochemicals) for 1 h. The samples were fixed in methanol for 1 h at Ϫ20°C and further fixed in 80% ethanol-50 mM glycine buffer (pH 2.0) at Ϫ20°C for 2 h. Incorporated BrdUrd was visualized using an anti-BrdUrd antibody and an alkaline phosphatase detection kit (Roche Molecular Biochemicals). The time of color development for alkaline phosphatase was precisely regulated to be identical for all samples.

RESULTS
Comparison of Amino Acid Sequences between hDREF/ KIAA0785 and Drosophila DREF-Previously, we isolated genes for DREFs from D. melanogaster (DmDREF; 709 amino acids) and D. virilis (D. virilis DREF; 742 amino acids) (8,13). Comparison of the deduced amino acid sequences allowed us to identify three highly conserved regions, CR1, CR2, and CR3. We have already demonstrated that CR1 is a domain required for DNA binding and homodimer formation by biochemical analysis of the DmDREF polypeptide (8). In addition, we have also provided evidence that CR1 might be important for Dm-DREF function using transgenic flies expressing this region under the control of the GAL4-UAS system (28). Thus, we performed a BLAST search using the amino acid sequence for DmCR1 as a query. As a result, significant similarities were found by a search of the GenBank TM and EMBL data bases using the tblastx program, particularly with human KIAA0785 (AB018328; 694 amino acids) (20). The KIAA0785 polypeptide shows significant identity (21.3%) and similarity (41.0%) to DmDREF in total. Especially, the regions corresponding to CR1 (27.7% identity and 48.1% similarity), CR2 (29.2% and 46.1%), and CR3 (21.1% and 51.9%) are highly conserved (Fig.  1A), suggesting that KIAA0785 might be a human homologue of Dm and D. virilis DREFs. It is therefore referred to here as hDREF/KIAA0785.
Interestingly, significant similarities with other proteins were also found by the BLAST search using the DmCR1 amino acid sequence as a query. As shown in Fig. 1B, KIAA0637 (AB014537) polypeptide (1171 amino acids) contains four CR1like sequences (29) and the N-terminal DNA binding region of the Drosophila BEAF32B (283 amino acids) (11). Alignment of amino acid sequences of the polypeptides revealed complete conservation of cysteine and histidine residues (shown by asterisks) and high conservation of surrounding amino acid residues, suggesting DNA binding activity in the N-terminal region containing C2H2-type zinc finger structure.
It should be noted that Esposito et al. (30) isolated a cDNA and the gene encoding KIAA0785 protein and called it Tramp. In the meantime we obtained KIAA0785 cDNA from the Kazusa DNA Research Institute and started functional analysis. They reported the Tramp gene to be localized on the X and Y chromosomes, and the amino acid sequence of the Tramp protein shows similarity with Drosophila DREF, although they did not demonstrate any biological activity.
hDREF/KIAA0785 Protein Is Localized in the Nuclei of HeLa Cells-Hart et al. and ourselves (8,12) have demonstrated that DmDREF is a ubiquitous nuclear protein. We raised a rabbit polyclonal antibody against bacterially produced GST-fused-hDREF/KIAA0785 protein. IgGs specifically recognizing hDREF/KIAA0785 protein were purified by serial passage of rabbit serum through affinity columns coupled with GST and GST-hDREF/KIAA0785 proteins. Immunoblot analysis revealed the antibody to react specifically with an 80-kDa polypeptide in a HeLa cell extract (Fig. 2A, lane 5). Preincubation of the antibody with GST-hDREF/KIAA0785 resulted in no detection (data not shown).
To investigate the subcellular localization of hDREF/ KIAA0785 protein, we immunostained HeLa cells. Signals for hDREF/KIAA0785 proteins were detected in nuclei (Fig. 2B,  panel a). To clarify whether hDREF/KIAA0785 is associated with a specific nuclear structure, HeLa cells treated with 0.2% Triton X-100 were immunostained. Significant amounts of fluorescence were then found in granular structures (Fig. 2B,  panel b), and the signals were resistant to treatment with 50 g/ml pancreatic DNase I (data not shown), suggesting that at least a part of hDREF/KIAA0785 may be tightly bound to nuclear structures. Subcellular localization of hDREF/KIAA0785 was also investigated with biochemical cell fractionation of HeLa cells into cytosolic and nuclear fractions by treatment with hypotonic buffer and low speed centrifugation. As shown in Fig. 2C, hDREF/KIAA0785 protein was detected in both cytosolic (lane 2) and nuclear (lane 3) fractions but was mainly localized in the nuclear fraction (lane 3). Subnuclear localization of hDREF/ KIAA0785 protein was further examined with use of detergent, salt, and nuclease. HeLa cells were first treated with CSK buffer containing 0.5% Triton X-100 and 0.15 M NaCl, which extracts not only cytoplasmic but also nuclear proteins not tightly bound to nuclear structures, and then the residual nuclei were treated with DNase I and/or various concentrations of NaCl. DNase treatment did not liberate hDREF/KIAA0785 from nuclei (Fig. 2C, lanes 4 -7), although DNase treatment under this condition liberated about two-thirds of core histones and DNA from the extracted nuclei (31). NaCl treatment at 0.15 M did not release hDREF/KIAA0785, but two-thirds and almost all of hDREF/KIAA0785 were released from nuclei in the presence of 0.35 and 0.55 M NaCl, respectively (Fig. 2C,  lanes 10 and 11). These results indicate that hDREF/KIAA0785 protein is associated with granular structures in a detergentand nuclease-resistant manner.
Determination of Consensus Sequences for hDREF/ KIAA0785 Binding-As described above, hDREF/KIAA0785 protein is localized in nuclei, like DmDREF (8). Therefore, we addressed the question of whether hDREF/KIAA0785 protein can directly bind to DNA and if so to which nucleotide sequence. For this purpose, we performed a series of CASTing experiments with a double-stranded oligonucleotide with defined ends for which PCR primers were available and a degenerate central core region of 26 nucleotides. Purified GST-hDREF/KIAA0785 fusion protein was incubated with degenerate double-stranded oligonucleotides and affinity-purified with glutathione-Sepharose beads. Unbound DNA was removed by washing, and bound DNA was amplified by PCR. Several cycles of affinity purification and PCR amplification were performed. Obtained DNA segments were cloned into pBluescript, and 50 independent isolates were sequenced. The relevant features of the determined sequences of 10 bp are shown in Fig. 3A. The consensus nucleotide sequence for hDREF/KIAA0785 protein is a palindromic 5Ј-TGTCG(C/) TGA(C/T)A. Notably, the half-site of this sequence (5Ј-(C/ T)GA(C/T)A) matches five of eight nucleotides of the DmDREF binding sequence (5Ј-TATCGATA) (8) and the main recognition motif of BEAF (5Ј-CGATA) (11) (Fig. 3B).
Next we tried to detect hDREF/KIAA0785 binding activity in HeLa cell nuclear extracts by EMSA. Oligonucleotides (hDRE) containing the 10-bp palindromic sequence (5Ј-TGTCGC-GACA) preferentially recognized by GST-hDREF/KIAA0785 protein was synthesized and used for EMSA as a probe (Fig.  4A). As shown in Fig. 4B, two protein-DNA complexes were formed with HeLa cell nuclear extracts (lanes 1 and 7). Therefore, we examined whether these shifted bands are complexes containing hDREF/KIAA0785 protein by EMSA experiments with anti-hDREF/KIAA0785 antibody. Preincubation of HeLa cell nuclear extracts with the rabbit anti-hDREF/KIAA0785 antibody eliminated the signals in a dose-dependent manner (Fig. 4B, lanes 8 -12), whereas addition of normal rabbit IgG had no effect (Fig. 4B, lanes 2-6). The result indicates that these two DNA-protein complexes formed between hDRE oligonucleotides and HeLa cell nuclear extract contain hDREF/ KIAA0785 protein, although it is not clear whether there are qualitative differences and what are the components of these two kinds of DNA-protein complexes.
Given the preference for the 10-bp nucleotide sequence selected by CASTing using GST-hDREF/KIAA0785, we further examined the nucleotide sequence sufficient for binding using HeLa cell nuclear extract. A series of oligonucleotides with base substitutions inside the hDREF/KIAA0785 binding site (mut1-mut10) were synthesized and used for EMSA as competitors (Fig. 4A). As shown in Fig. 4C, the two shifted bands were diminished by adding an excess amount of unlabeled oligonucleotides of wild-type (hDRE) (lanes 2 and 3) and mut6 with a single base substitution at the ninth C to T of the hDREF/ KIAA0785 binding sequence (lanes 14 and 15). The mut2, mut3, and mut4 with two base substitution mutations in the center region of the hDREF/KIAA0785 binding site did not compete for the binding at all (lanes 6 -11), and excess amounts of mut7 (lanes 16 and 17) and mut10 (lanes 22 and 23), with base substitutions inside the binding site, slightly competed, whereas the other mutations in the end of the binding site (mut1, mut5, mut8, and mut9) retained competition ability (lanes 4, 5, 12, 13, 16, 17, 20, and 21). These results indicate that the 6-bp sequence 5Ј-TCG(C/T)GA in the center of the hDREF/KIAA0785 binding site plays an important role in hDREF/KIAA0785 binding.
Search for Genes Carrying hDREF/KIAA0785 Binding Sites-To identify genes that might be under the control of hDREF/KIAA0785 protein, we searched the GenBank TM and EMBL data bases using BLAST and the TargetFinder network service (Telethon Institute of Genetics and Medicine). The simple BLAST program revealed that there exist more than 500 TGTCG(C/T)GA(T/C)A-like sequences (counting sequences with more than 7 bp of the 10-bp consensus sequence) in the human genome, and results for selected examples involving promoter regions are listed in Table I. Interestingly, we found hDREF/KIAA0785 binding sequences in a variety of genes related to cell proliferation, as reported for DmDREF (32). We categorized these genes by its function as follows: DNA replication (DNA synthesis), DNA repair, cell cycle regulation, transcription, regulation of chromatin structure, and protein synthesis. It should be noted that we also found DRE in the promoter regions of Drosophila genes homologous to those for human topoisomerase II␣ (33), DNA polymerase ␥ (34), c-Myb (35), rRNA (36), TGF-␤ (37), and K-ras (38). The findings suggest that hDREF/KIAA0785 might be a functional homologue of DmDREF.
The Promoter Region of the Histone H1 Gene Specifically Forms a Complex with hDREF/KIAA0785 Protein-Because we demonstrated that DmDREF regulates the expression of Drosophila genes related to DNA replication and cell proliferation (1, 7, 39 -41), we examined whether hDREF/KIAA0785 regulates human DNA replication-related genes. We focused on the histone H1 gene (referred to as FNC16/H1.5), because a single 10-bp sequence completely matching the hDREF/ KIAA0785 binding sequence was found in its promoter region, which has been cloned and for which several transcriptional regulatory elements have been characterized (21,22).
The hDREF/KIAA0785 binding sequence is positioned at Ϫ390 to Ϫ381 with respect to the translation start codon. We first examined whether the hDREF/KIAA0785 protein binds. Oligonucleotides containing the putative binding sequence in the histone H1 promoter (H1) or with base substitutions inside the hDREF/KIAA0785 binding site (H1m) were chemically synthesized and used for EMSA (Fig. 5A). As shown in Fig. 5B, four protein-DNA complexes were formed with the HeLa cell nuclear extract and the radiolabeled H1 as a probe (lane 1). Signals for the slowest and the second slowest migrating bands were diminished by adding excess amount of unlabeled H1 oligonucleotide (Fig. 5B, lanes 2 and 3) or oligonucleotides containing the consensus hDREF/KIAA0785 binding sequence (hDRE and mut6) (Fig. 5B, lanes 6 -9) but not by adding oligonucleotide H1m (Fig. 5B, lanes 4 and 5). Furthermore, the signals of the two retarded bands were diminished by preincubation with the rabbit anti-hDREF/KIAA0785 antibody in a dose-dependent manner (Fig. 5C, lanes 8 -12), whereas addition of normal rabbit IgG had no effect (Fig. 5C, lanes 2-6).
hDREF/KIAA0785 Protein Stimulates Histone H1 Gene Promoter Activity-To address the function of hDREF/KIAA0785 with regard to transcription of the histone H1 gene, we examined its effects on promoter activity. A DNA region containing the histone H1 promoter (Ϫ537 to Ϫ1) was amplified by PCR using genomic DNA from HeLa cells and cloned into a plasmid carrying the firefly luciferase reporter gene (H1-p/pGL3). Transient co-expression experiments with H1-p/pGL3 reporter plas- FIG. 4. Detection of a specific complex with the hDREF/KIAA0785 recognition site. A, nucleotide sequences containing the hDREF/ KIAA0785 recognition site (hDRE) and its base substitution mutants (mut1-mut10) used for EMSA. The hDREF/KIAA0785 binding sequence determined by CASTing (Fig. 3) and its base substitution sequences are boxed. Palindromic sequences are indicated by arrows. B and C, EMSA using 32 P-radiolabeled hDRE oligonucleotide as a probe. B, radiolabeled hDRE double-stranded oligonucleotide was incubated with HeLa nuclear extract and increasing amounts of normal rabbit IgG (lanes 2-6) or anti-hDREF/KIAA0785 IgG (lanes 8 -12). Lanes 1 and 7, no antibody. C, radiolabeled hDRE oligonucleotide was incubated with HeLa nuclear extract and various competitors as shown in A. The samples were electrophoretically separated in a 4% polyacrylamide gel and then the gel was dried and autoradiographed. mid and an hDREF/KIAA0785-bearing plasmid revealed that expression of the latter stimulated the H1 promoter activity in a dose-dependent manner (Fig. 6A). In contrast, expression of a deletion mutant for hDREF/KIAA0785 encoding N-terminal amino acid residues corresponding to the CR1 region did not result in activation (Fig. 6B). The results indicate that hDREF/ KIAA0785 protein positively regulates the histone H1 gene promoter.
Expression of hDREF/KIAA0785 Protein Is Induced in G 1 -S Phase of the Cell Cycle-Because transcription of the histone  H1 gene is transiently induced during G 1 -S phase (42,43), we examined fluctuation of hDREF/KIAA0785 protein during the cell cycle by immuno-Western blotting. We used primary cultures of normal human lung fibroblasts (HEL cells) for this analysis, because they can be readily induced to enter a quiescent state by serum deprivation, re-entering the cell cycle on addition of 10% fetal calf serum. We checked cell cycle progression after release from serum starvation by staining cells with propidium iodine and applying to flow cytometry (Fig. 7A) and by staining cells with anti-PCNA antibody, which is known as a G 1 -S phase marker (Fig. 7C) (44). Both analyses revealed that more than 85% of HEL cells arrested by 72 h of serum starvation synchronously re-enter the cell cycle after adding serum to the culture. As shown in Fig. 7, the total amount of hDREF/ KIAA0785 protein in quiescent HEL cells after 48 h of serum starvation was significantly reduced in comparison with that of asynchronous cultures of HEL cells (Fig. 7B, a). Levels began to increase in cells at 14 h after serum stimulation, reaching a maximum in S/G 2 -cell enriched conditions after 16 -18 h, and then gradually decreasing (Fig. 7B, a). Immunostaining of HEL cells also showed that expression of hDREF/KIAA0785 protein is regulated in line with the proliferating stage. Decrease in the amount of nuclear fluorescence with the anti-hDREF/ KIAA0785 antibody was observed in quiescent cells in serumdeprived culture. Release from serum starvation induced the reappearance of strong fluorescence within 14 h (data not shown). As shown in panel c of Fig. 7C, more than 70% of cells were then stained with hDREF/KIAA0785 antibody, and this expression pattern was similar to those of PCNA (Fig. 7C,  panel d). The period of induction of hDREF/KIAA0785 expression after serum stimulation seems to correspond well with the G 1 -S phase in the cell cycle (Fig. 7, A and B, a). The results thus indicate that expression of hDREF/KIAA0785 is induced during the G 1 -S transition.
The expression pattern of the histone H1 gene after serum stimulation was also examined by RT-PCR analysis. As reported previously (59), the signal for histone H1 mRNA was not detected in serum-starved cells. Expression of histone H1 mRNA was induced at 14 h after serum release, and the expression level reached a maximum at 18 -20 h (Fig. 7B, b). Fluctuation of histone H1 mRNA was quite similar to that of hDREF/KIAA0785 expression.

Reduction of hDREF/KIAA0785 Expression Caused Repression of DNA Synthesis and the Histone H1 Gene Expression-
The above results suggest a possibility that hDREF/KIAA0785 protein might play a role in G 1 -S progression. To address this question, we performed RNA interference targeting endogenous hDREF/KIAA0785 in HeLa cells. siRNAs against hDREF/ KIAA0785 (hDREFsiRNA1 and hDREFsiRNA2) were transfected, and the expression amounts of hDREF/KIAA0785 protein were determined. Immunoblot analysis revealed that amounts of hDREF/KIAA0785 protein were reduced to 12 and 70% by transfection of hDREFsiRNA1 and hDREFsiRNA2, respectively, relative to Scramble siRNA transfection (Fig. 8A,  a). Immunofluorescence staining also demonstrated that both hDREFsiRNA1 and hDREFsiRNA2 specifically reduced signals for hDREF/KIAA0785 protein in nuclei (Fig. 8A, panels  b-e). Under this condition, expression levels of histone H1 mRNA were measured by quantitative RT-PCR analysis. Introduction of hDREFsiRNA1 and hDREFsiRNA2 decreased amounts of histone H1 mRNA by 74 and 32% compared with Scramble siRNA transfection, respectively (Fig. 8B). Evidence strongly suggests that transcription of the histone H1 gene might be under the control of hDREF/KIAA0785.
Next we examined the effect of hDREF/KIAA0785 reduction on G 1 -S progression using the BrdUrd labeling method. hDREFsiRNAs-transfected cells failed to stain with BrdUrd (Fig. 8C, panels c and d), whereas about 30% of control cells incorporated BrdUrd (Fig. 8C, panels a and b). Noticeably, we observed that not a few cells transfected with hDREFsiRNAs exhibited flattened and enlarged morphology (Fig. 8, A (panels  d and e) and C (panels c and d)) and began to die after 5 days after transfection (date not shown). Thus, we concluded that hDREF/KIAA0785 might play an important role in G 1 -S progression.

DISCUSSION
In the present study, we identified KIAA0785 as a human DREF homologue based on conservation of amino acid sequences corresponding to CR1, CR2, and CR3 of Drosophila's DREFs. Complete conservation of cysteine and histidine residues and significant conservation of surrounding amino acid residues in C2H2-type zinc finger DNA binding domain suggest that KIAA0785 may recognize similar DNA sequence as Dm-DREF. Therefore, we determined the hDREF/KIAA0785 binding sequence in vitro by the CASTing method using recombi- FIG. 5. hDREF/KIAA0785 protein-specific binding to a recognition site in the histone H1 gene promoter. A, nucleotide sequences containing the hDREF/KIAA0785 recognition site (H1) and its base substitution mutant (H1m) in the histone H1 gene promote used for EMSA. The hDREF/KIAA0785 binding sequence and its base substitution sequences are boxed. B and C, EMSA using 32 P-radiolabeled H1 oligonucleotide as a probe. B, radiolabeled H1 oligonucleotide was incubated with HeLa nuclear extract and various competitors as shown in A and Fig. 4A. C, radiolabeled H1 double-stranded oligonucleotides was incubated with HeLa nuclear extract and increasing amounts of normal rabbit IgG (lanes 2-6) or anti-hDREF/KIAA0785 IgG (lanes 8 -12). Lanes 1 and 7, no antibody. The samples were electrophoretically separated in a 4% polyacrylamide gel and then the gel was dried and autoradiographed. nant hDREF/KIAA0785 protein. The sequence that appears to have the highest affinity to hDREF/KIAA0785 is 5Ј-TGTCG(C/ T)GA(C/T)A, and a half-site of this sequence matches 5 of the 8-bp consensus sequence for DmDREF.
We further detected hDREF/KIAA0785 DNA binding activity using HeLa cell nuclear extracts and determined the required nucleotide sequences by EMSA with oligonucleotides carrying a series of base substitution mutations as competitors.
The results indicate that the binding sequence with the highest affinity to hDREF/KIAA0785 protein is 5Ј-TGTCG(C/T)GA (C/T)A, and particularly the 6-bp sequence in the center (5Ј-TCG(C/T)GA) might be important for hDREF/KIAA0785 binding.
It is noteworthy that the consensus binding sequence for hDREF/KIAA0785 contains a CGCG sequence in its center, this being frequent in the promoter regions of ubiquitous-expressing house-keeping genes in the mammalian genome (45). Generally, non-methylated CpG is mainly found in CpG islands of promoter regions (45). Methylation prevents methylationsensitive transcription factors such as E2F from binding and strongly suppresses gene expression (46,47). In contrast, nonsensitive transcription factors such as Sp1 can bind to methylated CpG islands and activate transcription (48,49). We obtained results indicating that hDREF/KIAA0785 can bind to TGTCGCGACA and its methylated form, TGTmCGmCGACA, with almost the same affinity 2 so that it is thus likely that hDREF/KIAA0785 can regulate a wide variety of genes containing both hypo-and hypermethylated CpG islands.
Our search for the hDREF/KIAA0785 binding sequence in the human genome revealed its presence in the promoter regions of many kinds of genes involved in DNA replication, DNA repair, cell cycle regulation, and transcription. We have already reported this to be the case for DmDREF (32), and some of the genes have been demonstrated to be actually under the control of the DRE/DREF system (1, 7, 39 -41). The gene catalogue obtained by data base searching suggests that hDREF/ KIAA0785 may be a functional homologue of DmDREF and may similarly take part in the transcriptional regulation of genes related to cell proliferation.
We here focused on the human histone H1 gene (FNC16/ H1.5), because it carries a nucleotide sequence completely matching the 10-bp sequence exhibiting the highest affinity to the hDREF/KIAA0785 and is well known as a gene whose expression is stringently coupled with DNA replication. Several groups have extensively analyzed the ϳ180-bp region from the cap site of the histone H1 promoter and found two elements that might be important in the S phase-specific H1 transcription; an AC box, 5Ј-AAACACA-3Ј, and a CCAAT box, 5Ј-AC-CAATCACA-3Ј (50 -53), exist at Ϫ176 to Ϫ167 and Ϫ112 to Ϫ103, respectively (21). The putative hDREF/KIAA0785 binding site is located at Ϫ390 to Ϫ381 in a more distal region that has not yet been characterized by anyone. Therefore, we cloned the promoter region (Ϫ537 to Ϫ1) of the histone H1 gene (FNC16/H1.5) and examined whether it is regulated by hDREF/KIAA0785. We can conclude that hDREF/KIAA0785 specifically binds to the human histone H1 gene promoter and stimulates its activity for the following two reasons. 1) EMSA experiments with anti-hDREF/KIAA0785 antibody or competitor oligonucleotides demonstrated that hDREF/KIAA0785 protein specifically binds to the nucleotide sequence at Ϫ401 to Ϫ371. 2) Transient co-transfection assays revealed that expression of full-length hDREF/KIAA0785 protein stimulates the histone H1 gene promoter activity in dose-dependent manner, whereas expression of a truncated form of hDREF/KIAA0785 does not.
Promoters of most DNA replication-dependent human H4, H3, H2A, H2B, and H1 histone genes do not contain a typical E2F binding site, although the expression of these genes is coordinately controlled in a cell-cycle dependent fashion. Many kinds of elements and trans-acting factors have been identified in vitro using mutation analysis of promoter regions of histone genes and EMSA. For example, the HiNF-D complex (consisting of CDP/Cut, CDC2, and cyclin A, a retinoblastoma-related protein) (54 -56), HiNF-M/IRF-2 (56 -58), and HiNF-B/H1TF2 2 F. Hirose and N. Ohshima, manuscript in preparation.

FIG. 6. Effects of transfecting an hDREF/KIAA0785-expressing plasmid on histone H1 gene promoter-driven reporter expression.
A, HeLa cells were co-transfected with 100 ng of luciferase reporter plasmid containing histone H1 gene promoter (H1-p/pGL3), 0 ng-300 ng of HA-hDREF/KIAA0785/pcDNA3-HA, and 10 ng of pRL-TK as an internal control. B, 300 ng of HA-hDREF/KIAA0785, HA-CR1hDREF/KIAA0785expressing plasmid, or control plasmid (pcDNA3-HA) were co-transfected with 100 ng of H1-p pGL3 and 10 ng of pRL-TK into HeLa cells. After incubation for 48 h, cells were harvested, and luciferase activity was measured with a dual-luciferase reporter assay system. Firefly luciferase activity was normalized to sea pansy luciferase activity and expressed as the value relative to that obtained without expression plasmid. Transfections were performed several times with at least three independent plasmid preparations. (52,53) are all suggested to be involved in up-regulation of human histone genes during G 1 -S transition. Despite much effort to characterize these factors, fluctuation during the cell cycle and roles of each in regulation of cell cycle-specific histone gene expression have not yet been clearly demonstrated, other than for HiNF-M/IRF2 and CDP/Cut protein (55,56,58). Western blotting and immunohistochemical analysis here showed that expression of hDREF/KIAA0785 protein is induced by adding serum to serum-starved cultures of normal human fibroblasts, reaching a maximum in S phase. In addition, considering our previous finding that overexpression of DmDREF in the eye imaginal disc induces ectopic DNA synthesis in the post-mitotic cells by up-regulating many DNA replication genes (18), we hypothesize that hDREF/KIAA0785 protein may have an important role in up-regulation of histone gene expression. To directly demonstrate hDREF/KIAA0785 function in the histone H1 gene expression, we tested whether reduction of endogenous hDREF/ KIAA0785 protein diminishes the histone H1 gene expression. As expected, introduction of siRNAs against hDREF/KIAA0785 resulted in reduction of histone H1 mRNA. Therefore, we concluded that hDREF/KIAA0785 might regulate the histone H1 gene expression. Interestingly, we found that the histone H4 genes, organized in the cluster containing the histone H1 gene (FNC16/ H1.5), also have hDREF/KIAA0785 putative binding sites.
Knock-down of hDREF/KIAA0785 protein resulted in inhibition of G 1 -S progression. We have already reported (13) that expression of a dominant-negative form of DmDREF in cells of eye imaginal discs inhibited G 1 -S progression in the second mitotic wave. Similar phenotype induced by knock-down of hDREF/KIAA0785 and DmDREF protein also provide a piece of evidence that hDREF/KIAA0785 might be the DmDREF homologue.
In summary, we here characterized hDREF/KIAA0785 protein as a human homologue of DmDREF and demonstrated that it might be a positive transcriptional regulatory factor. We FIG. 7. Increased expression of hDREF/KIAA0785 protein in the G 1 to S phase. A, cell cycle profiles after serum stimulation. Serum-starved HEL cells were harvested at the indicated points after serum addition and stained with propidium iodide. Cells were analyzed by flow cytometry, and the percentages in each phase were calculated. B, expression patterns of hDREF/KIAA0785 protein and histone H1 mRNA through the cell cycle. a, the serum-stimulated cell extracts were electrophoretically separated in a 10% SDS-polyacrylamide gel and analyzed by immunoblotting with rabbit anti-hDREF/KIAA0785 antibody. b, total RNA was extracted from serum-stimulated cells and subjected to RT-PCR. PCR products were loaded on a 2% agarose gel and stained with ethidium bromide. C, immunostaining of quiescent and serum-stimulated HEL cells with anti-hDREF/KIAA0785 antibody. Serum-starved cells were released by serum addition for 0 h (panels a and b) or 18 h (panels c and d) and immunostained with rabbit anti-hDREF/KIAA0785 antibody (red; left columns) and mouse anti-human PCNA monoclonal antibody (green; right columns). Magnification was ϫ200.
FIG. 8. Effects of siRNAs against hDREF/KIAA0785 protein on BrdUrd incorporation and the histone H1 gene expression. siR-NAs targeting hDREF/KIAA0785 protein (hDREFsiRNA1 and hDREF-siRNA2) were transfected into HeLa cells, and cells were incubated for 72 h at 37°C. As a control, mock or Scramble siRNA transfection was performed. A, reduction of hDREF/KIAA0785 protein in hDREFsiR-NAs-transfected cells. a, whole cell lysates were prepared from 1.4 ϫ 10 6 cells, loaded on an 8% SDS-polyacrylamide gel, and analyzed by immunoblotting with rabbit anti-hDREF/KIAA0785 antibody. Panels b-e, siRNAs-transfected cells were fixed with methanol, permeabilized with Triton X-100, and then immunostained with rabbit anti-hDREF/ KIAA0785 antibody. Panel b, mock; panel c, Scramble; panel d, hDREF-siRNA1; panel e, hDREFsiRNA2. Magnification was ϫ20. B, effect of hDREFsiRNAs on expression of the histone H1 gene. Total RNA isolated from siRNA-transfected cells was subjected to RT-PCR for quantification of histone H1 mRNA in the presence (ϩ) or absence (Ϫ) of reverse transcriptase. PCR products were loaded on a 2% agarose gel and stained with ethidium bromide. C, effect of hDREFsiRNAs on BrdUrd incorporation. HeLa cells were transfected with siRNAs and incubated for 72 h and then labeled with BrdUrd for 1 h. Incorporated BrdUrd was visualized using an anti-BrdUrd antibody and an alkaline phosphatase. Arrows indicate cells labeled with BrdUrd. Panel a, mock; panel b, Scramble; panel c, hDREFsiRNA1; panel d, hDREFsiRNA2. also generated evidence suggesting that the human histone H1 gene is one of the targets of hDREF/KIAA0785. Although the data in the present study potentially suggests that hDREF/ KIAA0785 might be the homologue of DmDREF, we cannot rule out a possibility that hDREF/KIAA0785 is not the "true" homologue. To answer the issue, we should provide evidence that hDREF/KIAA0785 possesses similar activity as DmDREF.
We are now trying to establish that transgenic flies expressing chimera hDREF/KIAA0785 protein swapped its N-terminal DNA binding domain with that of DmDREF. Furthermore, studies on other genes with putative hDREF/KIAA0785 protein binding sequence should increase our understanding of the biological function of hDREF/KIAA0785.