Identification of a consensus cyclin-dependent kinase phosphorylation site unique to the nuclear form of human deoxyuridine triphosphate nucleotidohydrolase.

In the preceding report (Ladner, R.D., McNulty, D.E., Carr, S.A., Roberts, G.D., and Caradonna, S.J. (1996) J. Biol. Chem. 271, 7745-7751), we identified two distinct isoforms of dUTPase in human cells. These isoforms are individually targeted to the nucleus (DUT-N) and mitochondria (DUT-M). The proteins are nearly identical, differing only in a short region of their amino termini. Despite the structural differences between these proteins, they retain identical affinities for dUTP (preceding article). In previous work, this laboratory demonstrated that dUTPase is posttranslationally phosphorylated on serine residue(s) (Lirette, R., and Caradonna, S. (1990) J. Cell. Biochem. 43, 339-353). To extend this work and determine if both isoforms of dUTPase are phosphorylated, a more in depth analysis of dUTPase phosphorylation was undertaken. [32P]Orthophosphate-labeled dUTPase was purified from HeLa cells, revealing that only the nuclear form of dUTPase is phosphorylated. Electrospray tandem mass spectrometry was used to identify the phosphorylation site as Ser-11 in the amino-terminal tryptic peptide PCSEETPAIpSPSKR (the NH2-terminal Met is removed in the mature protein). Mutation of Ser-11 by replacement with Ala blocks phosphorylation of dUTPase in vivo. Analysis of the wild type and Ser-11 --> Ala mutant indicates that phosphorylation does not regulate the enzymatic activity of the DUT-N protein in vitro. Additionally, experiments with the Ser-11 --> Ala mutant indicate that phosphorylation does not appear to play a role in subunit association of the nuclear form of dUTPase. The amino acid context of this phosphorylation site corresponds to the consensus target sequence for the cyclin-dependent protein kinase p34(cdc2). Recombinant DUT-N was specifically phosphorylated on Ser-11 in vitro with immunoprecipitated p34(cdc2). Together, these data suggest that the nuclear form of dUTPase may be a target for cyclin-dependent kinase phosphorylation in vivo.

Human dUTPase 1 was first purified from HeLa cells by Caradonna and Adamkiewicz (2). The enzyme was characterized as a homodimer with a monomeric molecular weight of 22,500 and a K m for dUTP of 2.5 M. The dUTPase monomers associate in the presence of divalent cations such as magnesium or manganese to form the active enzyme. In later work, we identified the human enzyme as a serine phosphoprotein (1). Studies on herpesvirus infection of HeLa cells have shown that cellular dUTPase activity decreases postinfection, while the virus-encoded dUTPase activity increases. It was postulated that the associated decrease of cellular dUTPase activity was not due to rapid degradation but rather correlated with dephosphorylation of the host dUTPase protein. These data suggest that phosphorylation may play a role in regulating the enzymatic activity of the human dUTPase protein. More recently, Strahler and co-workers demonstrated that, upon peripheral blood lymphocyte stimulation, there is a large induction of the phosphorylated form of dUTPase. This induction of dUTPase protein coincided with the onset of DNA replication, suggesting a link between dUTPase phosphorylation and the proliferation status of the cell (4).
In this report we extend previous work involving dUTPase phosphorylation. A single site of phosphorylation correlating to Ser-11 of the nuclear isoform of dUTPase was identified. Although both the DUT-N and DUT-M contain the identical site, phosphorylation of Ser-11 is unique to the nuclear isoform. This site correlates with the consensus sequence for cyclin-dependent kinase phosphorylation and is specifically phosphorylated by p34 cdc2 in vitro, suggesting a link to the cyclin signaling pathway. Studies with a Ser-11 3 Ala mutant of DUT-N suggest that this modification is unrelated to both enzymatic activity and subunit association.

EXPERIMENTAL PROCEDURES
Cell Culture and Labeling-HeLa S3 cells (CCL 2.2) were purchased from the American Type Culture Collection and maintained in Dulbecco's modified Eagle's medium supplemented with 5% calf serum (Life Technologies, Inc.). COS 7 cells (CRL 1651) were purchased from the American Type Culture Collection and maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum (Life Technologies, Inc.). 32 P-Labeling of dUTPase was accomplished by incubating HeLa cells for 24 h in phosphate-deficient Dulbecco's modified Eagle's medium supplemented with 5% fetal calf serum and [ 32 P]orthophosphate.
Purification of dUTPase-Purification of dUTPase from HeLa, Cos-7, and Sf21 cells was performed using a modification of the method devel-* 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. This research was supported by NCI, National Institutes of Health, Grant CA42605 (to S. J. C.  Caradonna and Adamkiewicz (2) as detailed in the accompanying article (22).
Enzyme Assays-dUTPase activity was measured using a procedure described by Caradonna and Adamkiewicz (2).
Expression of Recombinant dUTPase-Baculovirus expression of the DUT-N gene is described in the previous paper (22). The DUT-N open reading frame was also cloned into the pEUK-C1 (Clonetech) eukaryotic expression vector and transfected into COS-7 cells with Lipofectin (Life Technologies, Inc.). Transfections were carried out as per the manufacturer's recommendations.
Antibodies-dUTPase-specific monoclonal antibodies were generated and prepared as described (1). These antibodies do not cross-react with species other than human and are only effective for immunoprecipitation and immunoaffinity chromatography. The monoclonal antibody immunoprecipitates both recombinant DUT-N and the Ser-11 3 Ala mutant protein with equivalent recoveries. dUTPase-specific polyclonal antibodies, useful for immunoblot analysis, were raised and affinitypurified as detailed in the previous paper (22). The polyclonal antibodies only recognize the human isoforms of dUTPase and do not crossreact with dUTPase derived from other species.
Site-directed Mutagenesis-The Ser-11 3 Ala mutation was introduced into the dUTPase cDNA using the Transformer Mutagenesis kit from Clonetech. Nucleotide 82 of the DUT-N cDNA sequence was changed from C to G, utilizing the following mutagenic primer: 5Ј-TTACTGGGTGCAATGGCGGGT-3Ј. The mutation was confirmed by sequence analysis.
Immunoprecipitation of p34 cdc2 and Kinase Assays-Immunoprecipitations were carried out according to the protocol described by Oncogene Science. Cellular extracts (1 mg/ml) were incubated with antibody for 18 h at 4°C. Protein A-Sepharose was added (150 l of a 10% solution) to pellet the antibody-antigen complexes. For kinase assays, pellets were washed five times in buffer containing 50 mM Tris-HCl, pH 7.4, 0.25 M NaCl, 0.1% (v/v) Nonidet P-40, 5 mM EDTA, and 50 mM NaF). Peptide competition studies were carried out by preincubating the p34 cdc2 antisera (CDC2 (Ab-1)) with a 5-fold molar excess of blocking peptide (CDC2 (peptide 1)) for 2 h at 4°C. The immunoprecipitation was then carried out as above.
dUTPase Immunoprecipitations and Western Blot Analysis-Total HeLa cell extracts (1 mg/ml) were incubated with approximately 10 g of monoclonal antibody for 18 h at 4°C. The amount of antibody utilized in these studies is in excess of the antigen. The immunoprecipitations deplete nearly all of the dUTPase protein in the extracts (as judged by Western blot). Protein A-Sepharose was added (150 l of a 10% solution) to pellet the antibody-antigen complexes. The complexes were washed five times in buffer containing 50 mM Tris-HCl, pH 7.4, 0.15 M NaCl, 1% (v/v) Nonidet P-40, 0.1% SDS, 0.5% sodium deoxycholate, and 5 mM EDTA. The samples were subsequently prepared for SDS-PAGE analysis. Western blot analysis was performed as described in the previous report.
HPLC Analysis/Purification-32 P-Labeled dUTPase was purified by the method described above and fractionated by 15% SDS-PAGE. In-gel reduction, alkylation, and tryptic digestion were performed as described in the previous report (22). The peptides were extracted from the gel slices three times with 150 l of 60% CH 3 CN, 0.1% trifluoroacetic acid for 20 min each with sonication. All supernatants were combined and evaporated to near dryness in a Speedvac concentrator. The tryptic digest was fractionated on a Vydac C18 column (2.1 ϫ 150 mm) with a linear AB gradient from 5 to 60% B (0.5% B/min), where A is 0.1% aqueous trifluoroacetic acid and B is 0.1% trifluoroacetic acid in acetonitrile (pH 2.0); flow rate was 0.5 ml/min; detection was at 214 nm (Waters HPLC). 0.25-ml fractions were collected, and radioactivity was determined by scintillation counting. The fractions containing the 32 P label were pooled and refractionated, and the phosphopeptide was collected manually.
Electrospray and Tandem Mass Spectrometry of Phosphopeptides-The phosphopeptide-containing HPLC fraction, containing from 10 to 25 pmol of peptide, was concentrated to near dryness on a Speedvac centrifuge (Savant Instruments) and reconstituted in 50 l of 1:1 methanol/H 2 O (v/v), 0.2% in formic acid. The sample was examined by electrospray mass spectrometry on a PE-Sciex API-III triple quadrupole mass spectrometer fitted with a fully articulated ionspray probe and an atmospheric pressure ionization source. Approximately 5 l of the sample (at 0.2-0.5 pmol/l) was introduced into the mass spectrometer by infusion with an infusion pump (Harvard Apparatus) at a flow rate of 1 l/min. Electrospray mass spectra were acquired by scanning quadrupole 1 from m/z 10 to 2400 with a mass step of 0.2 Da and 10-ms dwell/step. Doubly charged parent ions, (M ϩ 2H) 2ϩ , were selected for fragmentation and structural analysis with quadrupole 1. The massselected parent ion was subjected to collision-induced decomposition in quadrupole 2 of the triple quadrupole. Quadrupole 3 was scanned from m/z 10 to 2400 with a mass step of 1.0 Da and 10-ms dwell/step. Argon was used as the collision gas with a calculated collision energy of approximately 28 eV. Parent ion transmission was maximized by reducing the resolution of quadrupole 1 to transmit approximately a 2-3 m/z window about the selected parent ion.
Determination of the Stoichiometry of dUTPase Protein Phosphorylation-Procedures for determining the extent of dUTPase phosphorylation followed those of Sefton (3). Briefly, HeLa cells were labeled for 24 h with [ 32 P]orthophosphate. dUTPase protein was recovered by immunoprecipitation and fractionated by SDS-PAGE. dUTPase was quantitated by Coomassie staining of the gel and comparison to recombinant dUTPase protein standards run on the same gel. The 32 P-labeled dUTPase protein band was excised, and the amount of associated radioactivity was determined by scintillation counting. The specific activity of 32 P was determined by counting an aliquot of the media and dividing by the mol of inorganic phosphate in the media (based on fetal calf serum containing 3.8 Ϯ 0.7 mM inorganic phosphate; Life Technologies, Inc.). The mol of phosphate/mol of dUTPase protein was then calculated.

RESULTS
The Nuclear Form of dUTPase Is Phosphorylated-We have previously demonstrated that human dUTPase is a serine phosphoprotein (1). In order to determine if one or both of the dUTPase isoforms are phosphorylated in vivo, HeLa cells were labeled with [ 32 P]orthophosphate, and dUTPase was immunoprecipitated from total cell extracts. The immunoprecipitates were fractionated by 15% SDS-PAGE and subjected to Western blot analysis. The immunoblot shown in Fig. 1, lane 1, demonstrates the presence of both DUT-N and DUT-M isoforms. After quenching the chemiluminescent immunoblot reaction, the nitrocellulose membrane was exposed to x-ray film to detect [ 32 P]orthophosphate labeling. As seen in Fig. 1, lane 2, all of the radioactivity is associated with the lower molecular weight nuclear form of dUTPase (DUT-N) and not the mitochondrial isoform (DUT-M).
Identification of Ser-11 as the in Vivo Phosphorylation Site of dUTPase-HeLa cells were radiolabeled with [ 32 P]orthophos-FIG. 1. The nuclear form of dUTPase is phosphorylated. HeLa S3 cells were labeled with [ 32 P]orthophosphate and dUTPase was immunoprecipitated, utilizing dUTPase-specific monoclonal antibodies, from total cell extracts. The immunoprecipitates were fractionated by SDS-PAGE and transferred to nitrocellulose. dUTPase protein was detected by immunoblot analysis utilizing affinity-purified, dUTPasespecific polyclonal antibody. The protein bands were visualized using a chemiluminescent detection system. Lane 1, immunoblot analysis demonstrates the presence of the major nuclear form of dUTPase (DUT-N) and the larger, less abundant mitochondrial isoform (DUT-M). The chemiluminescent detection reaction was quenched by exposing to visible light, and the membrane was subjected to autoradiography to detect the 32 P labeling. Lane 2, autoradiograph demonstrating the phosphorylation of DUT-N. There is no detectable phosphorylation associated with the DUT-M isoform.
phate, and dUTPase protein was purified by immunoaffinity chromotography (see "Experimental Procedures"). Purified 32 Plabeled dUTPase was fractionated by 15% SDS-PAGE, and the protein band corresponding to DUT-N was cut out. The labeled dUTPase was then subjected to tryptic digestion, and fractionated by reversed-phase HPLC. A single peak of radioactivity was resolved at 23.5 min during a linear acetonitrile gradient (data not shown). The phosphopeptide was refractionated and subjected to mass spectrometry analysis.
The number of phosphate groups and their sequence locations in the phosphopeptide present in the major radioactive fraction was established by electrospray mass spectrometry (5-7). The major peptide signal observed had a determined M r of 1638.0. Tandem mass spectrometry (Ref. 7 and references therein) of the (M ϩ 2H) 2ϩ parent ion of the phosphorylated peptide provided the partial sequence PCSEETPAXpSP-, where pS is phosphoserine, Cys is carboxamidomethylated, and X is either Leu or Ile, which cannot be distinguished in this experiment ( Fig. 2; see legend for an explanation of the fragmentation). This subsequence corresponds to residues 2-12 of the mature protein (Fig. 4). The molecular weight calculated for the tryptic peptide Pro-2-Arg-15 is 1637.7. This is in close agreement with the predicted molecular weight for this peptide. These data establish the identity of this tryptic peptide as residues 2-15 of the protein in which Ser-11 is phosphorylated.
Mutagenesis of Ser-11 Prevents Phosphorylation in Vivo-In order to confirm the identity of the serine phosphorylation site, Ser-11 was changed to Ala by site-directed mutagenesis. The wild type DUT-N coding region and the Ser-11 3 Ala mutant were cloned into the eukaryotic expression vector pEUK-C1, and both constructs were individually transfected into COS-7 cells. After 60 h, the transfected COS-7 cells were harvested, and extracts were analyzed by immunoblot analysis (Fig. 3A). Expression of both the wild type and mutant dUTPase is evident (Fig. 3A, lanes 2 and 3), while there is no expression from the negative control (vector containing no insert) (Fig. 3A,  lane 1).
To further confirm that Ser-11 is the site of DUT-N phosphorylation, transfected cells were labeled with [ 32 P]orthophosphate for 10 h at 50 -60 h posttransfection. At 60 h, cells were harvested and extracts were prepared. Each sample was subjected to immunoprecipitation analysis using a human dUTPase-specific monoclonal antibody. The immunoprecipitates were resolved by 15% SDS-PAGE and visualized by autoradiography. The wild type DUT-N protein is readily phosphorylated in COS-7 cells (Fig. 3B, lane 2). Phosphorylation of the mutant DUT-N however, is blocked by the Ser-11 3 Ala mutation (Fig.  3B, lane 3). The monoclonal antibody used in this experiment was determined to quantitatively immunoprecipitate both the wild type and mutant forms of dUTPase (data not shown). These data further indicate that Ser-11 is the sole phosphorylation site of the nuclear form of human dUTPase in vivo.
Estimation of Stoichiometry of dUTPase Phosphorylation by Biosynthetic Labeling-The stoichiometry of dUTPase phosphorylation was determined by the procedure described by  (20,21). The numbering above the single-letter code sequence refers to y n ions formed by cleavage of the peptide bond of the nth amino acid from the COOH terminus with H-rearrangement to form a charged, COOH-terminal peptide fragment (NH 2 -CHR n -CO . . . NH-CHR1-CO 2 H ϩ H) ϩ . The numbering below the single-letter code sequence refers to b n ions formed by charge retention on the NH 2 -terminal acylium fragments (NH 2 -CHR1-CO . . . NH-CHR n CO ϩ ); loss of CO from the b n ions yields the a n ion series. The mass increment between the y 4 and y 5 ions (167 Da), and the presence of satellite peaks formed by the loss of either H 3 PO 4 (Ϫ98 Da) on all of the y n ions that are formed by cleavage COOH-terminal to the Pro-11 indicate that the phosphate is located on Ser-10.

FIG. 3. In vivo verification of phosphorylation on Ser-11 by site-directed mutagenesis.
A, Ser-11 was changed to Ala by sitedirected mutagenesis. The dUTPase native and mutant open reading frames were cloned into the eukaryotic expression vector pEUK-C1. COS-7 cells were transfected with these constructs and control vector (with no insert) using Lipofectin. Cells were harvested after 60 h. Immunoblot analysis utilizing affinity-purified polyclonal antibodies demonstrate the transient expression of native and mutant dUTPases. Lane 1, pEUK-C1 alone; lane 2, pEUK-C1/native dUTPase; lane 3, pEUK-C1/Ser-11 to Ala mutant dUTPase. B, COS-7 were transfected as above, and cells were labeled with [ 32 P]orthophosphate between 50 and 60 h posttransfection. At 60 h cells were harvested, and dUTPase was immunoprecipitated from total cell extracts. Immunoprecipitates were washed and fractionated by SDS-PAGE. The dried gel was exposed to x-ray film for 12 h at Ϫ80°C. Lane 1, pEUK-C1, no insert; lane 2, pEUK-C1/native dUTPase; lane 3, pEUK-C1/Ser-11 3 Ala mutant. Both the monoclonal and polyclonal antibodies to human dUTPase do not cross-react with COS-7 derived dUTPase. Sefton (3). Based on the mass spectrometry and site-directed mutagenesis data, the DUT-N protein contains a single phosphorylation site. The relative stoichiometry of dUTPase phosphorylation was estimated to be 0.81 (Ϯ 0.14) mol of phosphate/ mol of DUT-N protein. These data indicate that the majority of DUT-N protein is phosphorylated in HeLa cells.
Amino Acid Conservation among dUTPases: Identification of a Nonconserved Serine Phosphorylation Site-Amino acid sequence alignment between dUTPases from various sources (Fig. 4) illustrates five characteristic sequence motifs (enclosed in boxes) common to dUTPases (8). Analysis of the Escherichia coli dUTPase crystal structure suggests that the residues involved in dUTP binding and catalysis are contained within the five conserved motifs. Several serine residues in the human dUTPase amino acid sequence represent highly conserved amino acids when compared with other known dUTPases. The two most highly conserved (Ser-86 and Ser-160) are contained within motifs II and V. The DUT-N phosphorylation site, Ser-11, is located in the nonconserved, amino-terminal region of the enzyme (Fig. 4).
Although DUT-N and DUT-M differ in their NH 2 termini, the site of Ser phosphorylation is retained in both isoforms (Fig. 4). Despite this conservation, only the DUT-N isoform is phosphorylated at any detectable levels. Mass spectrometry analysis of the analogous tryptic peptide derived from DUT-M shows no change in mass indicative of a phosphorylated residue (22).
dUTPase Is Phosphorylated at a Consensus Cyclin-dependent Protein Kinase Site: in Vitro Phosphorylation of Ser-11 by p34 cdc2 -In an effort to identify the protein kinase responsible for dUTPase phosphorylation, the amino acid context of Ser-11 was inspected for consensus target motifs. The context of Ser-11 (pSPSK) conforms to the consensus target sequence of the cyclin-dependent kinase, p34 cdc2 ((pS/pT)PX(R/K), where X is a polar amino acid (9)). p34 cdc2 is a member of the cyclin-dependent kinases (CDKs), which are known to govern the ordered events of the cell cycle via phosphorylation of specific target substrates.
To examine the authenticity of the consensus CDK phosphorylation site, in vitro kinase assays were performed to determine if DUT-N could be phosphorylated specifically on Ser-11 by p34 cdc2 . Polyclonal antibodies generated against the carboxyl terminus of human p34 cdc2 were used to immunoprecipitate protein kinase activity from HeLa S3 cells. The immunoprecipitates were washed, and recombinant DUT-N protein (baculovirus-expressed) was added to in vitro kinase assays. The samples were subsequently fractionated by SDS-PAGE, and protein phosphorylation was detected by autoradiography. Fig. 5, lane 1, illustrates the in vitro phosphorylation of recombinant dUTPase by the immunoprecipitated p34 cdc2 protein kinase. Fig. 5, lane 2, demonstrates that dUTPase phosphorylation can be specifically blocked by the addition of the competing peptide to the immunoprecipitation reaction, verifying the identity of the immunoprecipitated kinase utilized in the in vitro assays. To confirm the specificity of phosphorylation in vitro, a recombinant Ser-11 3 Ala mutant was utilized in an identical kinase reaction. Fig. 5, lane 3, illustrates that the Ser-11 3 Ala mutant specifically prevents phosphorylation by p34 cdc2 , documenting the site of DUT-N phosphorylation by p34 cdc2 in vitro. Together, these experiments demonstrate that the recombinant DUT-N protein is phosphorylated by p34 cdc2 specifically on Ser-11 in vitro.
Effect of dUTPase Phosphorylation on Enzymatic Activity and Monomer Association-It was previously postulated that dUTPase phosphorylation may regulate enzymatic activity (1). In order to directly test this hypothesis, recombinant proteins corresponding to the wild type DUT-N and a Ser-11 3 Ala mutant were subjected to kinetic analysis. The genes encoding these proteins were cloned into the eukaryotic expression vector pEUK-C1, and the protein was expressed in COS-7 cells by transient transfection. As demonstrated in Fig. 3, COS-7 cells are capable of phosphorylating recombinant human DUT-N protein in vivo. The resulting dUTPase protein was purified by the methods described under "Experimental Procedures." Both the wild type and mutant forms of DUT-N were analyzed by  5. Phosphorylation of dUTPase by p34 cdc2 in vitro. Recombinant DUT-N protein was phosphorylated in vitro with immunoprecipitated p34 cdc2 from HeLa cells as described under "Experimental Procedures." DUT-N was subsequently immunoprecipitated with a monoclonal antibody and fractionated by SDS-PAGE. The dried gel was exposed to x-ray film for 12 h. Lane 1, recombinant DUT-N protein; lane 2, DUT-N protein, competition experiment with p34 cdc2 peptide; lane 3, Ser-11 3 Ala mutant. All reactions were carried out under identical conditions. enzyme assay and shown to be fully functional, possessing identical K m values of 2.5 M (data not shown). These values are in close agreement with previous determinations of the K m for the native DUT-N and DUT-M enzymes derived from HeLa cells (22). In addition, apparent velocities at saturating dUTP concentrations were similar for both the wild type and Ser-11 3 Ala mutant. Specific activity in each case is about 15 nmol of dUMP formed per min per g of purified protein. These experiments suggest that phosphorylation of Ser-11 does not significantly alter the enzymatic activity of the DUT-N protein in vitro.
We previously hypothesized that phosphorylation may regulate subunit association (1). Both the wild type and Ser-11 3 Ala mutant were assayed for the ability to undergo magnesium-dependent multimerization as described previously (2). Both the wild type and mutant forms of the DUT-N protein formed higher molecular weight complexes, demonstrating that monomer association is independent of the phosphorylation state of the protein (data not shown).

DISCUSSION
Nuclear and Mitochondrial dUTPase: Structural Differences-Data presented in the previous paper (22) indicate DUT-N and DUT-M differ only in a short region of their amino termini (see Fig. 4 of this report for comparison). Specifically, DUT-M contains an additional 19 amino acids. This presumably accounts for the higher molecular weight compared with DUT-N. In addition, the next 5 amino acids of DUT-N (Pro-20 through Pro-24) are different from DUT-M. The remaining amino acid sequence appears to be identical between DUT-N and DUT-M. In this study we further distinguish between these two isoforms on the basis of posttranslational phosphorylation. A single phosphorylation site correlating to Ser-11 of the nuclear isoform was identified. Stoichiometry analysis of phosphorylation indicates that approximately 81% of the DUT-N protein is phosphorylated in HeLa cells. Although both DUT-N and DUT-M contain the identical consensus phosphorylation site, only the nuclear form of dUTPase is phosphorylated in HeLa cells. A possible explanation for the exclusive phosphorylation of the DUT-N isoform is that the relevant protein kinase is sequestered away from the mitochondrial form. It is conceivable that the kinases acting on DUT-N are nuclear specific or are active only in the nucleus. There is evidence that p34 cdc2 requires specific associations with cyclin proteins to become functionally active (10). Association of p34 cdc2 with cyclin A occurs in the nucleus upon cell cycle-dependent nuclear transport of this cyclin during S phase (11). These findings imply that the protein kinase activity of CDKs is in part dependent on the localization of the relevant cyclins. Additional experimentation will be required to understand the implication of these observations with respect to dUTPase phosphorylation.
The Nonconserved dUTPase Phosphorylation Site Is a Consensus p34 cdc2 Sequence-There are five highly conserved motifs common to dUTPases (8) that are implicated in active site function. Two of these characteristic motifs (II and V) contain highly conserved serine residues. However, the DUT-N phosphorylation site lies in the nonconserved amino-terminal region of the protein. It is possible that the human dUTPase phosphorylation site is a relatively recent adaptation of this enzyme, performing a role that is unique to higher eukaryotes. It is presently unknown whether dUTPases from other mammalian or higher eukaryotes also posses phosphorylated residues.
The amino acid context of the DUT-N phosphorylation site corresponds to the consensus target sequence of the cyclin-dependent kinase p34 cdc2 (9). In order to determine that the Ser-11 phosphorylation site was authentic, we performed in vitro experiments demonstrating that dUTPase is specifically phosphorylated on Ser-11 by immunoprecipitated p34 cdc2 from HeLa S3 cells (Fig. 5). However, these experiments do not confirm that p34 cdc2 directly phosphorylates dUTPase in vivo. There is a family of related kinases that share extensive homology with p34 cdc2 (e.g. CDK2 through CDK5) (12). It is possible that DUT-N is a target for one or several of these p34 cdc2related kinases in vivo.
The Functional Role of dUTPase Phosphorylation-Protein phosphorylation/dephosphorylation controls a wide range of cellular functions including aspects of cell cycle control and signal transduction (13,14). Although relatively few CDK substrates have been identified, the functional significance of phosphorylation by CDKs appears to be diverse, including promoting protein complex formation and modulating enzymatic activity. For example, CDK phosphorylation of pp60 c-src is correlated with a 3-7-fold increase in pp60 c-src tyrosine kinase activity (15). Phosphorylation of lamins by p34 cdc2 plays a major role in lamina disassembly (16). In addition, it has been suggested that the phosphorylation of cyclins, which complex with CDKs such as p34 cdc2 , affects the regulation of kinase activity (17).
The role of dUTPase phosphorylation remains to be elucidated. In a previous report, we postulated that dUTPase phosphorylation may regulate enzymatic activity (1). Several lines of evidence, presented in this work, suggest that phosphorylation does not regulate the enzymatic activity of DUT-N. Mutagenesis of Ser-11 3 Ala prevents phosphorylation; however, experiments with the mutant protein demonstrate that enzymatic activity is not significantly altered in vitro. Second, a recombinant human dUTPase protein has been expressed that lacks the first 22 amino acids present in the nuclear form of dUTPase. This recombinant protein does not contain the phosphorylation site yet still retains full enzymatic activity (18). A third line of evidence arguing against the regulation of dUTPase activity by phosphorylation comes from a comparison of the enzymatic activities of the nuclear and mitochondrial isoforms. DUT-M contains the identical site for phosphorylation as DUT-N but is not phosphorylated in vivo. Although DUT-M is not phosphorylated, it exhibits identical kinetic characteristics (K m ϭ 2.5 M) to the phosphorylated nuclear form. Taken together, these observations suggest that the phosphorylation of dUTPase does not significantly govern enzymatic activity under the assay conditions utilized in vitro.
Another possible role of phosphorylation is the formation of the dUTPase multimer. Caradonna and Adamkiewicz (2) first described human dUTPase as a homodimer, although molecular modeling of human dUTPase based on the E. coli dUTPase crystal structure (19) suggests that the human protein is a homotrimer. 2 Experiments utilizing both the DUT-N recombinant and the Ser-11 3 Ala mutant demonstrate that multimerization is independent of Ser-11 phosphorylation. Additional evidence supporting this again comes from a truncated recombinant form of dUTPase lacking 22 amino-terminal residues. This protein was shown to trimerize independent of the Ser-11 phosphorylation site (18).
DUT-N phosphorylation may also regulate its intracellular localization. The DUT-N and DUT-M isoforms differ exclusively in their amino termini. This distinction appears to confer the ability of DUT-M to localize in the mitochondria. It is conceivable that the exclusive phosphorylation of DUT-N may play a role in nuclear targeting of this protein. Taken a step further, Ser-11 may confer the ability of DUT-N to localize in specific regions of the nucleus where the dUTPase function is 2 C. Tucker and S. Ealick, personal communication.
required. The Ser-11 3 Ala mutant should aid in the testing of these hypotheses.
In summary, we have continued our investigation of the detailed biochemistry of dUTPase in human cells, uncovering an additional layer of detail. Elucidation of two distinct isoforms localized to the mitochondria and nucleus, respectively, as well as identification of a CDK phosphorylation site specific to the nuclear isoform, all suggest that this enzyme function is highly regulated within the cell.