Article No. jmbi.1998.2475 available online at http://www.idealibrary.com on
J. Mol. Biol. (1999) 286, 475±487
Characterization of the Interactions Between Human cdc25C, cdks, Cyclins and cdk-Cyclin Complexes May C. Morris and Gilles Divita* Centre de Recherches de Biochimie Macromoleculaire UPR-1086 CNRS, 1919 Route de Mende, 34293 Montpellier Cedex 5, France
We have overexpressed and puri®ed human dual-speci®city phosphatase cdc25C from a prokaryotic expression system at high levels and in a soluble, active form, and have studied and quanti®ed its potential to interact with cdks, cyclins and preformed cdk-cyclin complexes by ¯uorescence spectroscopy and size-exclusion chromatography. Our data indicate that human cdc25C forms stable complexes, through hydrophobic contacts, with cdk and cyclin monomers, as well as with preformed cdk-cyclin complexes. In vitro, cdc25C interacts with cyclin monomers with high af®nity, with tenfold less af®nity with cdks, and with intermediate af®nity with cdk-cyclin complexes. Moreover, changes observed in the intrinsic ¯uorescence of cdks, cyclins and cdk-cyclin complexes upon interaction with cdc25C are indicative of concomitant conformational changes within cdks and cyclins. From our results, we propose that in vitro, in the presence of monomeric cdks and cyclins, cdc25C forms stable ternary complexes, ®rst through a high af®nity interaction with a cyclin, which may then help target cdc25C towards a cdk. We discuss the biological relevance of our results and propose that a similar, two-step mechanism of interaction between cdc25C and cdk-cyclin complexes may occur in vivo. # 1999 Academic Press
*Corresponding author
Keywords: cdc25C; cyclin; cdk; interactions; ¯uorescence spectroscopy
Introduction In higher eukaryotes, cell cycle progression is controlled by stoichiometric complexes of cyclindependent kinases (cdks) and of cyclins, which are themselves regulated by a large panel of upstream activators and inhibitors (reviewed by DoreÂe & Galas, 1994; Morgan, 1995). One of these activators, cdc25 was initially isolated in Schizosaccaromyces pombe as a dosage-dependent, rate-limiting inducer of mitotic control (Russell & Nurse, 1986). cdc25 was subsequently shown to be a dual-speci®city serine/threonine-tyrosine phosphatase which promotes entry into mitosis through speci®c dephosphorylation of two residues Thr14 and Tyr15 on the cdk subunit of the cdc2-cyclinB complex (Gautier et al., 1991; Kumagai & Dunphy, 1991; Abbreviations used: cdks, cyclin-dependent kinases; pNPP, para-nitrophenyl phosphate; mant-ATP, N-methyl-anthraniloyl-Adenosine triphosphate. E-mail address of the corresponding author:
[email protected] 0022-2836/99/070475±13 $30.00/0
Millar et al., 1991a; Strausfeld et al., 1991; Lee et al., 1992), thereby counteracting inhibitory phosphorylation by wee1/myt1/mik1 (Lundgren et al., 1991; Parker et al., 1992; Mueller et al., 1995). In humans, cdc25 is encoded by a multigene family of at least three members A, B and C, which can all fully rescue S. Pombe cdc25 mutant strains and display 40-50 % amino acid identity with one another, but which present different structural, functional, and temporal speci®cities. Human p55cdc25C (Sadhu et al., 1990) is functionally homologous to yeast cdc25 (Russell & Nurse, 1986) in its ability to dephosphorylate cdc2 at the G2/M transition (Strausfeld et al., 1991; Millar et al., 1991b), and in its key role in the checkpoint linking DNA synthesis and DNA damage to initiation of mitotic events (Dasso & Newport, 1990; Millar et al., 1991b; Enoch et al., 1991; Weinert, 1997; Furnari et al., 1997; Sanchez et al., 1997; Peng et al., 1997). p75cdc25 A is involved in the G1/S transition and has a potential for oncogenic transformation (Jinno et al., 1994; Hoffmann et al., 1994; Galaktionov et al., 1995a). p60-63edc25B is the most # 1999 Academic Press
476 abundant of the three isoforms and is expressed as several splicing variants (Nagata et al., 1991; Baldin et al., 1997a). Besides its function in the early stages of mitosis (Gabrielli et al., 1996), it has been reported to be associated with oncogenic and viral transformation (Nagata et al., 1991; Galaktionov et al., 1995a). Both cdc25B and C are themselves regulated through reversible phosphorylation in vivo and in vitro by cdc2-cyclin A and cdc2cyclin B, respectively; phosphorylation of cdc25B is a means of targeting it to proteasome-mediated degradation (Baldin et al., 1997b), whereas phosphorylation of cdc25C is required for its complete activation as a cdc2-speci®c dual-speci®city phosphatase in vivo (Izumi et al., 1992; Kumagai & Dunphy, 1992; Hoffmann et al., 1993; Izumi & Maller, 1993; Strausfeld et al., 1994; Lamb et al., 1994). cdc25 protein phosphatases have been shown to interact with a variety of different cell cycle proteins, including 14-3-3, Raf and protein kinase chk1 (Ogg et al., 1994; Conklin et al., 1995; Galaktionov et al., 1995b; Peng et al., 1997), as well as with their natural substrates, the cyclin dependent kinases. cdc25 proteins interact with cyclins and cdk-cyclin complexes in vitro, and can be co-immunoprecipitated with cdk-cyclin complexes from cell extracts (Galaktionov & Beach, 1991; Jessus & Beach, 1992; Zheng & Rudermann, 1993; Xu. & Burke, 1996; Saha et al., 1997). However, neither the mechanism, nor the kinetics of this interaction has been dissected, and no study on the precise nature and af®nity of these interactions has yet been undertaken. Similarly, the means by which cdc25 is regulated and activated to dephosphorylate cdk-cyclin complexes in vivo remains controversial. Activation of cdc25 proteins has been suggested to be controlled through interaction with the so-called ``Pbox'' cyclin domain (Galaktionov & Beach, 1991; Jessus & Beach, 1992; Zheng & Rudermann, 1993). Other authors have shown that phosphorylation of cdc25C on key proline-dependent phosphorylation sites is required for its activation in vivo (Izumi et al., 1992; Hoffmann et al., 1993; Izumi & Maller, 1993; Strausfeld et al., 1994). In vitro, however, nonphosphorylated recombinant human cdc25C can dephosphorylate and activate histone H1 activity of p34cdc2-cyclin B puri®ed from G2-arrested star®sh oocytes (Strausfeld et al., 1991). In order to provide a better understanding of the nature and mechanism of the interactions between cdc25C, cdks and cyclins, we devised a means to overexpress and purify high levels of recombinant human cdc25C in a soluble, active form, and investigated the potential of this protein to interact with monomeric recombinant cdks, cyclins and preformed cdk-cyclin complexes in vitro by ¯uorescence spectroscopy and size-exclusion chromatography. Our results indicate that recombinant human cdc25C can interact and form stable complexes through hydrophobic contacts with isolated cyclins and cdks, as well as with cdk-cyclin complexes, and that these interactions occur conco-
Interactions Between Human cdc25C, Cyclins and cdks
mitantly with conformational changes within cdks and cyclins.
Results Expression and purification of human cdc25C Despite many attempts to express and purify human cdc25C from prokaryotic systems, this phosphatase has proven extremely dif®cult to produce and, if not at low levels, has mostly been expressed in an insoluble, poorly active form. Hence, in order to undertake a biophysical study of the interactions between human cdc25C, cdks and cyclins, we ®rst set upon overexpressing and purifying human cdc25C in suf®cient amounts (i.e. >1 mg per l culture), in a soluble and active form. We succeeded in overexpressing high levels of soluble human cdc25C from the pOTSV-Hucdc25C/ Escherichia coli AR68 vector/strain system by heat shock induction at 42 C for three hours. Human cdc25C was puri®ed from the soluble fraction of bacterial cultures, by precipitation in ammonium sulphate, followed by a series of ion-exchange and size-exclusion chromatographic steps, including a nucleotide-af®nity Blue Sepharose column, which dissociated GroEL co-expressed in this system from cdc25C. The purity of cdc25C was assessed by SDS-PAGE electrophoresis and Western blotting with anti-cdc25C and anti-GroEL antibodies, as reported in Figure l(a). Using this expression system coupled with this puri®cation procedure, we were routinely able to obtain 3-5 mg recombinant, soluble human cdc25C from one litre of bacterial culture. That we should obtain such high amounts of cdc25C in a soluble form was inherent in the bacterial expression system, in which high levels of bacterial GroEL chaperone were co-expressed by heat shock. Although GroEL co-puri®ed along with cdc25C in the ®rst stages of puri®cation, it was predominantly eliminated during nucleotideaf®nity chromatography and completely eliminated in the ®nal ion-exchange and gel ®ltration chromatographies (Figure l(a)). The homogeneity and molecular mass of puri®ed recombinant cdc25C were con®rmed by HPLC size-exclusion chromatography in 150 mM potassium phosphate buffer. cdc25C eluted in a single peak, at 13 minutes, as a monomeric protein of 55 kDa (see Figure 6(b)). The phosphatase activity of puri®ed recombinant human cdc25C was measured in a standard para-nitrophenyl phosphate (pNPP) dephosphorylation assay (Figure 1(b)). From the values measured, we were able to determine the speci®c activity of recombinant cdc25C puri®ed from the pOTSV-/AR68 heat shock induction system as 61.3(11.5) nmol pNPP dephosphorylated per minute, per mg human cdc25C, a value slightly greater than that initially described for recombinant cdc25 puri®ed from inclusion bodies (Dunphy & Kumagai, 1991).
477
Interactions Between Human cdc25C, Cyclins and cdks
Interactions between human cdc25C, cdks, cyclins and cdk-cyclin complexes monitored by fluorescence spectroscopy In order to study the potential of unphosphorylated recombinant human cdc25C to interact with cdks and cyclins in vitro, we devised a strategy based on the use of tryptophan ¯uorescence spectroscopy. Both cdks and cyclins contain a wellde®ned number of conserved tryptophan groups, which confer on them an important intrinsic ¯uorescence, and which can be used as probes to monitor conformational changes and/or binding of other proteins (Burstein et al., 1973; Divita et al., 1993; Heitz et al., 1997). In contrast, human cdc25C presents a complete lack of tryptophan groups and consequently exhibits negligable intrinsic ¯uorescence compared to cdks and cyclins, between 310 and 400 nm, when excited at 290 nm. We took advantage of the absence of tryptophan groups in human cdc25C, and of the presence of conserved tryptophan groups in cdks and cyclins, to monitor interactions between cdc25C, cdks and cyclins by ¯uorescence spectroscopy. The Kd values obtained for the different interactions monitored are reported in Table 1. Human cdc25C interacts with cyclins in vitro Prior to investigating the potential of cdc25C to interact with cyclins by ¯uorescence spectroscopy, we characterized their intrinsic ¯uorescence spectra. In 150 mM potassium phosphate buffer (pH 6.8), both human cyclin A and B1 exhibited characteristic intrinsic tryptophan ¯uorescence with a maximal emission centered at 342 and 348 nm, respectively. Figure 2 presents the titration curves obtained for the interaction of cdc25C with cyclin A and Table 1. Dissociation constants of the interactions between human cdc25C and cdks, cyclins or preformed cdk-cyclin complexes Figure 1. Puri®cation of human cdc25C. (a) Puri®cation of recombinant human cdc25C-elimination of GroEL. Recombinant human cdc25C expressed from induced AR68 cultures co-puri®ed with GroEL in the initial steps of puri®cation, but was no longer associated with cdc25C at the end of the puri®cation. Samples from the different steps of the puri®cation procedure were analysed by SDS-PAGE, and by Western blotting with an af®nity-puri®ed antibody speci®c to human cdc25C and a polyclonal antibody recognizing GroEL. Lane 1, sample enriched by differential ammonium sulphate precipitation, and passed through an SP Sepharose column; lane 2, puri®ed cdc25C. The protein band corresponding to cdc25C is indicated with an arrow. (b) Phosphatase activity of puri®ed human cdc25C. Phosphatase activity of puri®ed human cdc25C was assayed against pNPP by measuring the increase of absorbance at 410 nm as a function of time. Calculation of the corresponding speci®c activity of cdc25C against pNPP yields a value of. 61 nmol pNPP dephosphorylated per minute, per mg of recombinant cdc25C.
Interactions
Kd (nM)
Cyclin A/cdc25C cyclin B1/cdc25C cdc2/cdc25C cdk2/cdc25C cdk7/cdc25C cdk2-cyclin A/cdc25C cdc2-cyclin B1/cdc25C cdk2/cdc25C-cyclin A cdc2/cdc25C-cyclin B1
72(4.5) 85(9) 698(70) 801(129) 460(80) 308(12) 339(21) 160(22) 152(18)
Dissociation constants of the interactions between human cdc25C and cdks, cyclins or preformed cdk-cyclin complexes were calculated from best curve ®tting of the values corresponding to relative changes in their intrinsic tryptophan ¯uorescence upon addition of increasing concentrations of human cdc25C, as described in the legends to Figures 2, 3 and 4. Dissociation constants of the interactions between cdks and cdc25C-cyclin were calculated from the changes in extrinsic ¯uorescence of mant-ATP pre-associated with cdks upon binding of cdc25C-cyclin complexes, as described in the legend to Figure 5.
478
Interactions Between Human cdc25C, Cyclins and cdks
Figure 2. Interactions of human cdc25C with cyclins. Binding of human cdc25C to cyclins A and B1 was monitored by measuring changes in the intrinsic tryptophan ¯uorescence spectrum of cyclins (100 nM) upon addition of increasing concentrations of cdc25C protein, between 310 and 400 nm following excitation at 290 nm. The maxima of the emission spectra were plotted against the concentration of human cdc25C, and values were ®tted according to a standard quadratic equation using the Gra®t software. Typical values obtained for cyclins A (*) and B1 (*) are presented together with their best ®tting curve.
cyclin B1. Upon addition of cdc25C, the intrinsic ¯uorescence of cyclins B1 and A increased, by a factor of 2.1 and 2.2, respectively, at saturating concentrations of cdc25C, without any obvious changes in their maximal emission wavelength. This increase in the ¯uorescence of cyclins A and B1 reveals that cdc25C effectively interacts with these cyclins in vitro, and that these interactions induce conformational changes which increase the hydrophobicity in the environment of the tryptophan groups. In both cases, titration curves are monophasic, and reveal that cdc25C presents very high af®nity for both cyclins with dissociation constant values of 72(4.5) nM for cyclin A and 85(9) nM for cyclin B1. Human cdc25C interacts with cdks in vitro We next monitored the potential of recombinant human cdc25C protein to interact with human cdc2, cdk2 and cdk7 in vitro by intrinsic ¯uorescence spectroscopy. The corresponding titration curves are reported in Figure 3(a) for cdc2 and cdk2, and in Figure 3(b) for cdk7. As for cyclins, the addition of cdc25C increased the intrinsic ¯uorescence of cdc2 and cdk2, by a factor of 1.9 and 2, respectively, at saturating concentrations of cdc25C. In contrast, binding of cdc25C to cdk7 lead to quenching of its ¯uorescence by 20 %. As in the case of cyclins, these changes in ¯uorescence are indicative of an interaction between cdc25C and cdc2, cdk2 and cdk7 in vitro, and of a concomitant conformational
Figure 3. Interactions of human cdc25C with cdks. Binding of human cdc25C to cdc2, cdk2 and cdk7 was monitored by measuring changes in the intrinsic tryptophan ¯uorescence spectrum of cdks (100 nM) upon addition of increasing concentrations of cdc25C protein, between 310 and 400 nm following excitation at 290 nm. The maxima of the emission spectra were plotted against the concentration of human cdc25C, and values were ®tted according to a standard quadratic equation using the Gra®t software. (a) Typical values obtained for cdk2 (*) and cdc2 (*), together with their best ®tting curve. (b) Typical values for cdk7 (~), together with their best ®tting curve.
change which modi®es the environment of the tryptophan groups involved. In the case of cdc2 and cdk2, the increase in intrinsic ¯uorescence implies that the tryptophan groups involved are surrounded by a more hydrophobic environment. In the case of cdk7, in contrast, quenching of intrinsic ¯uorescence suggests either that tryptophan groups are more exposed to the solvent, or that cdc25C interacts directly with a tryptophan group. In all three cases, titration curves are monophasic, and best ®ts yield Kd values of 698(70) nM for cdc2, 801(120) nM for cdk2, and of 460(80) nM for cdk7. Taken together, these results indicate that the interaction of cdc25C with cdk7 is of a different
Interactions Between Human cdc25C, Cyclins and cdks
479
nature compared to that with cdc2 and cdk2. This may be due to differences in the three-dimensional conformation of cdk7, or simply to punctual differences in the amino acid sequence of cdk7 compared to those of cdc2 and cdk2. Human cdc25C interacts with preformed cdk-cyclin complexes in vitro We examined whether cdc25C could interact with preformed cdk2-cyclinA and cdc2-cyclin B1 complexes, and whether the conformational changes induced in isolated cdks and cyclins upon their interaction with cdc25C still occurred when these were complexed to their cdk or cyclin partner. The titration curves obtained for these interactions are presented in Figure 4. cdc2 and cyclin B1 on the one hand, and cdk2 and cyclin A on the other, were mixed stoichiometrically in potassium phosphate buffer (pH 6.8) at a ®nal concentration of 100 nM and incubated for 30 minutes to form stable complexes, prior to addition of cdc25C protein. We veri®ed, by HPLC chromatography, that in these conditions, cdks and cyclins were fully complexed. As observed for isolated cyclins A and B1, cdc2 and cdk2, the addition of cdc25C increased the intrinsic ¯uorescence of cdc2-cyclin B1 and cdk2-cyclin A complexes. These changes in relative ¯uorescence were, however, less marked than those observed for isolated cyclins A and B1, cdc2 and cdk2, with only a 1.4-fold relative increase. As in Figures 2 and 3, titration curves are monophasic and yield Kd
Figure 5. Binding of cdks to preformed cdc25C/cyclin complexes. cdc25C/cyclin A and cdc25C/cyclin B1 complexes were formed by incubating stoichiometric concentrations of cdc25C and cyclins at 25 C for 30 minutes. cdk2 or cdc2 (0.1 mM) saturated with mant-ATP (1 mM) were titrated by increasing concentration of cdc25C/ cyclin A (*) or cdc25C/cyclin B1 (*), respectively. The enhancement of mant-ATP ¯uorescence was monitored at 450 nm upon excitation at 340 nm and data were plotted and ®tted as described in the legend to Figure 4.
values of 308(12) nM for cdk2-cyclin A, and 339(21) nM for cdc2-cyclin B1, which are intermediate between the Kd values obtained for cdks and those for cyclins. Binding of cdks to preformed cdc25C-cyclin complexes in vitro
Figure 4. Interactions of human cdc25C with cdkcyclin complexes. cdk2-cyclin A and cdc2-cyclin B1 complexes were formed by incubating stoichiometric concentrations (100 nM) of the corresponding cdk with a cyclin at 25 C for 30 minutes. Binding of human cdc25C to these preformed cdk-cyclin complexes was then monitored by measuring changes in the intrinsic tryptophan ¯uorescence of these complexes at 340 nm following excitation at 290 nm, upon addition of increasing concentrations of cdc25C protein. Data were plotted and ®tted according to a standard quadratic equation using the Gra®t software. Typical values obtained for cdk2cyclin A (*) and cdc2-cyclin B (*) are shown together with their best ®tting curve.
Given that cdc25C interacts with cdk-cyclin complexes with higher af®nity than with monomeric cdks, we asked whether the interaction between cdc25C and cdks might be regulated by the cyclin subunit. As both cdks and cyclins contain several tryptophan residues, interactions between cdks and preformed cdc25C-cyclin complexes could not be monitored easily by changes in intrinsic ¯uorescence. We have previously reported that the ¯uorescently labelled nucleotide analogue mantATP presents high af®nity for cdks, and constitutes a very sensitive probe for monitoring interactions between cdk and cyclin partners (Heitz et al., 1997; Morris et al., 1998). We therefore monitored binding of cdks to preformed cdc25C-cyclin complexes with cdks complexed to a mant-ATP group, by following changes in the ¯uorescence of mant-ATP. mant-ATP-cdk and cdc25C-cyclin complexes were preformed prior to binding experiments. MantATP-cdk complexes were then titrated with increasing concentrations of cdc25C-cyclin complexes from 50 nM to 1 mM. We con®rmed that at these concentration cdc25C and cyclins were fully complexed by gel ®ltration chromatography. As reported in Figure 5, binding of cdc25Ccyclin A to cdk2, and cdc25C-cyclin B1 to cdc2,
480
Interactions Between Human cdc25C, Cyclins and cdks
Figure 6. Interactions of human cdc25C with cdks, cyclins and cdk-cyclin complexes monitored by size-exclusion chromatography. Recombinant cdc25C was incubated with an equimolar concentration of cyclin A or cdk2-cyclin A in potassium phosphate buffer. Monomeric cdc25C, monomeric cyclin A, and the complexes formed by incubation of cdc25C with cyclin A, cdk2 or cdk2-cyclin A were separated by HPLC size exclusion chromatography on TSK Bio-Sil columns. (a) Chromatographs corresponding to the elution of monomeric human cdc25C (1), of monomeric cyclin A (2), of the binary cdc25C-cyclinA complex (4) and of the ternary cdc25C-cdk2-cyclin A complex (6) from HPLC sizeexclusion TSK Bio-Sil columns in 50 mM potassium phosphate buffer (pH 7.2), 150 mM NaCl. (b) Chromatographs corresponding to the elution of monomeric human cdc25C (1), of monomeric cdk2 (3) and of the binary cdc25C-cdk2 complex (5) from HPLC size-exclusion TSK Bio-Sil columns in 150 mM potassium phosphate buffer (pH 6.8). (c) Analysis of the components of the peaks shown above by electrophoresis on SDS-12.5 % polyacrylamide gels, followed by Western blotting with the corresponding antibodies. Lane 1, monomeric cdc25; lane 2, monomeric cyclin A; lane 3, monomeric cdk2; lane 4, cdc25C and cyclin A from the binary complex in peak 4 in (a); lane 5, cdc25C and cdk2 from the binary complex in peak 5 in (b); lane 6, cdc25C, cyclin A and cdk2 from the ternary complex in peak 6 in (a).
induced a twofold increase in the ¯uorescence of mant-ATP. The titration curves of these interactions are monophasic and yield Kd values of 160(22) nM for cdc25C-cyclin A and 152(18) nM for cdc25C/cyclin B1. Comparison of these results with those obtained for the interactions between cdks and monomeric cdc25C reveals that cdks present higher af®nity for preformed cdc25C-cyclin complexes than for monomeric cdc25C. This indicates that the presence of cyclin regulates and stabilizes association between cdc25C and cdks. Unlike cdc25C-cyclin complexes, cdc25C-cdk complexes were not stable enough at low concen-
trations (100 to 200 nM) to performed titrations of cdc25C-cdk complexes with cyclins. These data con®rm the idea that cyclins are essential for stabilizing the interaction between cdc25C and cdks. Interactions between human cdc25C, cdks, cyclins and cdk-cyclin complexes monitored by size-exclusion chromatography In order to validate the interactions measured between human cdc25C, cdks, cyclins and cdkcyclin complexes by ¯uorescence spectroscopy, we monitored the stability of these different complexes
481
Interactions Between Human cdc25C, Cyclins and cdks
by HPLC size-exclusion chromatography. cdc25C was incubated with equimolar amounts of cyclins, cdks, or cdk/cyclin complexes for 30 minutes at 25 C in potassium phosphate buffer, and then loaded onto HPLC Bio-Sil columns for separation by gel ®ltration chromatography. In each case, the volume at which peak samples eluted from the column was compared with that at which monomeric cdc25C, cdks or cyclins normally eluted, and peak fractions were analyzed by SDS-PAGE, followed by staining with Coomassie Blue or by Western blotting with adequate antibodies. As observed indirectly by ¯uorescence spectroscopy, cdc25C was found to form complexes with monomeric cyclins and cdks, as well as with preformed cdk/cyclin complexes. These complexes remained stable throughout the chromatographic run, despite the presence of high salt concentrations in the buffers, indicating that the interactions involved mainly hydrophobic contacts. Figure 6 presents the chromatographs of monomeric cdc25C, cdk2 and cyclin A, and of binary or ternary complexes from the HPLC size-exclusion chromatographic runs, together with an antigenic identi®cation of the proteins eluting within the corresponding peaks. In 50 mM potassium phosphate buffer, 150 mM NaCl (pH 7.2; Figure 6(a)), monomeric cdc25C and cyclin A eluted at 11 and 12 minutes, respectively. After incubation of equimolar concentrations of cdc25C with cyclin A, only one peak could be detected in the corresponding chromatograph, at 9.5 minutes, corresponding to the stoichiometric, quantitative formation of a stable cdc25C-cyclin A complex. Similarly, incubation of cdc25C with an equimolar ratio of cdk2cyclin A yielded a single peak which eluted at 8.5 minutes, indicating that cdc25C, cdk2 and cyclin A had stoichiometrically, quantitatively associated in a stable ternary complex, and had not dissociated, Table 2. Interaction of human cdc25C with cdks, cyclin and cdk-cyclin complexes monitored by size-exclusion chromatography Samples a
cdc25C cdk2b cdc2b Cyclin Aa Cyclin B1a cdk2/cdc25Cb cdc2/cdc25Cb Cyclin A/cdc25Ca Cyclin B1/cdc25Ca Cyclin A/cdk2/cdc25Ca Cyclin B1/cdc2/cdc25Ca
Retention time (min) 11(13)b 14 14 12 11 9 9.5 9.5 8 8.5 7
a Monomers and complexes were separated by HPLC sizeexclusion chromatography on TSK Bio-Sil columns. Monomeric cdc25C, cyclin A, cyclin B1, and the complexes formed by incubation of cdc25C with cyclins and cdk-cyclin complexes were eluted in 50 mM potassium phosphate buffer (pH 7.2), 150 mM NaCl as reported in Figure 6(a). b Monomeric cdk2, cdc2 and cdc25C/cdk complexes were eluted in 150 mM potassium phosphate buffer (pH 6.8) as reported in Figure 6(b).
even in a small proportion, into two binary complexes. In 150 mM potassium phosphate buffer (pH 6.8), monomeric cdc25C and cdk2 eluted at 13 and 14 minutes, respectively (Figure 6(b)). As for cdc25C and cyclin A, incubation of equimolar concentrations of cdc25C with cdk2, yielded a single chromatographic peak at nine minutes, corresponding to the stoichiometric, quantitative formation of a stable cdc25C-cdk2 complex. Similar results were obtained with cdc2 and cyclin B1, and the elution times for the different complexes are reported in Table 2. These results con®rm that the changes observed in the intrinsic tryptophan ¯uorescence of cdks and cyclins can be directly correlated with the binding of cdc25C to cyclins, cdks and cdk-cyclin complexes.
Discussion Expression and purification of recombinant human cdc25C We have developed a means to overexpress recombinant human cdc25C at high levels, and to purify it as a soluble, active protein, using the pOTSV/AR68 heat shock induction system, which naturally expresses high levels of the bacterial chaperone GroEL upon heat shock. That high levels of recombinant cdc25C should be maintained in a soluble form in this system is most likely due to the presence of co-expressed GroEL, which has been used in other studies to enhance the solubility of recombinant proteins (Amrein et al., 1995). However, GroEL was also found to co-purify with cdc25C throughout the ®rst steps of puri®cation, which lead us to optimize our puri®cation protocol, by including a nucleotide-af®nity chromatography step in which GroEL was completely dissociated from cdc25C. Thanks to this procedure, we were routinely able to obtain 3-5 mg of pure, soluble, active human cdc25C protein per litre of bacterial culture. Recombinant cdc25C was used to investigate in vitro molecular interactions between human cdc25C and its cdk and cyclin partners by ¯uorescence spectroscopy and size-exclusion chromatography. Our experiments reveal that, in vitro, human cdc25C forms complexes mainly stabilized by hydrophobic interactions with cdk-cyclin complexes as well as with monomeric cyclins and cdks. cdc25C interacts with cdks in the absence of cyclins in vitro As in vivo, cdc2 is targeted for inhibitory phosphorylation on Thr14 and Tyr15, through association with a mitotic cyclin (Gould & Nurse, 1989; Solomon et al., 1990; Meijer et al., 1991), it has generally been accepted that the main substrate of cdc25 is phosphorylated cdc2-cyclin B. However, in vitro, free cdc2, phosphorylated by recombinant v-src on tyrosine, can also undergo dephosphorylation upon addition by cdc25C (Gautier et al., 1991). Moreover, in vivo, cyclin-free cdc2, either immuno-
482 precipitated with anti-PSTAIRE antibodies (Devault et al., 1992), or puri®ed by gel ®ltration chromatography from HeLa cell extracts (Draetta & Beach, 1988), can be found phosphorylated on inhibitory residues before, and dephosphorylated after, the G2/M transition (Draetta & Beach, 1988; Gautier et al., 1989; Labbe et al., 1989). Hence, in vivo, cdc25C may effectively bind cdc2 even in the absence of a cyclin subunit, as we have observed in vitro. In agreement with our results, cdc25C has been reported to dephosphorylate the mitotic kinase cdc2, as well as the cdk2 protein kinase in vitro (Gautier et al., 1991; Strausfeld et al., 1991; Jessus & Beach, 1992; Gabrielli et al., 1992). In contrast, however, cdk7 is not subject to control via phosphorylation/dephosphorylation of residues Thrl4 and Tyr15 in vivo, and has not been documented to interact with cdc25C in any physiological context. Hence our in vitro data may re¯ect the potential of cdc25C to interact with different cdks in vitro, in the absence of any other factors or partners restricting speci®city. Such a binding potential would be dependent on conserved structural features involved in interactions, and located close to tryptophan groups (Heitz et al., 1997). In the structure of cdk2, tryptophan residues 168, 187 and 229 are situated in the T-loop, in a domain involved in interactions with cyclin A, and in the C-terminal L14-loop, where CKS proteins interact (De Bondt et al., 1993; Jeffrey et al., 1995; Bourne et al., 1996). In addition to these three tryptophan residues conserved in all cdks, cdc2 and cdk2 contain a fourth tryptophan group, Trp245 located close to Trp229 within the L14-loop; cdk7 contains an additional tryptophan, Trp122, located between a-helix 3 and b-sheet 6, close to the ATP-binding site and to residues Thr14 and Tyr15 of cdk2 (De Bondt et al., 1993; Jeffrey et al., 1995). Differences in the number and positions of tryptophan groups, as well as differences in the sequence and structure of cdk7 may explain the different effect we have observed upon interaction of cdk7 with cdc25C, compared to cdc2 and cdk2. Although highly speculative, the interaction between cdc25C and cdk7 might re¯ect an in vivo function of cdc25C, consisting of the targeting of cdk7 towards other cdks for their phosphorylation on Thr161. We are currently investigating this issue with tryptophan mutants of cdk7. cdc25C interacts with cyclins A and B in the absence of cdks in vitro We have shown that cdc25C can interact with recombinant cyclins non-associated to cdks with high af®nity in vitro. To date, the occurrence neither of cyclins as free subunits, nor of complexes of cyclins associated with cdc25C in the absence of cdks, have been reported in vivo. Moreover, within cells, the association of cyclins with other partners within more complex macromolecular structures is likely to limit the accessibility of cyclins to cdc25C.
Interactions Between Human cdc25C, Cyclins and cdks
Hence both the in vivo af®nity and speci®city of cdc25C/cyclin interactions may differ from our in vitro results. That cdc25C should interact with cyclin B1 could be expected, due to its expression pro®le and its function as a component of the mitotic cdc2 kinase. In contrast, although cdc2-cyclin A has been proposed to phosphorylate cdc25C (Lamb et al., 1994), its in vitro interaction with cdc25C was surprising, as cyclin A cannot be immunoprecipitated with cdc25C from cell extracts (Saha et al., 1997). Hence, as for cdks, our results may re¯ect the overall potential of cdc25C to interact with cyclins, irrespective of any cellular context, in which spatio-temporal expression pro®les, in particular, determine the speci®city of interactions. Such interactions most likely occur through conserved structural domains, containing conserved tryptophan groups. Cyclins A and B1 contain two conserved tryptophan residues, located at altogether very different areas of the cyclin molecule in the X-ray structure of cdk2-cyclin A: Trp217, located within the Pbox domain, within helix a1, along the same axis as the ATP-binding site, but on the ``opposite'' side of the cyclin molecule; and Trp372, located in the C terminus of the second cyclin box motif, in a loop between helices a30 and a40 at the ``bottom'' of the cyclin molecule (De Bondt et al., 1993; Jeffrey et al., 1995; Brown et al., 1995). Although we cannot exclude that cdc25C may interact with the C-terminal domain of cyclins, the Pbox domain seems to be a more obvious candidate for several reasons. First, the Pbox has been reported to be involved in interactions with cdc25 phosphatases (Galaktionov & Beach 1991; Zheng & Ruderman, 1993). Second, the Pbox tryptophan is not only conserved in cyclins A and B, but also in cyclins D1, D2, D3, E and F, and is therefore likely to be involved in hydrophobic contacts with other proteins, as recently shown with p27KIP1 (Brown et al., 1995; Russo et al., 1996). In this context, it should be noted that a cyclin-binding motif similar to the ``Cy motif'' of the p2l cdk inhibitor has been identi®ed in the N terminus of human cdc25A, as essential for in vitro and in vivo binding and dephosphorylation of cdk2-cyclins A and E by cdc25A, and has been suggested to interact with the Pbox of cyclins (Saha et al., 1997). Analysis of the primary sequence of human cdc25C does not reveal the presence of a similar motif, which suggests that this motif may be speci®c for binding of G1 and S phase-speci®c cyclins in vivo (Morgan, 1995; Chen, J., et al., 1995, 1996; Chen, I. T., et al., 1996; Luo et al., 1995; Zhu et al., 1995; Li et al., 1993; Hannon et al., 1993; Krek et al., 1995), but cdc25C may possess an analogous, N-terminal cyclin-binding motif, which would restrict its ability to interact with proteins containing G2 determinants in vivo.
483
Interactions Between Human cdc25C, Cyclins and cdks
cdc25C interacts with cdk/cyclin complexes in vitro: cyclins target cdc25C to cdks We have shown that cdc25C can form stable ternary complexes with preformed cdc2-cyclin B and cdk2-cyclinA, with an af®nity twofold greater than that observed for monomeric cdks. In addition, the interaction of cdc25C with these complexes induces only a 1.4-fold increase in their intrinsic ¯uorescence, compared to the twofold increase observed for monomeric cyclins and cdks. Hence the interaction of cdc25C with preformed cdk/cyclin complexes induces less marked conformational changes than those observed for monomeric cdks or cyclins and these do not correspond to the sum of the latter. By measuring the changes in the extrinsic ¯uorescence of mant-ATP pre-bound to cdks, to probe interactions between cdks and cdc25C-cyclin complexes, we have also demonstrated that the binding of a cyclin to cdc25C improves the af®nity of the latter for cdks ®vefold. These results, together with the low stability of the cdk-cdc25C complexes, suggest that the presence of a cyclin is required for stabilizing cdk-cdc25C interactions. Moreover, the presence of cdc25C may modulate the speci®city of the cdk-cyclin complex formation. Based on these results, we propose an ``in vitro`` model of a two-step mechanism of cdc25C binding with cdks and cyclins. In this model, cdc25C would ®rst interact with monomeric cyclin subunits with high af®nity, thus forming a stable binary complex, and then only bind the cdk partner of the cyclin. This model is in perfect agreement with the results of Galaktionov & Beach (1991), in which cdc25 was shown to be ``activated'' in its ability to dephosphorylate cdc2, upon incubation with cyclins, in the absence of cdks. This model is, however, less likely to apply in vivo, in particular as cyclins have not been found associated with cdc25C in the absence of cdks. In vivo cdc25C may instead associate with preformed cdk/cyclin complexes in a two-step process, mediated ®rst by a high af®nity interaction with the cyclin, which would then increase its af®nity for cdk binding. Unphosphorylated cdc25C interacts with cdks, cyclins and cdk-cyclin complexes We have shown that, in vitro, recombinant human cdc25C can interact with high af®nity and form stable complexes with cdk and cyclin partners, as well as with cdk-cyclin complexes, in the absence of phosphorylation of cdc25C. Similarly, Strausfeld et al. (1991) demonstrated that unphosphorylated, bacterially expressed puri®ed cdc25 could dephosphorylate and activate p34cdc2-cyclin B puri®ed from G2-arrested star®sh oocytes. However, in vivo, activation of cdc25C at the G2/M transition seems to depend on phosphorylation of its N-terminal domain (Strausfeld et al., 1994; Lamb et al., 1994, Hoffmann et al., 1993; Izumi &
Maller, 1993, 1995; Izumi et al., 1992). Hence, phosphorylation of the N-terminal domain of cdc25C may indeed be essential for interactions between cdc25C and its substrate in vivo, but may only contribute to specifying and stabilizing interactions shown to occur with its active site region (Xu & Burke, 1996). Alternatively, in vivo, phosphorylation of the N-terminal region of cdc25C may be involved in interactions with non-substrate proteins, thus targeting cdc25C to a speci®c locale or to speci®c protein complexes, where it would execute its function. Similarly, we cannot exclude that phosphorylation of cdks may modulate the speci®city and af®nity of these interactions in vitro. We are currently investigating these issues using phosphorylated cdks and phosphorylation site-mutants of human cdc25C. In conclusion, our results clearly demonstrate that human cdc25C interacts and forms stable complexes with cyclins, cdks and cdk-cyclin complexes in vitro. Although our data most certainly re¯ect the potential of cdc25C to interact with different cdks and cyclins, it should be kept in mind that a more restrictive speci®city is likely to exist in vivo, in part de®ned by the spatio-temporal expression pro®les of cdks and cyclins, but also by other controls which regulate the availability and accessibility of cdks and cyclins, including cell-cycledependent phosphorylation. From the results presented in this study, we propose that cdc25C interacts with cdks and cyclins in a two-step process, ®rst through high af®nity association with a cyclin subunit, which then co-operatively increases the af®nity of cdc25C for a cdk, and as such targets cdc25C towards a cdk substrate.
Materials and Methods Materials Restriction enzymes, Vent polymerase and phage T4 DNA ligase were obtained from Biolabs. PCR oligonucleotides were from Eurogentec. The pRK171-cdc25C construct was a gift from Dr P. Russell. The pOTSV vector was purchased from SmithKline & French Laboratories. Chelating Sepharose, Fast Flow SP, Q Sepharose and Blue Sepharose chromatography resins, as well as GSTaf®nity and Ni2-nitrolotriacetic acid-agarose resins, and HiLoadTM16/60 Superdex 75 and MonoQ HR5/5 FPLC columns were obtained from Pharmacia. TSK-250 and TSK-125 Bio-Sil columns were purchased from Biorad. The BCA Protein Assay Reagent kit was purchased from Pierce and para-nitrophenyl phosphate tablets (Sigma 104) were purchased from Sigma. Polyclonal rabbit antibodies against human cyclin A (#sc-751) and human cdk2 (#sc-163) were purchased from Tebu, Santa Cruz. An af®nity-puri®ed polyclonal antibody against human cdc25C was kindly provided by Dr P. Russell. Polyclonal rabbit antibodies against GroEL were purchased from Sigma. Horseradish peroxidaseconjugated anti-rabbit donkey secondary antibodies for Western blotting were purchased from Amersham.
484 Cloning human cdc25C into the pOTSV expression vector The open reading frame coding for human cdc25C was ampli®ed from the pRK171-cdc25C using two primers 50 CGA TGG ATC CGT CTA CGG AAC 30 and 50 CTG CTC TAG ATC ATG GGC TCA TGT 30 and 40 cycles of denaturation at 94 C for one minute, annealing at 56 C for two minutes, and ampli®cation at 72 C for two minutes. This product was then digested with BamHI and XbaI, and cloned into the pOTSV vector. Expression and purification of human cdc25C The pOTSV-cdc25C construct was transfected into the E. coli AR68 strain, and positive clones were grown into cultures overnight at 32 C in Terri®c Broth (12 g/l bactotryptone, 24 g/l bactoyeast, 4 ml/l glycerol, 2.31 g/l KH2PO4, 12.5 g/l K2HPO4) supplemented with 50 mg/ml ampicillin. Cultures were induced to overexpress human cdc25C by heat shock induction at 42 C for three hours. For puri®cation of cdc25C, bacteria were pelleted by centrifugation for 15 minutes at 4500 g/4 C, resuspended in K buffer (15 mM KH2PO4, l5 mM K2HPO4 (pH 6.8), 1 mM EDTA) supplemented with 1 mM PMSF, treated with 100 mg/ml lysozyme for 20 minutes at 25 C, and frozen. The pellet was thawed and cells were lysed by sonication. The soluble fraction was separated from the insoluble by centrifugation at 25,100 g at 4 C, precipitated in 25 %, then 45 % NH4SO4, and resuspended in 3 ml K buffer. This suspension was desalted by gel ®ltration chromatography on chelating Sepharose resin equilibrated in K buffer, and further puri®ed by passage through an SP Sepharose column and a Blue Sepharose column, both equilibrated in K buffer (pH 6.8). The resulting ¯owthrough was set to pH 8.8 with 2 M Tris (pH 13), and loaded onto a Q Sepharose column equilibrated with 30 mM Tris-HCl (pH 8.8), 1 mM EDTA. Fractions containing cdc25C were eluted between 400 M and 1 M salt, pooled and precipitated in 70 % NH4SO4, then loaded onto a HiloadTM 16/60 Superdex 75 Pharmacia FPLC gel ®ltration column, and separated in 30 mM Tris-HCl (pH 8.8), 1 mM EDTA, l00 mM NaCl at a ¯ow rate of 0.8 ml per minute. Fractions containing cdc25C were pooled, diluted 1:1 in 30 mM TrisHCl (pH 8.8), 1 mM EDTA, loaded onto a Mono Q HR 5/5 Pharmacia FPLC anion-exchange column and eluted with a salt gradient from 200 mM to 450 mM NaCl. cdc25C typically eluted as a single protein in a peak at 380-400 mM NaCl. Protein fractions were routinely analyzed by SDS-PAGE, according to Laemmli (1970) on SDS-12.5 % polyacrylamide gels, or 10 % when necessary, for identi®cation of cdc25C and closely migrating GroEL. Puri®ed samples were concentrated using a Centricon concentrator (Schleicher & Schuell) and their concentration was determined using the BCA Protein Assay Reagent kit, and bovine serum albumin as a standard (Smith et al., 1985). Homogeneity of recombinant cdc25C puri®ed according to this protocol was con®rmed by HPLC size-exclusion chromatography in 150 mM potassium phosphate buffer, at a ¯ow rate of 0.8 ml per minute using two TSK Bio-Sil columns, TSK-250 and TSK-125 in series, as described (Heitz et al., 1997). The molecular mass of recombinant cdc25C was determined by comparison with the elution pro®le of the following protein standards in the same conditions: 1.35 kDa vitamin B12 (25.1 minutes), 17 kDa myoglobin (17.2 minutes), 44 kDa ovalbumin (14.1 minutes), p5l subunit of HIV-1 reverse transcriptase (13.5 minutes), p66 subunit of HIV-1
Interactions Between Human cdc25C, Cyclins and cdks reverse transcriptase (12.5 minutes), 158 kDa IgG (6.88 minutes), 670 kDa thyroglobulin (5.9 minutes). Determination of phosphatase activity: pNPP assays The phosphatase activity of puri®ed, recombinant human cdc25C was measured in standard pNPP assays, essentially as described by Dunphy & Kumagai (1991), in a 100 ml reaction containing 50 mM Tris-HCl (pH 8.0), 250 mM NaCl, 5 mM EDTA, l0 mM DTT, l00 mM pNPP (Sigma 104) for 0 to 40 minutes at 37 C. The reaction was terminated by the addition of 2 ml of 10 M NaOH, and phosphatase activity was determined as a function of the increase of absorbance at 410 nm. For calculation of speci®c phosphatase activity against pNPP, a molar absorptivity value of pNPP of 1.78 104 Mÿ1 cmÿ1 was used (Tonks et al., 1988; Dunphy & Kumagai, 1991). Expression and purification of cdks and cyclins cdks and cyclins were overexpressed in E. coli and puri®ed essentially as described (Heitz et al., 1997). Human cdk2 and cdk7 were puri®ed as GST-fusion proteins by af®nity chromatography, followed by removal of the GST tag with factor X and further puri®ed by gel ®ltration chromatography (Heitz et al., 1997). Human cyclin A was puri®ed as described by Lorca et al. (1992). Human cyclin B1 was prepared as a maltose-binding fusion protein, followed by removal of the maltose-binding protein with factor X. N-terminal polyhistidinetagged human cdc2 was puri®ed by af®nity chromatography on Ni2-nitrolotriacetic acid-agarose as described by Heitz et al. (1997). Protein concentrations were routinely determined using the BCA Protein Assay Reagent kit, using bovine serum albumin as a standard (Smith et al., 1985). Fluorescence experiments Fluorescence measurements were performed essentially as described (Heitz et al., 1997) at 25 C, using a Spex II Jobin Yvon ¯uorolog spectro¯uorometer, with spectral band-passes of 2 nm for both excitation and emission, in l50 mM potassium phosphate buffer (pH 6.8), containing 5 % (v/v) glycerol to stabilize the ¯uorescence signal. Protein samples were incubated in this buffer for 30 minutes before starting experiments. Intrinsic tryptophan ¯uorescence was excited at 290 nm and the corresponding emission spectra were recorded between 310 and 400 nm. Titration experiments were performed at least four times with a ®xed 100 nM concentration of cdks, cyclins, or cdk-cyclin complexes and increasing concentrations of cdc25C. The maxima of the emission spectra were corrected as described by Divita et al. (1993) and plotted against the concentration of human cdc25C. Binding of preformed cdc25C-cyclin complexes to cdks was performed using cdks complexed to mant-ATP, by monitoring enhancement of mant-ATP ¯uorescence at 450 nm following excitation at 340 nm, as described by Heitz et al. (1997). mant-ATP-cdk complexes were kept at a constant concentration of 0.1 mM and incubated for 15 minutes in the presence of different concentrations of cdc25C-cyclin complexes before starting the experiments. Values were ®tted according to a standard quadratic equation using the Gra®t software (Erithacus Software Ltd) with the following equation:
Interactions Between Human cdc25C, Cyclins and cdks
F Fini ÿ f
F
Et L Kd ÿ
Et L Kd 2 ÿ 4Et L1=2 g=2Et in which F is the relative ¯uorescence intensity, Fini is the relative ¯uorescence intensity at the beginning of the titration, F is the variation of ¯uorescence intensity between the initial value and at the saturating concentrations of substrate (L), and Kd is the dissociation of the enzyme-substrate complex. Et corresponds to the total concentration of cdks, cyclins or cdk/cyclin complexes in intrinsic ¯uorescence measurements or of mant-ATP/ cdk complexes in extrinsic ¯uorescence measurements. HPLC gel filtration experiments Gel ®ltration HPLC were performed as described (Heitz et al., 1997), using two Bio-Sil HPLC columns, TSK-250 and TSK-125 (Bio-Rad) in series. Before injecting onto the columns, samples of 50 mg cdc25C, cyclins, cdks and preformed cdk-cyclin complexes were incubated either alone or with the same amount of recombinant cdc25C for 30 minutes at 25 C in 50 mM potassium phosphate buffer (pH 7.2), 150 mM NaCl (for cyclins and cdk-cyclin complexes), or in l50 mM potassium phosphate buffer (pH 6.8; for cdks), to allow the formation of complexes. HPLC runs were performed in the same buffer at a ¯ow rate of 0.8 ml per minute. Peak fractions were analysed by electrophoresis on SDS-12.5 % polyacrylamide gels, followed by Western blotting.
Acknowledgements This work was supported by grants from the CNRS and from L'Association de Recherche Contre le Cancer, ARC 1244, and La Ligue de Recherche Contre le Cancer to G.D. We thank Dr M. DoreÂe for continuous support of this work, as well as for fruitful discussions and critical reading of the manuscript. Likewise, we thank Dr F. Heitz and Dr S. I. Reed for critical reading of the manuscript and for helpful discussions.
References Amrein, K. E., Takacs, B., Stieger, M., Molnos, J., Flint, N. A. & Burn, P. (1995). Puri®cation and characterization of recombinant human p50csk protein-tyrosine kinase from an Escherichia coli expression system overproducing the bacterial chaperones GroES and GroEL. Proc. Natl Acad. Sci. USA, 92, 1048-1052. Baldin, V., Cans, C., Superti-Furga, G. & Ducommun, B. (1997a). Alternative splicing of the human cdc25B tyrosine phosphatase. Possible implications for growth control? Oncogene, 14, 2485-2495. Baldin, V., Cans, C., Knibiehler, M. & Ducommun, B. (1997b). Phosphorylation of human cdc25B phosphatase by cdk1-cyclin A triggers its proteasomedependent degradation. J. Biol. Chem. 272, 3273132734. Bourne, Y., Watson, M. H., Hickey, M. J., Holmes, W., Rocque, W., Reed, S. I. & Tainer, J. A. (1996). Crystal structure and mutational analysis of the human
485 cdk2 kinase complex with cell cycle regulatory protein CksHs1. Cell, 84, 863-874. Brown, N. R., Noble, M. E. M., Endicott, J. A., Garman, E. F., Wakatsuki, S., Mitchell, E., Rasmussen, B., Hunt, T. & Johnson, L. N. (1995). The crystal structure of cyclin A. Structure, 3, 1235-1247. Burstein, E. A., Vedenkina, N. S. & Ivkova, M. N. (1973). Fluorescence and the location of tryptophan residues in protein molecules. Photochem. Photobiol. 18, 263-279. Chen, I. T., Akamatsu, M., Smith, M. L., Lung, F. D. T., Duba, D., Roller, P. P., Fornace, A. J. & O'Connor, P. M. (1996). Characterization of p2l (Cipl/Wa¯) peptide domains required for cyclin E/cdk2 and PCNA interaction. Oncogene, 12, 595-607. Chen, J., Jackson, P. K., Kirschner, M. W. & Dutta, A. (1995). Separate domains of p2l involved in the inhibition of cdk kinase and PCNA. Nature, 374, 386-388. Chen, J., Saha, P., Kornbluth, S., Dynlacht, B. & Dutta, A. (1996). Cyclin-binding motifs are essential for the function of p21CIP1. Mol. Cell. Biol. 16, 4673-4682. Conklin, D. S., Galaktionov, K. & Beach, D. (1995). 14-33 proteins associate with cdc25 phosphatases. Proc. Natl Acad. Sci. USA, 92, 7892-7896. Dasso, M. & Newport, J. W. (1990). Completion of DNA replication is monitored by a feedback system that controls the initiation of mitosis in vitro: studies in Xenopus. Cell, 61, 811-823. De Bondt, H., Rosenblatt, J., Jancank, J., Jones, H. D., Morgan, D. O. & Kim, S.-H. (1993). Crystal structure of cyclin-dependent kinase 2. Nature, 363, 595602. Devault, A., Fesquet, D., Cavadore, J.-C., Garrigues, A.-M., LabbeÂ, J.-C., Lorca, T., Picard, A., Philippe, M. & DoreÂe, M. (1992). Cyclin A potentiates maturation-promoting factor activation in the early Xenopus embryo via inhibition of the tyrosine kinase that phosphorylates cdc2. J. Cell Biol. 118(5), 1109-1120. Divita, G., Goody, R. S., Gautheron, D. C. & Di Pietro, A. (1993). Structural mapping of catalytic site with respect to a-subunit and noncatalytic site in yeast mitochondrial F1-ATPase using ¯uorescence resonance energy transfer. J. Biol. Chem. 268, 13178-13186. DoreÂe, M. & Galas, S. (1994). The cyclin-dependent protein kinases and the control of cell division. FASEB J. 8, 1114-1121. Draetta, G. & Beach, D. H. (1988). Activation of cdc2 protein kinase during mitosis in human cells: cell cycle-dependent phosphorylation and subunit rearrangement. Cell, 54, 17-26. Dunphy, W. G. & Kumagai, A. (1991). The cdc25 protein contains intrinsic phosphatase activity. Cell, 67, 189196. Enoch, T., Gould, K. L. & Nurse, P. (1991). Mitotic checkpoint control in ®ssion yeast. Cold Spring Harbor Symp. Quant. Biol. 56, 409-416. Enoch, T., Carr, A. M. & Nurse, P. (1992). Fission yeast genes involved in coupling mitosis to completion of DNA replication. Genes Dev. 6, 2035-2046. Furnari, B., Rhind, N. & Russell, P. (1997). Cdc25 mitotic inducer targeted by Chk1 DNA damage checkpoint kinase. Science, 277, 1495-1497. Gabrielli, B. G., Lee, M. S., Walker, D. H., PiwnicaWorms, H. & Maller, J. L. (1992). Cdc25 regulates the phosphorylation and activity of the Xenopus cdk2 protein kinase complex. J. Biol. Chem. 267(25), 18040-18046.
486 Gabrielli, B. G., De Souza, C. P. C., Tonks, I. D., Clark, J. M., Hayward, N. K. & Ellem, K. A. O. (1996). Cytoplasmic accumulation of cdc25B phosphatase in mitosis triggers centrosomal microtubule nucleation in HeLa cells. J. Cell Sci. 109, 1081-1093. Galaktionov, K. & Beach, D. (1991). Speci®c activation of cdc25 tyrosine phosphatases by B-type cyclins: evidence for multiple roles of mitotic cyclins. Cell, 67, 1181-1194. Galaktionov, K., Lee, A. K., Eckstein, J., Draetta, G., Meckler, J., Loda, M. & Beach, D. (1995a). Cdc25 phosphatases as potential human oncogenes. Science, 269, 1575-1577. Galaktionov, K., Jessus, C. & Beach, D. (1995b). Raf1 interaction with cdc25 phosphatase ties mitogenic signal transduction to cell cycle activation. Genes Dev. 9, 1046-1058. Gautier, J., Matsukawa, T., Nurse, P. & Maller, J. (1989). Dephosphorylation and activation of Xenopus p34cdc2 protein kinase during the cycle. Nature, 339, 626-629. Gautier, J., Solomon, M. J., Booher, R. N., Bazan, J. F. & Kirshner, M. W. (1991). cdc25 is a speci®c tyrosine phosphatase that directly activates p34cdc2. Cell, 67, 197-211. Gould, K. L. & Nurse, P. (1989). Tyrosine phosphorylation of the ®ssion yeast cdc2 protein kinase regulates entry into mitosis. Nature, 342, 39-45. Hannon, G. J., Demetrick, D. & Beach, D. (1993). Isolation of the Rb-related p130 through its interaction with cdk2 and cyclins. Genes Dev. 7, 2378-2391. Heitz, F., Morris, M. C., Fesquet, D., Cavadore, J.-C., DoreÂe, M. & Divita, G. (1997). Interactions of cyclins with cyclin-dependent kinases: a cominon interactive mechanism. Biochemistry, 36, 4995-5003. Hoffmann, I., Clarke, P. R., Marcote, M. J., Karsenti, E. & Draetta, G. (1993). Phosphorylation and activation of human cdc25-C by cdc2-cyclin B and its involvement in the self-ampli®cation of MPF at mitosis. EMBO J. 12(1), 53-63. Hoffmann, I., Draetta, G. & Karsenti, E. (1994). Activation of the phosphatase activity of human cdc25A by a cdk2-cyclin E dependent phosphorylation at the Gl/S transition. EMBO J. 13(18), 4302-4310. Izumi, T. & Maller, J. L. (1993). Elimination of cdc2 4 phosphorylation sites in the cdc25 phosphatase block initiation of M-phase. Mol. Biol. Cell, 4, 13371350. Izumi, T. & Maller, J. L. (1995). Phosphorylation and activation of the Xenopus cdc25 phosphatase in the absence of cdc2 and cdk2 kinase activity. Mol. Biol. Cell, 6, 215-226. Izumi, T., Walker, D. H. & Maller, J. L. (1992). Periodic changes in phosphorylation of the Xenopus cdc25 phosphatase regulate its activity. Mol. Biol. Cell, 3, 927-939. Jeffrey, P. D., Russo, A. A., Polyak, K., Gibbs, E., Hurwitz, J., MassagueÂ, J. & Pavletich, N. P. (1995). Mechanism of cdk activation revealed by the structure of a cyclin A-cdk2 complex. Nature, 376, 313320. Jessus, C. & Beach, D. (1992). Oscillation of MPF is accompanied by periodic association between cdc25 and cdc2-cyclin. Cell, 68, 323-332. Jinno, S., Suto, K., Nagata, A., Igarashi, M., Kanaoka, Y., Nojima, H. & Okayama, H. (1994). Cdc25A is a novel phosphatase functioning early in the cell cycle. EMBO J. 13(7), 1549-1556.
Interactions Between Human cdc25C, Cyclins and cdks Krek, W., Xu, G. & Livingstone, D. M. (1995). Cyclin A-kinase regulation of E2F-1 DNA binding function underlies suppression of an S phase checkpoint. Cell, 83, 1149-1158. Kumagai, A. & Dunphy, W. G. (1991). The cdc25 protein controls tyrosine dephosphorylation of the cdc2 protein in a cell-free system. Cell, 64, 903-914. Kumagai, A. & Dunphy, W. G. (1992). Regulation of the cdc25 protein during the cell cycle in Xenopus extracts. Cell, 70, 139-151. LabbeÂ, J.-C., Picard, A., Peaucellier, G., Cavadore, J.-C., Nurse, P. & DoreÂe, M. (1989). Puri®cation of MPF from star®sh: identi®cation as the H1 histone kinase p34cdc2 and a possible mechanism for its periodic activation. Cell, 57, 253-263. Laemmli, U. K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature, 227, 680-685. Lamb, N. J. C., Morris, M. C. & Fernandez, A. (1994). Mitotic Activity of cdc25C protein phosphatase: control by phosphorylation and subcellular localization in normal mammalian cells. Advan. Protein Phosphatases, 8, 143-161. Lee, M. S., Ogg, S., Xu, M., Parker, L. L., Donoghue, D. J., Maller, J. L. & Piwnica-Worms, H. (1992). cdc25 encodes a protein phosphatase that dephosphorylates p34cdc2. Mol. Biol. Cell, 3, 73-84. Li, Y., Graham, C., Lacy, S., Duncan, A. M. V. & Whyte, P. (1993). The adenovirus ElA-associated 130 KD protein is encoded by a member of the retinoblastoma gene family and physically interacts with cyclins A and E. Genes Dev. 7, 2366-2377. Lorca, T., LabbeÂ, J.-C., Devault, A., Fesquet, D., Capony, J.-P., Cavadore, J.-C., Le, Bouffant F. & DoreÂe, M. (1992). Dephosphorylation of cdc2 on threonine 161 is required for cdc2 kinase inactivation and normal anaphase. EMBO J. 11, 2381-2390. Lundgren, K., Walworth, N., Booher, R., Dembski, M., Kirschner, M. & Beach, D. (1991). Mik1 and wee1 cooperate in the inhibitory tyrosine phosphorylation of cdc2. Cell, 64, 1111-1122. Luo, Y., Hurwitz, J. & Massague, J. (1995). Cell-cycle inhibition by independent cdk and PCNA binding domains in p21-CIP1. Nature, 375, 159-161. Meijer, L., Azzi, L. & Wang, J. Y. J. (1991). Cyclin B targets for p34cdc2 for tyrosine phosphorylation. EMBO J. 10(6), 1545-1554. Millar, J. B. A., McGowan, C. H., Lenaers, G., Jones, R. & Russell, P. (1991a). p80cdc25 Mitotic inducer is the tyrosine phosphatase that activates p34cdc2 kinase in ®ssion yeast. EMBO J. 10(13), 4301-4309. Millar, J. B. A., Blevitt, J., Gerace, L., Sadhu, K., Featherstone, C. & Russell, P. (1991b). p55cdc25 is a nuclear protein required for the initiation of mitosis in human cells. Proc. Natl Acad. Sci. USA, 88, 1050010504. Morgan, D. O. (1995). Principles of cdk regulation. Nature, 374, 131-134. Morris, C. M., Heitz, F. & Divita, G. (1998). Kinetics of dimerization and interactions of p13suc1 with cyclindependent kinases. Biochemistry, 37, 14257-14266. Mueller, P. R., Coleman, T. R., Kumagai, A. & Dunphy, W. G. (1995). Mytl: a membrane-associated inhibitory kinase that phosphorylates cdc2 on both threonine-14 and tyrosine-15. Science, 270, 86-90. Nagata, A., Igarashi, M., Jinno, S., Suto, K. & Okayama, H. (1991). An additional homolog of the ®ssion yeast cdc25 gene occurs in humans and is highly
Interactions Between Human cdc25C, Cyclins and cdks expressed in some cancer cells. New Biol. 3(10), 959968. Ogg, S., Gabrielli, B. & Piwnica-Worms, H. (1994). Puri®cation of a serine kinase that associates with and phosphorylates human cdc25C on serine 216. J. Biol. Chem. 269(48), 30461-30469. Parker, L. L., Atherton-Fessler, S. & Piwnica-Worms, H. (1992). p107wee1 is a dual-speci®city kinase that phosphorylates p34cdc2 on tyrosine 15. Proc. Natl Acad. Sci. USA, 89, 2917-2921. Peng, C.-Y., Graves, P. R., Thoma, R. S., Wu, Z., Shaw, A. S. & Piwnica-Worms, H. (1997). Mitotic and G2 checkpoint control: regulation of 14-3-3 protein binding by phosphorylation of cdc25C on serine 216. Science, 277, 1501-1505. Russell, P. & Nurse, P. (1986). cdc25 Functions as an inducer in the mitotic control of ®ssion yeast. Cell, 45, 145-1533. Russo, A. A., Jeffrey, P. D., Patten, A. K., Massague, J. & Pavletich, N. P. (1996). Crystal structure of the p27KIPI cyclin-dependent kinase inhibitor bound to the cyclin A-cdk2 complex. Nature, 382, 325-331. Sadhu, K., Reed, S. I., Richardson, H. & Russell, P. (1990). Human homolog of ®ssion yeast cdc25 mitotic inducer is predominantly expressed in G2. Proc. Natl Acad. Sci. USA, 87, 5139-5143. Saba, P., Eichbaum, Q., Silberman, E. D., Mayer, B. J. & Dutta, A. (1997). p2lC1P1 and JD cdc25A: competition between an inhibitor and an activator of cyclin-dependent kinases. Mol. Cell. Biol. 17(8), 43384345. Sanchez, Y., Wong, C., Thoma, R. S., Richman, R., Wu, Z., Piwnica-Worms, H. & Elledge, S. J. (1997). Conservation of the Chk1 checkpoint pathway in mammals: linkage of DNA damage to cdk regulation through cdc25. Science, 277, 1497-1501.
487 Smith, P. K., Krohn, R. I., Hermanson, G. T., Mallia, A. K., Gartner, F. H., Provenzano, M. D., Fujimoto, E. K., Goeke, N. M., Olson, B. J. & Klenk, D. C. (1985). Measurement of protein using bicinchonic acid. Anal. Biochem. 150, 76-85. Solomon, M. J., Glotzer, M., Lee, T. H., Philippe, M. & Kirschner, M. W. (1990). Cyclin activation of p34cdc2. Cell, 63, 1013-1024. Strausfeld, U., Labbe, J.-C., Fesquet, D., Cavadore, J.-C., Picard, A., Sadhu, K., Russell, P. & DoreÂe, M. (1991). Dephosphorylation and activation of a p34cdc2/cyclin B complex in vitro by human cdc25 protein. Nature, 351, 242-245. Strausfeld, U., Fernandez, A., Capony, J.-P., Girard, F., Lautredou, N., Derancourt, J., Labbe, J.-C. & Lamb, N. J. C. (1994). Activation of p34cdc2 protein kinase by microinjection of human cdc25C into mammalian cells: requirement for prior phosphorylation of cdc25C by p34cdc2 on sites phosphorylated at mitosis. J. Biol. Chem. 269(8), 5989-6000. Tonks, N. K., Diltz, C. D. & Fischer, E. H. (1988). Characterization of the major protein tyrosine phosphatases of human placenta. J. Biol. Chem. 263, 6731-6737. Weinert, T. (1997). A DNA damage checkpoint meets the cell cycle engine. Science, 277, 1450-1451. Xu, X. & 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(9), 5118-5124. Zheng, X.-F. & Ruderman, J. V. (1993). Functional analysis of the P box, a domain in cyclin B required for the activation of cdc25. Cell, 75, 155-164. Zhu, L., Harlow, E. & Dynlacht, B. D. (1995). p107 uses a p2l-related domain to bind cyclin/cdk2 and regulate interactions with E217. Genes Dev. 9, 1740-1752.
Edited by J. Karn (Received 7 July 1998; received in revised form 4 December 1998; accepted 9 December 1998)