Specificity of Natural and Artificial Substrates for Human Cdc25A

Specificity of Natural and Artificial Substrates for Human Cdc25A

Analytical Biochemistry 289, 43–51 (2001) doi:10.1006/abio.2000.4906, available online at http://www.idealibrary.com on Specificity of Natural and Ar...

162KB Sizes 0 Downloads 11 Views

Analytical Biochemistry 289, 43–51 (2001) doi:10.1006/abio.2000.4906, available online at http://www.idealibrary.com on

Specificity of Natural and Artificial Substrates for Human Cdc25A Johannes Rudolph,* ,1 David M. Epstein,* ,2 Laura Parker,† and Jens Eckstein* ,3 *Mitotix, Inc., Cambridge, Massachusetts; and †BASF Bioresearch Corp., Worcester, Massachusetts

Received July 5, 2000

Cdc25A is a dual-specific protein phosphatase involved in the regulation of the kinase activity of Cdk– cyclin complexes in the eukaryotic cell cycle. To understand the mechanism of this important regulator, we have generated highly purified biochemical reagents to determine the kinetic constants for human Cdc25A with respect to a set of peptidic, artificial, and natural substrates. Cdc25A and its catalytic domain (dN25A) demonstrate very similar kinetics toward the artificial substrates p-nitrophenyl phosphate (k cat/K m ⴝ 15–25 M ⴚ1 s ⴚ1) and 3-O-methylfluorescein phosphate (k cat/K m ⴝ 1.1–1.3 ⴛ 10 4 M ⴚ1 s ⴚ1). Phospho-peptide substrates exhibit extremely low second-order rate constants and a flat specificity profile toward Cdc25A and dN25A (k cat/K m ⴝ 1 to 10 M ⴚ1 s ⴚ1). In contrast to peptidic substrates, Cdc25A and dN25A are highly active phosphatases toward the natural substrate, T14- and Y15bis-phosphorylated Cdk2/CycA complex (Cdk2-pTpY/ CycA) with k cat/K m values of 1.0 –1.1 ⴛ 10 6 M ⴚ1 s ⴚ1. In the context of the Cdk2-pTpY/CycA complex, phosphothreonine is preferred over phospho-tyrosine by more than 10-fold. The highly homologous catalytic domain of Cdc25c is essentially inactive toward Cdk2-pTpY/ CycA. Taken together these data indicate that a significant degree of the specificity of Cdc25 toward its Cdk substrate resides within the catalytic domain itself and yet is in a region(s) that is outside the phosphate binding site of the enzyme. © 2001 Academic Press

1 To whom reprint requests should be addressed at current address: Department of Biochemistry, Duke University Medical Center, Box 3813, Durham, NC 27710. Fax: (919) 613-8642. E-mail: [email protected]. 2 Current address: Department of Cancer and Osteoporosis Research, Bayer Pharmaceutical, West Haven, CT. 3 Current address: Enanta Pharmaceuticals, Cambridge, MA.

0003-2697/01 $35.00 Copyright © 2001 by Academic Press All rights of reproduction in any form reserved.

The eukaryotic cell cycle is tightly regulated by the action of the cyclin-dependent kinases (Cdks) 4 (1, 2). The kinase activity of these complexes is directly responsible for the initiation and progression of successive phases of the cell cycle by modification of proteins required for the activation of genes involved in DNA synthesis and in the structural reorganization of the cell during mitosis (3). Therefore, reversible regulation of Cdk kinase activity is key to controlling the cell cycle. Monomeric Cdks are inactive and associate with the appropriate cyclin partner to form a heterodimeric complex (i.e., Cdc2/CycB, Cdk2/ CycA). Kinase activation is brought about by phosphorylation of Thr160 in the T-loop of the Cdk/cyclin complexes by the Cdk-activating kinase (CAK) (4). Additional Cdk/cyclin regulation consists of both association with inhibitory proteins (i.e., members of the INK4 and WAF1/CIP1 inhibitor classes 5) and a series of phosphorylation and dephosphorylation events that take place on two residues, Thr14 and Tyr15 in both Cdc2 and Cdk2. These two residues are located within the sequence motif GEGTYG in the conserved glycine-rich loop (Gly-loop) of the Cdk catalytic subunit. Phosphorylation of Thr14 and Tyr15 is mediated by the protein kinases Wee-1 (6, 7) and Myt-1 (8). Dephosphorylation of these residues by the dual-specificity protein phosphatase Cdc25 generates an active kinase complex (reviewed by 9 –11). Hence, Thr14 and/or Tyr15 dephosphorylation appears to be the trigger for a number of cell cycle transitions. 4 Abbreviations used: Cdk, cyclin-dependent kinase; CAK, Cdk-activating kinase; PTPase, protein-tyrosine phosphatase; DSP, dual-specific phosphatase; TPCK, tosyl-L-phenylalanine chloromethyl ketone; TLCK, tosyl-L-lysine chloromethyl ketone; PIN, protease inhibitor cocktail; IPTG, isopropyl ␤-D-thiogalactoside; DTT, dithiothreitol; GST, glutathione S-transferase; ATP, adenosine triphosphate; pNPP, p-nitrophenyl phosphate; MFP, O-methyl fluorescein phosphate; TFA, trifluoroacetic acid; TCA, trichloroacetic acid; GSH, glutathione (reduced); VHR, vaccinia H1-related phosphate; PVDF, polyvinylidene fluoride.

43

44

RUDOLPH ET AL.

Three homologs of Cdc25 exist in humans, Cdc25A, -B, and -C (12–14). The Cdc25 proteins are approximately 500 amino acids long. The N-termini have low sequence homology (identity from 20 –25%) and contain numerous phosphorylation sites as well as nuclear import/export signals. The catalytic cores of the Cdc25s are located in the homologous C-termini (⬃60% identity over 200 amino acids). The Cdc25 phosphatases belong to a class of phosphatases that include the protein-tyrosine phosphatases (PTPases) and the dual-specific phosphatases (DSPs) (15, 16). These phosphatases contain the active site motif HCX 5 R, where H is a highly conserved histidine residue, C is the catalytic cysteine, the X 5 motif forms a loop wherein the five amide nitrogens hydrogen-bond to the phosphate of the protein substrate, and R is the conserved arginine that binds the phosphoamino acid of the substrate through two additional hydrogen bonds. Like other DSPs, Cdc25 phosphatases display no sequence identity with other PTPases outside of this signature motif. The structures of the catalytic domains of Cdc25A and -B have recently been solved to high resolution (17, 18). They revealed that the Cdc25 catalytic domains have a completely different fold than other PTPases or DSPs, having a topology similar to that of the sulfur transfer protein rhodanese. Interestingly, the architecture surrounding the active site was found to be solvent exposed and open. Also, the structures did not reveal an auxiliary loop containing the expected catalytic acid residue known from other PTPases and DSPs. Several lines of evidence confirm the essential role of Cdc25C-mediated dephosphorylation of Cdc2/CycB at the G2/M transition (12, 19 –21). Cdc25C appears to be the key mitotic regulator at the G2/M transition, yet it is Cdc25A and Cdc25B that have been implicated in cancer (22–25). However, the substrates of Cdc25A and Cdc25B in the dephosphorylation of Cdks during cell cycle progression are not well defined. Data suggest that Cdc25A acts on phosphorylated Cdk2/CycE and/or Cdk2/CycA in G1/S (13, 26, 27) and in the control of Cdk4 and Cdk6 phosphorylation mediated by TGF-␤ (28). The substrates of Cdc25B are postulated to be the Cdk2/CycE and/or Cdk2/CycA complexes (29). Thus, while it is known that Cdc25s play a vital role in cell cycle control, little is known about their substrate specificity or catalytic activity. Herein we report the purification of native Cdc25A, its catalytic domain ⌬N25A, and the preparation of bona fide T14- and Y15-phosphorylated substrate Cdk2-pTpY/CycA. We show that Cdk2/CycA is an efficient substrate for Cdc25A, especially when compared with phosphorylated peptides, and that the catalytic domains of the Cdc25 phosphatases appear to contain crucial specificity determinants for substrate recognition.

MATERIALS AND METHODS

Reagents. The protease inhibitors tosyl-L-phenylalanine chloromethyl ketone (TPCK), N-␣-tosyl-L-lysine chloromethyl ketone (TLCK), and a broad-spectrum protease inhibitor cocktail (PIN) were obtained from Boehringer Mannheim. IPTG was purchased from Gibco/BRL. DEAE, SP-Sepharose, and S-200 Superdex chromatography resins were obtained from Pharmacia. Peptides were purchased from the peptide synthesis facility at Tufts Medical School and were HPLC purified prior to use. All other reagents were obtained from Sigma/Aldrich and were of the highest purity available. Preparation of Cdc25A. The open reading frame containing human cdc25A was cloned into pET-3d (Novagen) by incorporation of a NcoI site at the start codon and a HinDIII site following the stop codon. The resulting plasmid was transformed into BL-21 (DE3) and the protein was overproduced by induction of mid-log cells with 0.5 mM IPTG for 3 h at 25°C. All steps in the purification were performed at 4°C. In a typical preparation, 40 g of frozen cell pellets was thawed in 200 mL of buffer A, pH 8.0 (25 mM Tris–HCl, 25 mM Bis-Tris-propane, 1 mM EDTA, 1 mM DTT, and the protease inhibitors PIN, TPCK, and TLCK at 0.1 mg/ mL). Cells were lysed by mild sonication (3 ⫻ 30 s) and freeze–thawing. The lysate was cleared by centrifugation at 25,000g for 30 min. Cdc25A was precipitated by addition of ammonium sulfate to 20% saturation and the protein was collected by centrifugation at 25,000g for 30 min. Following resuspension of the pellet in 200 mL of buffer A, undissolved material was removed by another centrifugation. The protein was then passed over a DEAE–Sepharose column (32 mL) and the column was washed with an additional 75 mL of buffer A at pH 8.0 containing 100 mM NaCl. The pooled flowthrough and wash were then carefully adjusted to pH 6.5 by the addition of HCl. Cdc25A was then bound in batch to 10 mL of SP-Sepharose for 1 h. Following two 50-mL washes of the column with buffer A at pH 6.5 containing 100 mM NaCl, the protein was eluted with 50 mL of buffer A at pH 6.5 with 200 mM NaCl. Following adjustment of the pH to 8.0, the protein was subjected to concentration and S-200 Superdex chromatography in buffer A at pH 8.0 containing 200 mM NaCl. The final yield was 5–10 mg of purified protein per preparation with a specific activity of 42 nmol/ min/mg using 70 mM pNPP at 25°C. Preparation of ⌬N25A. Cloning and expression of the open reading frame for the catalytic domain ⌬N25A was performed as for the full-length protein with the placing of a methionine N-terminal to the sequence GTIENI and continuing to the natural stop codon. All steps in the purification were performed at 4°C and phosphatase activity was followed by assays using

SUBSTRATE SPECIFICITY FOR HUMAN Cdc25A

pNPP as a substrate. In a typical preparation, 33 g of frozen cell pellets were thawed in 150 mL of buffer B (3 mM potassium phosphate, pH 7.4, 75 mM NaCl, 1 mM EDTA, 1 mM DTT, and the protease inhibitors PIN, TPCK, TLCK at 0.1 mg/mL). The cells were lysed by sonication (3⫻, 1 min). Following centrifugation at 25,000g for 30 min. the cleared lysate was bound to 15 mL of SP-Sepharose equilibrated in buffer B. dN25A was eluted with buffer B containing 150 mM NaCl and further purified by S-200 chromatography in phosphatase reaction buffer (see below). The final yield was 85 mg of purified protein per preparation with a specific activity of 120 nmol/min/mg using 70 mM pNPP at 25°C. Preparation of other phosphatases: VHR, PTP1b, dN25B, and dN25C. dN25B, the catalytic domain of Cdc25B, and dN25C, the catalytic domain of Cdc25C, were prepared as described previously (41). dN25B and dN25C have specific activities of 130 and 90 nmol/min/ mg, respectively, using 70 mM pNPP at 25°C. VHR was purified from an Escherichia coli overexpression strain to 90% homogeneity as described (30). PTP1b was purified from an E. coli overexpression strain to 95% homogeneity as described (31). Preparation of Cdk2-pTpY/CycA. The Cdk2/CycA complex, in which the CycA is truncated and encompasses residues 174 – 432, was prepared as described for its crystallization (32). The purity, as assessed by SDS–PAGE, was ⬎95% and indicated a 1:1 complex. Myt-1 was cloned and prepared as follows: A 1.8-kb fragment containing full length myt-1 c-DNA was amplified from lambda phage DNA prepared from a Xenopus laevis oocyte c-DNA library (Clontech) using primers of the sequence (amino terminal) GGCCCGGGATGCCTGTTCCAGGGG and (carboxy terminal) GGCCCGGGGTCATGGCGATATCATGAA. The myt-1 cDNA was ligated into the PCRII cloning vector to generate PCRIImyt-1 and sequenced. PCRIImyt-1 was then digested with SmaI and the myt-1 cDNA was ligated into SmaI-cut pACG2T. The resulting plasmid was transfected into insect cells using the Baculogold transfection system (Invitrogen). Expression was optimized by plaque purification of the recombinant baculovirus and all steps in the purification were performed at 4°C. The frozen cells from a 4-L infection were thawed in 50 mL lysis buffer (10 mM Hepes (pH 7.4), 150 mM NaCl, 5 mM EGTA, 0.5% Triton-X100, and the protease inhibitors PIN, TPCK, and TLCK at 0.1 mg/ mL) and subjected to mild sonication (3 ⫻ 5 s). GSTMyt-1 was batch-bound to 5 mL of GSH–Sepharose. Following washes with 20 vol of lysis buffer containing 1 M NaCl, the column was washed with 20 vol of kinase buffer (50 mM Tris–HCl (pH 7.5), 100 mM NaCl, 10 mM MgCl 2, and 0.1% Triton X-100). The protein was

45

stored in kinase buffer at ⫺70°C bound to the GSH beads following freezing by liquid N 2 immersion. Phosphorylation of the Cdk2/CycA complex was carried out for 2.5 h at 25°C in kinase buffer containing 3.3 mM [␥- 32P]ATP (200 Ci/mol) and 0.5 equivalents of GST-Myt-1 for each equivalent of Cdk2/CycA as measured using the BioRad protein assay solution. Following the phosphorylation reaction, GST-Myt-1 bound to the GSH beads was removed by centrifugation at 1000g for 5 min. Unincorporated ATP was removed by G-50 chromatography (1 ⫻ 15 cm) in 40 mM Hepes (pH 7.5), 200 mM NaCl, and 5 mM DTT. Wee-1 was prepared as described (7) and Y15 phosphorylations of Cdk2/CycA using Wee-1 were performed analogously to the Myt-1 reactions with the following differences. No detergent was used in the phosphorylation reaction, which contained 300 ␮M ATP instead of the 3.3 mM used in the Myt1 reaction. Phosphatase reactions. All phosphatase reactions were performed in either three-component buffer at pH 7.0 (100 mM Na-acetate, 50 mM Tris, 50 mM Bis–Tris) or in 50 mM Tris–HCl (pH 8.0), 50 mM NaCl, 1 mM EDTA, and 1 mM DTT at 25°C. Reactions using pnitrophenyl phosphate (pNPP, Sigma) and 3-O-methyl fluorescein phosphate (mFP) (33) were followed by continuous UV–Vis spectroscopy at 410 nm (␧ ⫽ 18,000 M ⫺1 cm ⫺1) and 477 nm (␧ ⫽ 27,200 M ⫺1 cm ⫺1), respectively. Typical enzyme concentrations in steady-state assays using mFP were 1–5 nM, whereas 1–5 ␮M of enzyme was used in the burst kinetics experiments that were performed and analyzed as described previously (41). Phospho-peptide reactions were typically performed using 0.2 to 1 ␮M enzyme. Product formation in these reactions were either followed by continuous monitoring of the increase of absorbance at 282 nm (␧ ⫽ 280 M ⫺1 cm ⫺1) (34) or by HPLC analysis of fixed time points (reverse phase liquid chromatography using an acetonitrile gradient in 0.1% TFA) (35). Dephosphorylation reaction of phospho-peptide substrates was verified in each case by LC-MS analysis (LCQ system). Reactions with the natural substrate were performed using 0.2 to 2 nM enzyme in the presence of 1 mg/mL bovine serum albumin and fixed timepoints were quenched by the addition of 0.3 equivalents (by volume) of 30% TCA. The supernatant containing the released phosphate was subjected to scintillation counting following centrifugation of the precipitated protein at 14 K and 4°C for 10 min. The precipitated protein was subjected to phospho-amino acid analysis as described by (36). RESULTS AND DISCUSSION

Preparation and Initial Characterization of Enzymes Previous attempts at expression and purification of soluble native full-length Cdc25 has proven difficult to

46

RUDOLPH ET AL.

FIG. 1. Coomassie-stained SDS–PAGE monitoring the purification of full-length (Cdc25A full-length) and catalytic domain (dN25A) of Cdc25A as expressed in E. coli. The steps in the purifications are described in detail in the experimental section and the molecular weight markers (MW) are labeled on the right side of the gel.

date, thereby preventing quantitative biochemical characterization of these dualspecific protein phosphatases. Initial attempts using bacterially expressed human or Drosophila Cdc25A required that the enzyme be renatured from inclusion bodies and showed low phosphatase activities (i.e., with k cat/K m s of ⬍1 M ⫺1 s ⫺1 using pNPP, whereas VHR has a k cat/K m s of ⬎10 3 M ⫺1 s ⫺1) (33, 37–39). While full-length human Cdc25A and Cdc25B have been expressed as soluble GST fusion proteins (13), they have not been purified extensively or cleaved from the GST tag. Except for a recent report of the purification of full-length Cdc25C (40), only the catalytic domains of the human Cdc25s have been highly purified (33, 41, 42). We report here the expres-

sion of human Cdc25A in E. coli as a full-length native protein and its purification to greater than 90% homogeneity with 2- to 10-mg yields (Fig. 1). The protein migrates with a molecular weight of 190 kDa on S-200 gel filtration chromatography; thus Cdc25A is apparently a trimeric protein. Purified Cdc25A has a k cat value of 1.5 ⫾ 0.3 s ⫺1 and a K m value of 34 ⫾ 3 ␮M using the artificial substrate mFP. These values are similar to those obtained from the GST fusion proteins of Cdc25A and -B (33), as well as the purified GST fusion protein of the catalytic domain of Cdc25B (GSTdN1B) (41). The full-length native protein Cdc25A exhibits rapid burst kinetics using mFP as a substrate under stop-flow conditions, as does GST-dN1B, suggesting the buildup of a phospho-enzyme intermediate during turnover (Table 1). The burst size of 0.4 equivalent of enzyme was not due to partial oxidation of the active site cysteine as LC-MS analysis indicated a fully reduced protein. These burst kinetics give evidence that the active site is correctly formed and primed to catalyze dephosphorylation reactions. The purified catalytic domain of dN25A was initially identified as a stable and catalytically active product of the full-length native enzyme following limited trypsin proteolysis. N-terminal sequence was obtained by sequencing from PVDF membrane (G-T-I-E-N-I-L). Analysis of this protein by mass spectrometry indicated that the C-terminus remained intact upon trypsin treatment. Cloning and expression of this proteolytic product generated high yields (25–50 mg/L) of purified and active phosphatase (Fig. 1). Similar to full-length protein, dN25A has a k cat value of 1.3 ⫾ 0.2 s ⫺1 and a K m value of 22 ⫾ 2 ␮M using the artificial substrate mFP. Like the full-length protein, it also exhibits burst kinetics with mFP. Unlike the full-length protein, it migrates as a monomer on S-200 chromatography, indicating that the N-terminus is responsible for a pre-

TABLE 1

Specificity Constants (k cat/K m ) for Various Substrates of Cdc25A at pH 7.0 Substrate pNPP mFP N-ac-pY-NH 2 N-ac-GV-pY-NH 2 N-ac-GE-pY-NH 2 N-ac-GEGT-pY-GVV-NH 2 N-ac-GEG-pTY-GVV-NH 2 N-ac-GEG-pTpY-GVV-NH 2 Cdk2-pTpY/CycA (Myt1-phosphorylated) Cdk2-pY/CycA (Wee1-phosphorylated)

Cdc25A (M ⫺1 s ⫺1)

dN25A (M ⫺1 s ⫺1)

Comment

15 13,000 Burst amplitude ⫽ 0.3 nd nd nd 1.5 1.1 4.1 1.7 1.1 ⫻ 10 6 6.2 ⫻ 10 4 5.5 ⫻ 10 4

35 11,000 Burst amplitude ⫽ 0.4 1 1 4 1.9 1.2 10 4.5 1.0 ⫻ 10 6 6.1 ⫻ 10 3 6.1 ⫻ 10 3

Artificial substrate Artificial substrate pY pY pY pY pT pY pT pT pY pY

phosphorolysis phosphorolysis phosphorolysis phosphorolysis phosphorolysis phosphorolysis phosphorolysis phosphorolysis phosphorolysis phosphorolysis

SUBSTRATE SPECIFICITY FOR HUMAN Cdc25A

47

viously unknown role in the oligomerization of Cdc25A in vitro. Peptidic Substrates As previously reported, phospho-peptides can serve as slow substrates for Cdc25s (13, 20). We prepared short mono- and bis-phosphorylated peptides based on the consensus Cdk inhibitory phosphorylation sequence (GGEGTYGVV) to quantify the hydrolysis rates as catalyzed by Cdc25 and compare these with the rates using the other artificial substrates, pNPP and mFP. Cdc25A and dN25A dephosphorylate short mono- and bis-phosphorylated peptides; however, they are poor substrates. No saturation behavior is observed up to 10 mM and their specificity constants are extremely low (Table 1). It is of interest that peptidic substrates that contain the consensus sequence of the native substrate (i.e., G-E-G-pT-pY-G-V-V) are not significantly better than other random phospho-peptides (data not shown). Preparation and Initial Characterization of Cdk2pTpY/CycA The lack of bona fide preparation of pure, phosphorylated Cdk/cyclin complex in amounts suitable for biochemical characterization has held back mechanistic investigations of the Cdc25s. In general, phosphorylated protein substrates for Cdc25s have been prepared from native cell lysates through affinity chromatography (20, 38, 43– 45) or coexpression of kinase and cyclin in insect cells (39, 46). All of these procedures result in protein preparations that cannot be accurately quantitated or characterized for purity or phosphorylation status. In addition, published methods yield Cdk/Cyclin complexes in which the phosphorylation states of the cyclin and/or kinase (i.e., T14, Y15, T167) are variable or may contain other proteins which copurify with Cdk/cyclin (e.g., p13suc1 or members of the INK4 or WAF1/Cip1 family of inhibitors). These covalent and noncovalent modifications may affect the interaction of Cdc25 with its protein substrate (see also 37). Therefore, preparation of Cdk2-pTpY/CycA was performed in a multistep procedure that ensured the integrity and homogeneity of the substrate. The kinase Cdk2 was prepared as a monomer from insect cells infected with baculovirus containing an expression vector and was not modified by phosphorylation (expected MW, 33,975 Da; MW determined by MS, 33,971 Da). Truncated CycA, prepared from E. coli, also was monomeric and unphosphorylated (expected MW, 29,721 Da; MW determined by MS, 29,716 Da). In our hands, the Cdk2/CycA complex formed from these purified monomers was stable, could be purified as a 1:1 complex, and showed a low kinase activity that could

FIG. 2. (A) Autoradiography following SDS–PAGE of a reaction mixture containing 1 ␮g Xenopus GST-Myt1 and 3.3 mM [␥- 32P]ATP in the presence (lane 1) or absence (lane 2) of 3 ␮g of human Cdk2/ CycA demonstrating that Myt1 undergoes self-phosphorylation as well as phosphorylating Cdk2. (B) Autoradiography following SDS– PAGE of a reaction mixture containing Myt1-phosphorylated Cdk2/ CycA following no addition (lane 1), treatment with Cdc25A (lane 2), or treatment with Cdc25A (C430S) (lane 3), indicating that the Myt1-phosphorylated complex is a substrate for active Cdc25A but not for its active site mutant.

be activated 200- to 300-fold by CAK-phosphorylation (data not shown), as reported previously (47). Phosphorylation of the complex by Myt-1 was monitored and quantified by radiolabeling. In addition, we utilized mass spectrometry and phospho-amino acid analysis to verify the degree of phosphorylation and to characterize the residues phosphorylated by Myt-1. In contrast to previous reports, CAK-phosphorylation of Thr160 was not required for Myt-1 phosphorylation (8). Quantitative phosphorylation of Cdk2 was shown by the observation of two additional phosphate groups (MW, 2 ⫻ 80 Da) in the molecular weight determination by mass spectrometry (expected MW, 34,135 Da; MW determined by MS, 34,130 Da). As expected, cyclin A remained unmodified by Myt-1 phosphorylation. Phospho-amino acid analysis showed the presence of both phospho-tyrosine and phospho-threonine, consistent with phosphorylation of T14 and Y15 (see Fig. 3B). Treatment of the Cdk2-pTpY/CycA with Cdc25A (but not with the dominant negative active-site mutant C430S mutant) led to complete dephosphorylation of the complex, again consistent with Myt-1 phosphorylation of T14 and Y15 (Fig. 2B). Wee-1 phosphorylation, which was not quantitative (10 –30% yield, based on incorporation of labeled phosphate), led, as expected, to the appearance of phospho-tyrosine alone when analyzed by phospho-amino acid analysis (data not shown). From the sum of these data, we conclude that the Cdk2-pTpY/CycA we have prepared is a bona fide substrate for Cdc25, namely a stable 1:1 complex of Cdk2 and Cyclin A, which has the requisite kinase activity, can be quantitatively phosphorylated and can be dephosphorylated by Cdc25.

48

RUDOLPH ET AL.

FIG. 3. (A) Dephosphorylation reaction kinetics of Cdk2-pTpY/ CycA (100 pmol) as catalyzed by Cdc25A (3.1 pmol) in the presence of 1 mg/mL BSA. Fixed time points were quenched by the addition of 0.3 equivalents (by volume) of 30% TCA. The supernatant containing the released phosphate was subjected to scintillation counting following centrifugation of the precipitated protein at 14 K and 4°C for 10 min. Note the biphasic nature of the reaction giving a slower rate at 50% turnover (⫽ 100 pmol). (B) Phospho-amino acid analysis of time-points corresponding to the reaction in A. The radiolabeled Cdk2-pTpY/CycA was quenched in 6 M HCl for 1 h at 95°C in order to hydrolyze the protein into individual amino acids. Two-dimensional phospho-amino acid analysis was then performed by standard procedures and visualized by autoradiography. The dotted circles correspond to unlabeled standards that were added to the samples prior to electrophoresis and visualized by Ninhydrin spray.

Cdk2-pTpY/CycA Is a Good Substrate for Cdc25A The time course of a typical biphasic dephosphorylation reaction of Cdk2-pTpY/CycA catalyzed by Cdc25A is shown in Fig. 3A. The first important result to note is the fast reaction rate. Fits to the initial linear part of the reaction indicate a k cat/K m of 1.1 ⫻ 10 6 M ⫺1 s ⫺1 (see Table 1). This is 5 orders of magnitude higher than the best peptidic substrate investigated and almost 100fold higher than the best artificial substrate, mFP. The K m for the complex was estimated to be around 1 ␮M from experiments where the concentration of Cdk2pTpY/CycA was varied from 50 nM to 1 ␮M (data not shown). Thus, Cdc25A is indeed a highly potent and specific phosphatase, even though it shows such poor reactivity with peptidic substrates. This result is in striking contrast to many other enzymes that utilize protein substrates. Proteases, for example, are highly efficient at using peptidic and artificial substrates (e.g., caspases with k cat/K m of 10 3–10 6 M ⫺1 s ⫺1 (48)) and much

useful enzymology has been performed with peptidic substrates. Most protein kinases also effectively phosphorylate peptidic substrates and show substrate preference for those peptidic sequences that mimic the true phosphorylation site (e.g., protein kinase A or p38-␣ MAP kinase with k cat/K m of 10 3–10 6 M ⫺1 s ⫺1 (49, 50)). More closely related, most cysteine phosphatases, including VHR and low-molecular-weight protein tyrosine phosphatases, show high reactivity with artificial and peptidic substrates (k cat/K m of 10 5–10 6 M ⫺1 s ⫺1 (30, 51)). Thus, Cdc25, which belongs to the much smaller class of enzymes that works effectively only with protein substrates, in some way appears to require the restricted tertiary structure of the Cdk2/ CycA complex. This is especially intriguing in light of the flat and open active site region seen in the recent structural elucidations of the catalytic domains of Cdc25A and Cdc25B (17, 18) and suggests that a complete active site is not formed until the Cdc25 enzyme meets its protein substrate. The second important result to note is the biphasic nature of the complete reaction time course. The presence or addition of reaction product, i.e., unphosphorylated complex, did not affect the dephosphorylation rates discarding the possibility of product inhibition (data not shown). Therefore, this biphasic behavior is most directly explained by the presence of two substrates that are dephosphorylated at different rates. Indeed, phospho-amino acid analysis shows that both phospho-threonine and phospho-tyrosine are present at the early time points, whereas only phospho-tyrosine is detected at the later time points (Fig. 3B). This data indicate that phospho-threonine is the preferred initial target of Cdc25A (k cat/K m ⫽ 1.1 M ⫺1 s ⫺1, Table 1), leading to a mono-phosphorylated Cdk2-pY/ CycA intermediate which is subsequently dephosphorylated at a slower rate (k cat/K m ⫽ 6.2 ⫻ 10 4 M ⫺1 s ⫺1, Table 1). The 20-fold substrate preference for phosphothreonine has been confirmed by measuring the dephosphorylation rates of monophosphorylated complexes (i.e., following specific tyrosine phosphorylation using Wee-1 (k cat/K m ⫽ 5.5 ⫻ 10 4 M ⫺1 s ⫺1, Table 1). Thus, the data do not support a mechanism whereby Cdc25 functions by a progressive mechanism involving only one binding step for the removal of the two neighboring phosphates from T14 and Y15. Instead, two separate binding steps Cdc25 are required for the complete sequential conversion of Cdk2-pTpY/CycA to Cdk2-pY/CycA to Cdk2/CycA. Relative Specificity of Other Phosphatases for Cdk2pTpY/CycA Next, we investigated the specificity of other phosphatases for the Cdk2-pTpY/CycA complex. As seen in Fig. 4, Cdc25A was by far the best phosphatase, fol-

SUBSTRATE SPECIFICITY FOR HUMAN Cdc25A

49

FIG. 4. Comparative phosphatase activity measurements of various phosphatases using the substrate Cdk2-pTpY/CycA. Reaction mixtures contained 100 nM Cdk2-pTpY/CycA and varying amounts of the phosphatases. The reactions were quenched with TCA after 30 min at room temperature and the supernatants of the samples were subjected to scintillation counting following centrifugation of the precipitated protein containing unreacted starting material.

lowed closely by dN25A (see also Table 1). The similar reactivity of dN25A toward Cdk2-pTpY/CycA indicates that all of the specificity elements required for the recognition of Cdk2-pTpY/CycA are located in the catalytic domain. The “kink” in the dephosphorylation curve around 50% dephosphorylation reflects the slower rate of tyrosine dephosphorylation by dN25A. This 100-fold rate difference was quantitatively confirmed by measuring the dephosphorylation of the monophosphorylated complex (Table 1). Therefore, although the N-terminus of Cdc25A is not required for recognition of Cdk2-pTpY/CycA, it appears to be required for the recognition of Cdk2-pY/CycA, suggesting that the presentation of the phospho-threonine vs the phospho-tyrosine in the Cdk complexes toward Cdc25A may be significantly different. In an investigation of the specificity of the other Cdc25s toward the Cdk2-pTpY/CycA substrate, the catalytic domain of Cdc25B (dN25B) was found to recognize this substrate almost as efficiently as full-length Cdc25A; the catalytic domain of Cdc25C (dN25C), by contrast, was four orders of magnitude less effective in this reaction (Fig. 4). This result, in combination with the essentially equivalent activity of all three Cdc25s toward the artificial substrates pNPP and mFP, suggests that the catalytic domains themselves contain crucial recognition elements for their natural substrates. That is, given that Cdc2/CycB is the putative natural substrate for Cdc25C, it is perhaps not surprising that it does not recognize the Cdk2/CycA complex efficiently in comparison to Cdc25A. Two non-cell-cycle phosphatases were also tested for their activity toward Cdk2-pTpY/CycA. Although both the dual-specificity phosphatase VHR and the protein tyrosine phosphatase PTP1b are much more active toward pNPP and

phospho-peptides than the Cdc25s, they were found to be at least four orders of magnitude less effective at dephosphorylating Cdk2-pTpY/CycA. Thus, the phosphates on Cdk2-pTpY/CycA must be at least somewhat protected (i.e., unlike a peptide free in solution) and somehow only presentable to Cdc25A and Cdc25B. How Cdc25A (and Cdc25B) achieve the specific recognition of Cdk2/CycA is unclear at present, even in light of the recent structure of the catalytic domain of Cdc25A and Cdc25B as determined both by homology modeling and X-ray crystallography (17, 18, 52). The active site region of the Cdc25s appears to be extremely shallow and open compared to other phosphatases. There is no obvious binding pocket for the phosphoamino acid and no obvious interaction surface to facilitate complex formation with Cdk2/cyclins. Additionally, there is no auxiliary loop extending over the active site that could provide the catalytic acid presumably necessary to catalyze the dephosphorylation of phospho-threonine so efficiently. Further investigations to identify the origin of the specificity displayed by the Cdc25s are underway. ACKNOWLEDGMENTS We thank J. Dixon and N. Tonks for providing the clones for VHR and PTP1b, respectively. We thank Rong-Rong Zhu (BASF) for excellent LC-MS assistance.

REFERENCES 1. Morgan, D. O. (1995) Principles of CDK regulation. Nature 374, 131–134. 2. Nigg, E. A. (1995) Cyclin-dependent protein kinases: Key regulators of the eukaryotic cell cycle. BioEssays 17, 471– 480. 3. Nigg, E. A. (1993) Targets of cyclin-dependent protein kinases. Curr. Opin. Cell Biol. 5, 187–193.

50

RUDOLPH ET AL.

4. Solomon, M. J., Lee, T., and Kirschner, M. W. (1992) Role of phosphorylation in p34cdc2 activation: Identification of an activating kinase. Mol. Biol. Cell 3, 13–27. 5. Elledge, S. J., and Harper, J. W. (1994) Cdk inhibitors: On the threshold of checkpoints and development. Curr. Opin. Cell Biol. 6, 874 – 878. 6. McGowan, C. H., and Russell, P. (1995) Cell cycle regulation of human WEE1. EMBO J. 14, 2166 –2175. 7. Parker, L., and Piwnica-Worms, H. (1992) Inactivation of the p34cdc2– cyclin B complex by the human wee1 tyrosine kinase. Science 257, 1955–1957. 8. Mueller, P. R., Coleman, T. R., Kumagai, A., and Dunphy, W. G. (1995) Myt1: A membrane-associated inhibitory kinase that phosphorylates Cdc2 on both threonine 14 and tyrosine 15. Science 270, 86 – 89. 9. Hoffmann, I., and Karsenti, E. (1994) The role of cdc25 in checkpoints and feedback controls in the eukaryotic cell cycle. J. Cell Sci. S18, 75–79. 10. Dunphy, W. G. (1994) The decision to enter mitosis. Trends Cell Biol. 4, 202–207. 11. Draetta, G., and Eckstein, J. (1997) Cdc25 protein phosphatases in cell proliferation. Biochim. Biophys. Acta 1332, M53–M63. 12. Sadhu, K., Redd, S. I., Richardson, H., and Russell, P. (1990) Human homolog of fission yeast cdc25 mitotic inducer is predominantly expressed in G2. Proc. Natl. Acad. Sci. USA 87, 5139 – 5134. 13. Galaktionov, K., and Beach, D. (1991) Specific activation of cdc25 tyrosine phosphatases by B-type cyclins: Evidence for multiple roles of mitotic cyclins. Cell 67, 1181–1194. 14. Nagata, A., Igarashi, M., Jinno, S., Suto, K., and Okayama, H. (1991) An additional homolog of the fission yeast cdc25⫹ gene occurs in humans and is highly expressed in some cancer cells. New Biol. 3, 959 –968. 15. Denu, J. M., Stuckey, J. A., Saper, M. A., and Dixon, J. E. (1996) Form and function in protein dephosphorylation. Cell 87, 361– 364. 16. Barford, D., Das, A. K., and Egloff, M. P. (1998) The structure and mechanism of protein phosphatases: Insights into catalysis and regulation. Annu. Rev. Biophys. Biomol. Struct. 27, 133–164. 17. Fauman, E. B., Cogswell, J. P., Lovejoy, B., Rocque, W. J., Holmes, W., Montana, V. G., Piwnica-Worms, H., Rink, M. J., and Saper, M. A. (1998) Crystal structure of the catalytic domain of the human cell cycle control phosphatase, Cdc25A. Cell 93, 617– 625. 18. Reynolds, R. A., Yem, A. W., Wolfe, C. L., Deibel, M. R. J., Chidester, C. G., and Watenpaugh, K. D. (1999) Crystal structure of the catalytic subunit of Cdc25B required for G2/M phase transition of the cell cycle. J. Mol. Biol. 293, 559 –568. 19. Millar, J. B. A., Blevitt, J., Gerace, L., Sadhu, K., Featherstone, C., and Russell, P. (1991) p55cdc25 is a nuclear protein required for the initiation of mitosis in human cells. Proc. Natl. Acad. Sci. USA 88, 10500 –10504. 20. Dunphy, W. G., and Kumagai, A. (1991) The cdc25 protein contains an intrinsic phosphatase activity. Cell 67, 189 –196. 21. Hoffmann, I., Clarke, P. R., Marcote, M. J., Karsenti, E., and Draetta, G. (1993) Phosphorylation and activation of human cdc25c by cdc2-cyclin B and its involvement in the self-amplification of MPF at mitosis. EMBO J. 12, 53– 63. 22. Galaktionov, K., Lee, A. K., Eckstein, J., Draetta, G., Meckler, J., Loda, M., and Beach, D. (1995) Cdc25 phosphatases as potential human oncogenes. Science 269, 1575–1577.

23. Galaktionov, K., Chen, X., and Beach, D. (1996) Cdc25 cell-cycle phosphatase as a target of c-myc. Nature 382, 511–517. 24. Dixon, D., Moyana, T., and King, M. J. (1998) Elevated expression of the cdc25A protein phosphatase in colon cancer. Exp. Cell Res. 240, 236 –243. 25. Wu, W. G., Fan, Y. H., Kemp, B. L., Walsh, G., and Mao, L. (1998) Overexpression of cdc25A and cdc25B is frequent in primary non-small cell lung cancer but is not associated with overexpression of c-myc. Cancer Res. 58, 4082– 4085. 26. Jinno, S., Suto, K., Nagata, A., Igarashi, M., Kanaoka, Y., Nojima, H., and Okayama, H. (1994) Cdc25a is a novel phosphatase functioning early in the cell cycle. EMBO J. 13, 1549 – 1556. 27. Hoffmann, I., Draetta, G., and Karsenti, E. (1994) Activation of the phosphatase activity of human cdc25a by a cdk2-cyclin E dependent phosphorylation at the G1/S transition. EMBO J. 13, 4302– 4310. 28. Iavarone, A., and Massague´, J. (1997) Repression of the cdk activator cdc25A and cell-cycle arrest by cytokine TGF-␤ in cells lacking the cdk inhibitor p15. Nature 387, 417– 421. 29. Sebastian, B., Kakizuka, A., and Hunter, T. (1993) Cdc25M2 Activation of cyclin-dependent kinases by dephosphorylation of threonine-14 and tyrosine-15. Proc. Natl. Acad. Sci. USA 90, 3521–3524. 30. Denu, J. M., Zhou, G., Wu, L., Zhao, R., Yuvaniyama, J., Saper, M. A., and Dixon, J. E. (1995) The purification and characterization of a human dual-specific protein tyrosine phosphatase. J. Biol. Chem. 270, 3796 –3803. 31. Tonks, N. K., Diltz, C. D., and Fischer, E. H. (1988) Characterization of the major protein-tyrosine phosphatases of human placenta. J. Biol. Chem. 263, 6731– 6737. 32. Jeffrey, P. D., Russo, A. A., Polyak, K., Gibbs, E., Hurwitz, J., Massague´, J., and Pavletich, N. P. (1995) Mechanism of CDK activation revealed by the structure of a CyclinA-CDK2 complex. Nature 376, 313–320. 33. Eckstein, J. W., Beer-Romero, P., and Berdo, I. (1996) Identification of an essential acidic residue in Cdc25 protein phosphatase and a general three-dimensional model for a core region in protein phosphatases. Protein Sci. 5, 5–12. 34. Zhang, Z.-Y., Maclean, D., Thieme-Sefler, A. M., Roeske, R., and Dixon, J. E. (1993) A continuous spectrophotometric and fluorimetric assay for protein tyrosine phosphatase using phosphotyrosine-containing peptides. Anal. Biochem. 211, 7–15. 35. Zhang, Z.-Y., Walsh, A. B., Wu, L., McNamara, D. J., Dobrusin, E. M., and Miller, W. T. (1996) Determinants of Substrate Recognition in the Protein-tyrosine phosphatase PTP1. J. Biol. Chem. 271, 5386 –5392. 36. Cooper, J. A., Sefton, B. M., and Hunter, T. (1983) Detection and Quantification of Phosphotyrosine in Proteins. Methods Enzymol. 99, 387– 402. 37. Gautier, J., Solomon, M. J., Booher, R. N., Bazan, J. F., and Kirschner, M. W. (1991) Cdc25 is a specific tyrosine phosphatase that directly activates p34cdc2. Cell 67, 197–211. 38. Kumagai, A., and Dunphy, W. G. (1991) The cdc25 protein controls tyrosine dephosphorylation of the cdc2 protein in a cell-free system. Cell 64, 903–914. 39. Strausfeld, U., Labbe´, J. C., Fesquet, D., Cavadore, J. C., Picard, A., Sadhu, K., Russell, P., and Dore´e, M. (1991) Dephosphorylation and activation of a p34cdc2/cyclin B complex in vitro by human cdc25 protein. Nature 351, 242–245. 40. Morris, M. C., and Divita, G. (1999) Characterization of the interactions between human cdc25C, cdks, cyclins and cdk-cyclin complexes. J. Mol. Biol. 286, 475– 487.

SUBSTRATE SPECIFICITY FOR HUMAN Cdc25A 41. Gottlin, E., Epstein, D. M., Eckstein, J., and Dixon, J. (1996) Kinetic analysis of the catalytic domain of human cdc25B. J. Biol. Chem. 272, 27445–27449. 42. Xu, X., and Burke, S. P. (1996) Roles of active site residues and the NH2-terminal domain in the catalysis and substrate binding of human Cdc25. J. Biol. Chem. 271, 5118 –5124. 43. Jessus, C., Ducommun, B., and Beach, D. (1990) Direct activation of cdc2 with phosphatase: identification of p13suc1-sensitive and insensitive steps. FEBS Lett. 266, 4 – 8. 44. Borgne, A., and Meijer, L. (1996) Sequential dephosphorylation of p34cdc2 on Thr-14 and Tyr-15 at the prophase/metaphase transition. J. Biol. Chem. 271, 27847–27854. 45. Kumagai, A., and Dunphy, W. G. (1992) Regulation of the cdc25 protein during the cell cycle in Xenopus extracts. Cell 70, 139 – 151. 46. Kumagai, A., and Dunphy, W. G. (1995) Control of the Cdc2/ Cyclin B Complex in Xenopus Egg Extracts arrested at a G2/M Checkpoint with DNA synthesis inhibitors. Mol. Biol. Cell 6, 199 –213. 47. Russo, A. A., Jeffrey, P. D., and Pavletich, N. P. (1996) Structural basis of cyclin-dependent kinase activation by phosphorylation. Nat. Struct. Biol. 3, 696 –700.

51

48. Garcia-Calvo, M., Peterson, E. P., Rasper, D. M., Vaillancourt, J. P., Zamboni, R., Nicholson, D. W., and Thornberry, N. A. (1999) Purification and catalytic properties of human caspase family members. Cell Death Differ. 6, 362–369. 49. Shaffer, J., and Adams, J. A. (1999) An ATP-linked structural change in protein kinase A precedes phosphoryl transfer under physiological magnesium concentrations. Biochemistry 38, 5572–5581. 50. Wang, S., Tabernero, L., Zhang, M., Harms, E., Van Etten, R. L., and Stauffacher, C. V. (2000) Crystal structures of a low-molecular weight protein tyrosine phosphatase from Saccharomyces cerevisiae and its complex with the substrate p-nitrophenyl phosphate. Biochemistry. 51. Bucciantini, M., Stefani, M., Taddei, N., Chiti, F., Rigacci, S., and Ramponi, G. (1998) Sequence-specific recognition of peptide substrates by the low M-r phosphotyrosine protein phosphatase isoforms. FEBS Lett. 422, 213–217. 52. Hofmann, K., Bucher, P., and Kajava, A. V. (1998) A model of Cdc25 phosphatase catalytic domain and Cdk-interaction surface based on the presence of a rhodanese homology domain. J. Mol. Biol. 282, 195–208.