The novel human DNA helicase hFBH1 is an F-box protein.

We have identified a novel DNA helicase in humans that belongs to members of the superfamily I helicase and found that it contains a well conserved F-box motif at its N terminus. We have named the enzyme hFBH1 (human F-box DNA helicase 1). Recombinant hFBH1, containing glutathione S-transferase at the N terminus, was expressed in Sf9 cells and purified. In this report, we show that hFBH1 exhibited DNA-dependent ATPase and DNA unwinding activities that displace duplex DNA in the 3' to 5' direction. The hFBH1 enzyme interacted with human SKP1 and formed an SCF (SKP1/Cullin/F-box) complex together with human Cullin and ROC1. In addition, the SCF complex containing hFBH1 as an F-box protein displayed ubiquitin ligase activity. We demonstrate that hFBH1 is the first F-box protein that possesses intrinsic enzyme activity. The potential role of the F-box motif and the helicase activity of the enzyme are discussed with regard to regulation of DNA metabolism.

DNA helicases are ubiquitous enzymes that play essential roles in various DNA transactions involved in replication, repair, and recombination (1,2) and have been implicated in a number of human genetic disorders (3,4). A trademark of these enzymes is a set of well conserved amino acid sequences termed "helicase motifs" (5,6). Helicases are generally believed to generate single-stranded DNA by catalyzing the melting of stable helical DNA structures utilizing energy derived from the hydrolysis of nucleoside triphosphates. Analysis of the genome of Saccharomyces cerevisiae indicates that there are 134 open reading frames containing with helicase-like features that represent ϳ2% of its genome (7). The presence of such a large number of helicases or helicase-like proteins in cells, many of unknown function in vivo, most likely reflects a complexity of nucleic acid metabolic reactions and the distinct structural template requirements for a given helicase. Alternatively, many helicases may play roles that functionally overlap, making it difficult to determine a specific role for a given helicase. For example, each mutant cell of the two yeast helicases, Sgs1 and Srs2, grows with wild-type kinetics, but the combination of an sgs1 mutation with loss of the helicase srs2 resulted in a severe growth defect (8,9). A majority of sgs1 and srs2 mutant cells stops dividing stochastically as large budded DNA (9). Such double mutant cells are unable to replicate DNA at restrictive temperature and unable to efficiently transcribe rDNA (8). This suggests that these two helicases provide a redundant but essential activity for DNA replication and rDNA transcription. Despite well conserved motifs specific to helicase family proteins, there are several proteins for which no helicase activity has been demonstrated, indicating that helicase motifs cannot, without additional information, be used to define a protein as a helicase. For example, the chromatin-remodeling factor SWI2/SNF2 containing these motifs lacks helicase activity, although its ability to hydrolyze ATP is stimulated by DNA (10,11), suggesting that ATP hydrolysis provides the energy required to alter protein-DNA structure rather than duplex DNA or RNA structure. This suggests that proteins with helicase motifs may have functions that do not involve unwinding of nucleic acids.
As a continued effort to understand the role of Schizosaccharomyces pombe DNA helicase I that we reported previously (12), we identified a human homolog of this fission yeast enzyme and named it hFBH1 (human F-box DNA helicase 1) because it contains a well conserved functional F-box motif (13). In this report, we present data that hFBH1 is not only an ATPase/DNA helicase but also is an F-box protein that can form an SCF (SKP1-CUL1 (Cdc53)-Rbx1-F-box protein) complex with human SKP1 and CUL1. SCF complexes constitute a new class of E3 1 ligase that plays important roles in cell cycle regulation and signal transduction by catalyzing ubiquitinmediated proteolysis (14 -18). The substrate specificity of an SCF complex is governed by the interchangeable F-box protein subunit, which recruits a specific set of substrates for ubiquitination to the SCF core complex composed of SKP1, Cdc53, ROC1, and the E2 enzyme Cdc34. The polyubiquitinated proteins are rapidly captured by the 26 S proteasome, an abundant, self-compartmentalized protease particle (19). To date, hFBH1 is the first example of an F-box protein that contains intrinsic enzymatic activity, suggesting that a helicase can play a role in a certain aspect of DNA metabolism that requires ubiquitin-dependent proteolysis. The potential biological role of this interesting human DNA helicase will be discussed.

MATERIALS AND METHODS
Enzymes, Antibodies, and Nucleotides-The following proteins were obtained commercially: restriction endonucleases, the Klenow fragment * This work was supported by a grant from the Creative Research Initiatives of the Korean Ministry of Science and Technology (to Y.-S. S.). 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.
ʈ To whom correspondence should be addressed. of Escherichia coli DNA polymerase I, and polynucleotide kinase from KOSCO Inc. (Taejeon). The antibodies were obtained commercially: mouse monoclonal anti-FLAG antibodies from Sigma and mouse monoclonal anti-GST-horseradish peroxidase and rabbit polyclonal anti-human SKP1 from Santa Cruz Biotechnology. All of the secondary antibodies were from Amersham Biosciences. ⌽X174 single-stranded circular DNA was purchased from New England Biolabs. Radioisotopes, [␥-32 P]ATP (Ͼ5000 Ci/mmol), and [␣-32 P]dCTP (Ͼ6000 Ci/mmol) were purchased from Amersham Biosciences.
Cloning of the Full-length cDNA of hFBH1-A search for a human homolog of S. pombe DNA helicase I, using the TBLASTN homology search program (20) with the S. pombe DNA helicase I amino acid sequence as a query, resulted in the detection of an expressed sequence tag (EST) clone (GenBank TM accession number AA349988). This sequence was amplified by reverse transcription-PCR using a 5Ј hFBH1 EST primer (5Ј-GGA GAA TTC CTG CAC TCA GGC C-3Ј) and 3Ј hFBH1 EST primer (5Ј-GGA CGG AGT TGA CCA TGA AGG G-3Ј). An amplified band was excised from an agarose gel, purified using QIAquick gel extraction kit (Qiagen), and subcloned into pCR2.1 TOPO vector (Invitrogen). The EcoRI fragment derived from the subclone was labeled using [␣-32 P]dCTP with a Prime It II kit (Stratagene). A HeLa cell cDNA library in Uni-ZAP TM XR vector (Stratagene) was screened for a cDNA of hFBH1 according to the manufacturer's instructions. For the cloning of full-length cDNA, the 5Ј-end of the cDNA was extended by a 5Ј RACE-PCR. An M13F primer (5Ј-GGA AAC AGC TAT GAC CAT GAT TAC GC-3Ј), complementary to a flanking sequence of inserts in the library plasmid and a hFBH1-RACE1 primer (5Ј-GGG TCT CCG GAG GCA GAA GAG CAG TCG-3Ј), specific to the hFBH1 cDNA, were used for the first PCR using a human cDNA library as template. The second PCR using DNA obtained in the first round of amplification was performed with the M13F primer and a hFBH1-RACE2 primer (5Ј-CGC TGG ATG TCA TTC ACA CTG-3Ј). The reaction conditions for the first RACE-PCR were 30 cycles of 30 s at 95°C, 45 s at 60°C, and 1 min at 72°C. The subsequent nested PCR was carried out as described for the first PCR amplification except that annealing was performed at 65°C. The reaction products were excised from the gel, purified using QIAquick gel extraction kit, and ligated into the pCR2.1 TOPO. The sequences of several independent clones were determined with an automated DNA sequencer (ABI PRISM 310 Genetic Analyzer; PerkinElmer Life Sciences).
Expression and Purification of Recombinant hFBH1-To express hFBH1 as GST fusion protein in insect cells, the full-length cDNA of hFBH1 was cloned into pFastBac1 (Invitrogen) as follows. A PCRamplified GST gene from pGEX vector (Amersham Biosciences) was first subcloned into the BamHI and EcoRI sites of pFastBac1. An N-terminal portion of hFBH1 was amplified using a pair of primers (5Ј-GAA TTC A 1 TG GCC AAA AGC AAT TCT G 19 -3Ј and 5Ј-G 264 GC CTC CCT AGG CCT TGT GCA TGG C 240 -3Ј), and the remaining Cterminal region of hFBH1 was amplified using a pair of primers (5Ј-G 240 CCA TGC ACA AGG CCT AGG GAG GCC 264 -3Јand 5Ј-GC GGC CGC T 2910 CA GAA GAC GAG GAA GAG CAG 2890 -3Ј). The superscript numbers in the primers above and hereafter indicates the position of the nucleotides in hFBH1 cDNA with respect to the first A 1 of the initiation codon. The amplified N-terminal and C-terminal fragment of hFBH1 was digested with EcoRI/AvrII and AvrII/NotI, respectively, and then ligated into EcoRI/NotI-cleaved pFastBac1 that has a gene encoding GST cloned at BamHI and EcoRI sites. This procedure generated GST-hFBH1/pFastBac1. All of the PCR fragments were subcloned into PCR2.1 TOPO vector and sequenced to verify the absence of PCR errors.
The resulting plasmid was used to construct a recombinant baculovirus expressing a GST-tagged hFBH1 according to the manufacturer's instructions. The recombinant baculovirus was used to infect Sf9 insect cells for 72 h at a multiplicity of infection of 10. The infected cells (1 ϫ 10 6 cells/ml, 1.3 liters) were harvested, resuspended in 30 ml of lysis buffer (50 mM Tris-HCl, pH 8.0, 300 mM NaCl, 1 mM EDTA, 10% glycerol, 5 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride, 0.1 mM benzamidine, 1.25 g/ml leupeptin, 0.625 g/ml pepstatin A), and disrupted by sonication (three cycles of a 30-s pulse and a 2-min cooling interval). The extracts were cleared by centrifuging at 37,000 rpm for 1 h in a Beckman 45 Ti rotor, the supernatant (12.1 mg/ml, 33.5 ml) was incubated with 2.5 ml of Glutathione-Sepharose 4B beads (Amersham Biosciences) for 8 h at 4°C, and then the mixture was packed into a column. The column was washed with 10 column volumes of lysis buffer and eluted with 20 ml of glutathione elution buffer (10 mM glutathione, 25 mM Tris-HCl, pH 8.0, 300 mM NaCl, 1 mM EDTA, 10% glycerol, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride). The peak fractions (0.24 mg/ml, 8.1 ml) were pooled, equilibrated with T buffer (25 mM Tris-HCl, pH 7.5, 1 mM EDTA, 10% glycerol, 1 mM phenylmethylsulfonyl fluoride) containing 200 mM NaCl (buffer T200; hereafter, the number indicates the concentration of NaCl added to the buffer T), and then loaded onto a Hitrap-heparin column (1 ml; Amersham Biosciences). The column was eluted with a linear gradient (200 -800 mM) of NaCl in buffer T (10 ml). Fractions containing the DNA-dependent ATPase activity were pooled and dialyzed against buffer T100. The dialyzate (0.16 mg/ml, 7 ml) was loaded onto a Hitrap Q column (1 ml; Amersham Biosciences) and eluted with a linear gradient (100 -800 mM) of NaCl in buffer T (10 ml). The active fractions (0.21 mg/ml, 3.6 ml) were pooled and subjected again to glutathione-Sepharose 4B (200 l) to remove the residual polypeptides present in this fraction. The active fractions (Ͼ95% in purity) were pooled and stored at Ϫ70°C.
Immunologic Techniques: Immunodepletion and Coimmunoprecipitation-To establish that ATPase and helicase activities are intrinsic to the purified recombinant enzyme, immunodepletion experiments were carried out using Glutathione-Sepharose 4B beads. A mixture (200 l) containing beads (10 l) and purified GST-hFBH1 (400 ng) was adjusted to 300 mM NaCl and incubated for 4 h at 4°C with occasional rocking. The mixture was spun down, and the supernatant was examined for the presence of ATPase and helicase activities. M2 anti-FLAG beads (Sigma) were used as a negative control. Coimmunoprecipitation was carried out using cell extracts obtained 48 h after transfection. The cells were incubated in 300 l of immunoprecipitation buffer (10 mM Tris-HCl, pH 7.4, 150 mM NaCl, 0.3% Triton X-100, 1 mM EDTA, 1 mM EGTA, 0.2 mM phenylmethylsulfonyl fluoride) for 30 min at 4°C. The cell extracts were cleared by centrifuging at 15,000 ϫ g for 15 min and incubated for 4 h at 4°C with 20 l of appropriate resin as indicated. After quick spin, the sedimented resin was washed twice with 1 ml of immunoprecipitation buffer and 1 ml of phosphate-buffered saline once. The proteins were released from the resin by adding 1ϫSDS-PAGE sample buffer and subjected to SDS-PAGE. Western blot analyses were performed with antibodies as indicated in each experiment.
Preparation of Helicase Substrates-The DNA substrate used for standard helicase assays (see Fig. 2C) were exactly as described previously (12). In brief, the substrate consists of ⌽X174 single-stranded cDNA with a 20-nucleotide oligonucleotide (5Ј-GGC GAT TGC GTA CCC GAC GA-3Ј) annealed to it. To prepare DNA substrates to measure direction of hFBH1 translocation (see Fig. 3), a 52-mer oligonucleotide (5Ј-CGA ACA ATT CAG CGG CTT TAA CCG GAC GCT CGA CGC CAT TAA TAA TGT TTT C-3Ј; 10 pmol) containing a sequence complementary to ⌽X174 single-stranded cDNA at nucleotides 703-754 was annealed to ⌽X174 single-stranded cDNA (2 pmol). The annealed oligonucleotide was labeled either at its 3Ј-end by incorporating [␣-32 P]dCTP with the presence of Klenow fragment or at its 5Ј-end by incorporating [␥-32 P]ATP with polynucleotide kinase. The partial duplex region of the 5Ј-end labeled DNA substrate was cleaved by HpaII to generate a linear DNA in which a labeled 23-nucleotide fragment present at the 5Ј-end of the template was labeled. The cleavage by HpaII of the 3Ј-end labeled DNA substrate resulted in generation of linear DNA with a labeled 29-nucleotide fragment at the 3Ј-end of the template. These two substrates were used to measure the translocation direction of the hFBH1 enzyme.
Helicase activity was measured in a reaction mixture (20 l) containing 25 mM Tris-HCl (pH 7.8), 2 mM MgCl 2 , 2 mM dithiothreitol, 2 mM ATP, 0.25 mg/ml bovine serum albumin, and the 3Ј-32 P-labeled partial duplex DNA substrate (15 fmol). The reactions were incubated at 37°C for 10 min and stopped with 4 l of 6ϫ stop solution (60 mM EDTA, pH 8.0, 40% (w/v) sucrose, 0.6% SDS, 0.25% bromphenol blue, 0.25% xylene cyanol). The reaction products were subjected to electrophoresis for 1.5 h at 150 volts through 10% polyacrylamide gels containing 0.1% SDS in 0.5ϫ TBE (45 mM Tris base, 45 mM boric acid, 1 mM EDTA). The gel was dried on DEAE-cellulose paper and autoradiographed. Labeled DNA products were quantitated with the use of a PhosphorImager. The background level detected in the absence of added DNA helicase was less than 2% of the input substrate, and this value was subtracted from the amount of displaced products formed in the presence of the DNA helicase.
To construct hFBH1/pAS2-1, an EcoRI-AvrII fragment was isolated from GST-hFBH1/pFastBac1 described above, and a C-terminal fragment was amplified using a pair of primers (5Ј-G 240 CCA TGC ACA AGG CCT AGG GAG GCC 264 -3Ј and 5Ј-GTC GAC T 2910 CA GAA GAC GAG GAA GAG CAG 2890 -3Ј). The amplified C-terminal fragment was restricted with AvrII and SalI. The two fragments were simultaneously ligated into pAS2-1 vector digested with EcoRI and SalI to construct hFBH1/pAS2-1. For construction of ⌬FhFBH1/pAS2-1, we first amplified the N-terminal region using a pair of primers (5Ј-GAA TTC A 1 TG GCC AAA AGC AAT TCT G 19 -3Ј and 5Ј-A 426 GG CAG GCA TCG ATT GTG GCT CAG 403 -3Ј) and digested the fragment with EcoRI and ClaI. This and the ClaI-SalI fragment from hFBH1/pAS2-1 were ligated into pAS2-1 digested with EcoRI and SalI. This procedure resulted in an internal deletion of the F-box motif. To construct N-hFBH1/pAS2-1, we isolated EcoRI fragment encoding amino acids 1-345 from hFBH1/ pAS2-1 and subcloned it into the EcoRI site of pAS2-1. The orientation was confirmed by enzyme digestion. For C-hFBH1/pAS2-1, we amplified the C-terminal fragment using a pair of primers (5Ј-CAT ATG A 853 AT GAC ATC CAG CGA CTG 870 -3Ј and 5Ј-GTC GAC T 2910 CA GAA GAC GAG GAA GAG CAG 2890 -3Ј). The resulting NdeI-SalI fragment encoded amino acids 285-969 and was subcloned into the NdeI-SalI site of pAS2-1.
The cDNA coding for human SKP1 was inserted into pGAD424 for GAL4 activation domain fusion. Yeast two-hybrid assays were carried out as described (21). The host strain used was S. cerevisiae Y190 (CLONTECH), and transformation was performed as described (22). The ␤-galactosidase assays were performed using liquid cultures as described (23,24).
Cell Culture and Transfection-Human 293T cells were maintained in Dulbecco's modified Eagle's medium (Invitrogen) containing 10% fetal bovine serum at 37°C with 5% CO 2 in a humidified incubator. For experiment in which the in vivo interactions of hFBH1 with proteins of interest were examined, plasmids (2 g each) were transiently transfected into 3 ϫ 10 5 cells in 6-well plates using the standard calcium phosphate precipitation method (25).
Constructs for Protein Expression in Human Cell Lines-To verify the in vivo interaction of hFBH1 with other proteins, an oligonucleotide encoding the FLAG epitope (MDYKDDDD) was subcloned in the Hin-dIII and NotI sites of pcDNA3 (Invitrogen), followed by insertion of the cDNAs of full-length hFBH1 or ⌬FhFBH1 into the NotI site. Thus, the hFBH1 and ⌬FhFBH1 proteins were expressed as a FLAG epitope fusion protein at their N termini. Full-length cDNAs encoding human SKP1 and CUL1 were amplified from a human cDNA library (CLON-TECH) using pairs of primers (upstream and downstream primers contained a BamHI and a KpnI site, respectively) complementary to the 5Ј-and 3Ј-ends of each cDNA. The BamHI-KpnI fragments were then ligated into pEBG for expression as GST fusion proteins. The fulllength cDNA of human ROC1 was amplified from a human cDNA library by PCR. An oligonucleotide encoding the HA epitope (MYPYD-VPDYA) was first subcloned into pcDNA3, followed by insertion of the full-length cDNA of ROC1 into the EcoRI and XhoI sites of modified pcDNA3.
Isolation of E1, E2, and 32 P-Labeled Ubiquitin-The hCdc34 protein, E2 conjugation enzyme, was purified from E. coli harboring a pET3a plasmid expressing the wild-type Cdc34 protein. The purification of E2 was carried out as reported previously (26). Human E1 was isolated from cytosolic extracts of HeLa cells using an ubiquitin affinity column as described previously (27). 32 P-Labeled ubiquitin was prepared as described previously (28). Phosphorylation of 32 P-labeled ubiquitin (7 g) was carried out in a reaction mixture (20 l) containing 20 mM Tris-HCl, pH 7.4, 12 mM MgCl 2 , 2 mM NaF, 50 mM NaCl, 25 M ATP, 5 Ci of [␥-32 P]ATP, 0.1 mg/ml bovine serum albumin, and 1 unit of cAMP kinase (Sigma). The reaction mixture was incubated at 37°C for 30 min. To inactivate the kinase, the mixture was heated at 70°C for 3 min.

RESULTS
Cloning and Analysis of a cDNA Encoding a Human F-box DNA Helicase-Using the TBLASTN homology search program, we identified a human EST cDNA clone (GenBank TM accession number AA349988) that possessed a significant homology to the S. pombe DNA helicase I (Ref. 12; GenBank TM accession number AAK58077). Screening of human cDNA libraries using the EST clone as a probe resulted in the isolation of two partial clones lacking the 5Ј region of the cDNA. Using a 5Ј RACE-PCR method, we isolated the missing 5Ј one-third of the full-length cDNA. The 5Ј cDNA amplified by reverse transcription-PCR, together with the two partial cDNAs, was used to reconstitute the full-length cDNA (GenBank TM accession number AF380349). The open reading frame of the cloned cDNA encoded a polypeptide of 969 amino acids with all seven conserved helicase motifs (Fig. 1A), and the protein had 28% identity and 44% similarity at the amino acid level to the S. pombe homolog and 95% identity to the mouse protein (data not shown). Interestingly, it was noted that all three homologs (humans, mice, and S. pombe) included a well conserved F-box motif (Fig. 1B) that is required to form a complex with SKP1, a key component of the SCF complexes involved in ubiquitin-dependent proteolysis (16). In addition, all three homologs contained the seven well conserved helicase motifs (Fig. 1C). Therefore, we named the human enzyme hFBH1 (human F-box DNA helicase 1).
The hFBH1 Has Both DNA-dependent ATPase and DNA Unwinding Activities-To determine whether hFBH1 contained helicase activity, we expressed the enzyme as a GSTfused recombinant protein (GST-hFBH1) in insect cells as described under "Materials and Methods." The expression of recombinant GST-hFBH1 was monitored by Western blot analyses using two antibodies: anti-GST antibody, which detects only a polypeptide with an intact N terminus, and anti-hFBH1 polyclonal antibody ␣-hFBH1, which was raised against the N-terminal 345-aa region of hFBH1. The full-length GST-hFBH1 protein was detected by both anti-GST and anti-hFBH1 antibodies in the crude extracts prepared from Sf9 insect cells infected with the recombinant baculovirus expressing hFBH1 ( Fig. 2A, lanes 2, 5, and 8), whereas it was not detected in control extracts prepared from uninfected cells ( Fig. 2A, lanes  1, 4, and 7).
Because the crude extracts prepared from the baculovirusinfected cells displayed elevated levels of single-stranded DNAdependent ATPase activity compared with control extracts (data not shown), we purified GST-hFBH1 by monitoring ATP hydrolysis in the presence and absence of single-stranded DNA as a cofactor. The hFBH1 enzyme purified according to the procedure described under "Materials and Methods" was also subjected to SDS-PAGE and was detected by both antibodies (Fig. 2A, lanes 3, 6, and 9). To demonstrate that the ATPase and helicase activities detected are intrinsic to hFBH1, we examined whether they were specifically depleted by glutathione-coupled beads (Fig. 2, B and C). The addition of buffer alone (Fig. 2, B, lanes 2 and 6, and C, lane 3) or unrelated antibodies such as anti-FLAG monoclonal antibody (Fig. 2, B, lanes 4 and 8, and C, lane 5) failed to deplete either ATPase (Fig. 2B) or DNA unwinding activity (Fig. 2C). In contrast, the addition of glutathione beads efficiently depleted the ATPase and helicase activities in the solution (Fig. 2, B, lanes 3 and 7, and C, lane 4), demonstrating that both activities are intrinsic to the purified hFBH1 enzyme. When we examined the unwinding direction of hFBH1, it displaced the 23-mer oligonucleotide annealed to the 5Ј-end region of the template (Fig. 3,  lanes 6 and 9), indicating that the enzyme translocated in the 3Ј to 5Ј direction. The enzyme was not dependent on fork structures for its unwinding activity (data not shown).
The hFBH1 Enzyme Interacts with SKP1, Forming a SCF Complex-Because the F-box protein, a component of a SCF complex, interacts with SKP1, we investigated whether the hFBH1 enzyme physically interacts with the human SKP1 protein. For this purpose, we constructed plasmids expressing bait proteins that contained different regions of hFBH1 fused to a GAL DNA-binding domain; these included the full-length protein, F-box-deleted protein, the N-terminal 1-345-aa fragment, and the C-terminal 285-969-aa fragment of hFBH1 (hFBH1, ⌬FhFBH1, N-hFBH1, and C-hFBH1, respectively; Fig. 4A). The F-box motif is present in the full-length hFBH1 and N-hFBH1 constructs as shown Fig. 1 (A and B). The human SKP1 gene was fused to a GAL4 activation domain to prepare a prey plasmid as described under "Materials and Methods." The yeast transformants that did not express hFBH1 (vector) or those that expressed either ⌬FhFBH1 lack-ing the F-box motif or the C-terminal fragment of hFBH1 failed to activate transcription of the reporter gene in combination with the human Skp1 expression vector (Fig. 4A). Such transformants were unable to grow on selective media containing 10 mM 3-aminotriazole as shown in Fig. 4A. In contrast, transformants obtained with the full-length hFBH1 and the N-terminal fragment containing the F-box motif grew in the presence of 10 mM 3-aminotriazole (Fig. 4A). Consistent with this, we detected significant ␤-galactosidase activity in the extracts derived from the growing cells (Table I), suggesting that interaction between the hFBH1 and SKP1 proteins occurs in vivo.
To further verify that hFBH1 formed a complex in vivo with SKP1 in an F-box-dependent manner, we transfected transiently a plasmid expressing GST-tagged human SKP1 (GST-SKP1) together with a plasmid expressing either FLAG-tagged full-length (FLAG-hFBH1) or F-box-deleted hFBH1 (FLAG-⌬FhFBH1, lacking amino acids 141-185) proteins into the human 293T cell line and prepared extracts for coimmunoprecipitation experiments (Fig. 4, B and C). FLAG-hFBH1 and GST-hSKP1 were coimmunoprecipitated by anti-FLAG antibodies (Fig. 4B, lane 2). In the absence of FLAG-hFBH1, GST-hSKP1 was not precipitated (Fig. 4B, lane 1). Moreover, the coimmunoprecipitation did not occur when the F-box motif was deleted (Fig. 4B, lane 3). This was also the case when coimmunoprecipitation was performed using glutathione beads (Fig.  4C). These results demonstrate that hFBH1 interacts with hSKP1 in vivo and that the interaction between SKP1 and hFBH1 is dependent on the presence of the functional F-box motif.
Because we confirmed that hFBH1 interacts with SKP1 and that the interaction required a functional F-box motif, we examined whether the hFBH1-SKP1 complex contained Cullin (CUL1), another protein that interacts directly with SKP1 to form an SCF complex. For this purpose, we constructed an additional plasmid expressing GST-tagged human CUL1 (GST-hCUL1) that was transiently transfected in combination with a plasmid expressing either FLAG-hFBH1 or FLAG-⌬FhFBH1 into 293T cells. FLAG-hFBH1 was precipitated by FLAG antibody resin (Fig. 5A, lanes 2-4), and GST-hCUL1 was coimmunoprecipitated only when FLAG-hFBH1 with a functional Fbox motif was expressed (Fig. 5A, compare lanes 2 and 3). This was also the case when the coimmunoprecipitation was performed using glutathione beads (Fig. 5B, lane 2). Consistent with the fact that hFBH1 interacts with hSKP1 (Fig. 4), we were able to detect endogenous hSKP1 in the immunocomplex precipitated by anti FLAG antibody resin (Fig. 5A, lanes 2 and  4) regardless of exogenous expression of GST-hCUL1. The endogenous SKP1 was also observed only in the presence of the functional F-box motif of hFBH1 (Fig. 5A, compare lanes 2 and  3). These results confirm that hFBH1 interacts with endogenous hSKP1 and demonstrates that it forms a complex with hCUL1.
It has been known that CUL1 forms a direct complex with ROC1, the fourth subunit of human SCF complexes, and that CUL1 interacts with hSKP1 (28). Therefore, F-box proteins interact indirectly with CUL1 and ROC1 through SKP1. Thus, the association of hFBH1 with these two proteins should depend on a functional F-box motif. We tested this possibility as shown in Fig. 6A. Consistent with the previous finding (28), ROC1 was coimmunoprecipitated with hCUL1 (Fig. 6A, lane 1). In the absence of hCUL1 and ROC1 expression, neither hFBH1 nor the mutant hFBH1 lacking the F-box motif (FLAG-⌬FhFBH1) was immunoprecipitated by glutathione beads (Fig.  6A, lanes 2 and 3, receptively). However, hFBH1 was coimmunoprecipitated with GST-hCUL1 together with HA-ROC1 (Fig.  6A, lane 4). This interaction was dependent upon the presence of a functional F-box motif in hFBH1, because FLAG-⌬FhFBH1 failed to coimmunoprecipitate with the CUL1-ROC1 (Fig. 6A, lane 5).
The SCF Complex Containing hFBH1 Has Ubiquitin Ligase Activity-Because we confirmed that hFBH1 formed a complex with all three other subunits (SKP1, CUL1, and ROC1) for an SCF complex, we examined whether the SCF complex containing hFBH1 as an F-box protein (SCF hFBH1 ) possesses ubiquitin ligase activity as is expected for an SCF complex. For this purpose, we purified E1, the Ub-activating enzyme, and hCdc34 (E2), the Ub-conjugating enzyme, which are required for polymerization of ubiquitin by E3 ligase activity of an SCF complex, as described under "Materials and Methods." Plasmids for expressing hCUL1, hFBH1, or hROC1 were transiently transfected into human 293T cells in various combinations as depicted in Fig. 6B. Immunoprecipitated proteins were used as an enzyme source to measure the ability to polymerize 32 P-labeled ubiquitin in the presence of E1 and E2. Although the reaction mixture alone (negative control) produced limited ubiquitin polymerization (Fig. 6B, lane 1), GST-hCUL1/HA-ROC1 (positive control) immunoprecipitated with glutathione beads efficiently produced long polyubiquitin chains that consist of more than 10 ubiquitin monomers (Fig. 6B, lane 2), in keeping with the previous result (28). When we transfected a plasmid expressing FLAG-⌬FhFBH1 lacking the F-box motif Helicase and ATPase activities were measured in standard reaction mixtures described under "Materials and Methods." The helicase substrate is the standard substrate described previously (12). The amount of ATP hydrolyzed or the substrate unwound is presented below the autoradiograms. ss, single-stranded. along with the plasmids expressing GST-hCUL1 and HA-ROC1, the complex immunoprecipitated by FLAG antibody resin failed to produce long chains of polyubiquitin (Fig. 6B,  lane 4), similar to negative control (Fig. 6B, lane 1). In contrast, the immunoprecipitated complex with FLAG-hFBH1 displayed ubiquitin ligase activity (Fig. 6B, lane 3). Formation of ubiquitin polymers by immunoprecipitated complexes containing GST-hCUL1 and HA-ROC1 is more efficient than that by those containing hFBH1 (Fig. 6B, compare lanes 2 and 3). This reflects the fact that glutathione beads against GST-hCUL1 precipitated a large number of SCF complexes in the extracts, whereas FLAG antibody precipitated only a subset of the SCF complexes that contains hFBH1 only. When E1 or E2 was removed from the reaction mixture, no polyubiquitin chain was detected (Fig. 6B, lanes 5 and 6, respectively), indicating that the formation of polyubiquitin chain by SCF hFBH1 is dependent on E1 and hCdc34. This result confirms that SCF hFBH1 contains ubiquitin ligase activity and interacts with the ubiquitination machinery. DISCUSSION In this report, we demonstrate that hFBH1 is a DNA helicase that translocates in the 3Ј to 5Ј direction and interacts with SKP1 in a manner that requires a functional F-box motif. Besides, hFBH1 interacted with CUL1/ROC1 and formed a SCF complex that contained all known four subunits (SKP1, hFBH1, CUL1, and ROC1). Moreover, SCF hFBH1 displayed ubiquitin ligase activity in a F-box motif-dependent fashion. The polyubiquitination activity of the SCF hFBH1 complex was dependent upon functional E1 and E2 enzymes, demonstrating that the complex interacts with the ubiquitination machinery. Our study showed that hFBH1 is the first example of an F-box protein that harbors an intrinsic enzymatic activity, catalyzing a helicase reaction. In addition, hFBH1 is the first F-box protein implicated in nucleic acid metabolism. Therefore, our findings raise the possibility that the role of hFBH1 is most likely   FIG. 4. hFBH1 forms a complex with SKP1 in a F-box motif-dependent fashion. A, hFBH1 interacts with SKP1 in the yeast twohybrid assays. A yeast strain Y190 expressing full-length (hFBH1, 1-969 aa), F-box-deleted (⌬FhFBH1), N-terminal (N-hFBH1, 1-345 aa), or C-terminal (C-hFBH1, 285-969 aa) fragment of hFBH1 was transformed with hSKP1/GAD424 to detect two-hybrid interactions between the two proteins. Aliquots containing transformants (5 ϫ 10 4 , 5 ϫ 10 3 , 5 ϫ 10 2 , 5 ϫ 10 1 , and 5 ϫ 10 0 ) were spotted onto synthetic dextrose minimal plates lacking His, Leu, and Trp. The cells were grown at 30°C in duplicate in the presence (10 mM) or absence of 3-aminotriazole. B and C, hFBH1 coimmunoprecipitates with SKP1. Plasmids (2 g each) for expressing GST-tagged human SKP1 (GST-hSKP1) and FLAGtagged human FBH1 (wild type, FLAG-hFBH1; F-box-deleted, FLAG-⌬FhFBH1) proteins were transiently transfected into the 3 ϫ 10 5 cells of human 293T cells in combinations as indicated. Blank vectors with tags only were used in the minus lanes. Anti-FLAG (␣-FLAG) and glutathione-coupled (GTT) beads (B and C, respectively; 20 l each) were used for immunoprecipitation. Coimmunoprecipitated proteins were visualized by Western blot analyses using anti-GST (␣-GST; top blot, 12% SDS-PAGE) or anti-FLAG (␣-FLAG; bottom blot, 8% SDS-PAGE) antibodies. One-fifth (24 g) of crude extracts used for immunoprecipitation were loaded in extract lanes ( lanes 5-8). IP, immunoprecipitation.  to recruit a target protein, most likely involved in some aspect of DNA metabolism, to the protein degradation pathway. Thus, a protein involved in DNA metabolism would be marked by hFBH1 for degradation via the highly regulated ubiquitin degradation pathway.
As an initial attempt to evaluate the role of the F-box motif of the enzyme in this regard, we decided to identify a protein(s) that interacts specifically with hFBH1 using the yeast two-hybrid screen. This screen resulted in the isolation of human MAT1 (data not shown), which is a regulatory subunit of the Cdk-activating kinase (29,30). It is unlikely that MAT1 is a direct substrate of SCF hFBH1 E3 ligase activity because a substrate phosphorylation is a prerequisite step prior to recognition by an F-box protein (31,32). MAT1 in Cdk-activating kinase may play a role leading to the phosphorylation of a substrate in proximity with hFBH1, thereby facilitating the subsequent ubiquitination of the substrate by the SCF hFBH1 complex. In support of this possibility, we failed to detect any ubiquitination of MAT1 (data not shown). These findings raise the possibility that MAT1 may play a direct role in the phosphorylation of hFBH1 instead of acting as a substrate for the E3 ligase activity of the SCF hFBH1 complex. Therefore, the level of hFBH1 itself is likely to be regulated by posttranslational modifications that involve both Cdk-activating kinase and E3 ligase activities. Because it has been reported that F-box proteins can be degraded autocatalytically in a ubiquitination-dependent fashion (33), the F-box motif of hFBH1 may regulate the level of hFBH1 through auto-ubiquitination during cell cycle progression. However, we were not able to demonstrate this, because the endogenous levels of hFBH1 were too low to be detected with the antibodies that we have (data not shown). We also explored whether hFBH1 is ubiquitinated in vivo. Although the hFBH1 was indeed ubiquitinated in vivo (data not shown), the efficiency of hFBH1 ubiquitination was independent of the intact F-box motif, suggesting that ubiquitination occurred in trans by some other ubiquitinating activity. Reconstitution of the SCF complex containing hFBH1 with the purified proteins will help to address this question.
Although the biological function in vivo of hFBH1 is unclear at present, mutational studies of the S. pombe fdh1 ϩ gene (encoding the S. pombe homolog of hFBH1) have provided some clues. When the S. pombe fdh1 ϩ gene was deleted, the cells showed elongated cell morphology and uneven distribution of chromosomes in dividing cells. 2 In addition, deletion of either the F-box motif or a single amino acid change in the ATP-binding motif resulted in phenotypes similar to those of the null mutation. 2 These observations suggest that both the F-box and ATPase/helicase activity of hFBH1 are required for the physiological function of the enzyme. Currently, both genetic and biochemical studies are underway using an S. pombe model system to delineate the precise biological role of hFBH1.