Biochemical Characterization of the ~ 3 4 ' ~ ' ~ Protein Kinase Component of Purified Maturation-promoting Factor from Xenopus Eggs*

Genetic studies in the fission yeast Schizosaccharomyces pombe and biochemical data in oocytes and eggs of Xenopus laevis have implicated the product of the cdc2+ gene as critical for the G2 to M transition in the cell cycle. The product of the cdc2+ gene is a 34-kDa serinefthreonine protein kinase, designated ~ 3 4 ~ " , that is a component of purified maturation-promoting factor (MPF) and also of purified mammalian growth-associated histone H1 kinase. The biochemical prop-erties of ~34"'"' H1 kinase activity in the MPF complex were studied. Phosphorylation of the ~ 4 5 ~ ' " " component in the MPF complex by ~ 3 4 ~ " exhibited kinetics consistent with an intramolecular reaction. On glycerol gradient centrifugation, MPF kinase against several substrates sedimented with an apparent M, = 45,000-55,000. ~ 3 4 ~ " was found to utilize ATP, GTP, and adenosine 5'-0-(3-thiotriphosphate) with apparent K,,, values of 75, 700, and 250 PM, respec-tively. The kinase activity was inhibited by 0-glycer-ophosphate, NaF, and zinc, whereas p-nitrophenyl phosphate was slightly stimulatory. The relative rates of phosphorylation of various substrates by MPF and growth-associated H1 kinase were similar. These findings should prove useful in further work of 15 pl of glacial phosphorylated isolated on Radioactivity quantified Reactions triplicate and for radioactivity in the absence of enzyme. For some experiments the reactions were terminated by the of one-fourth of five times concentrated electrophoresis

Recent years have seen remarkable advances in our understanding of the molecular basis of cell cycle control in mitosis ("phase). New insights have come primarily from genetic analysis in the fission yeast Schizosaccharomyces pombe (1) and from biochemical analysis in oocytes and eggs of the frog, Xenopus luevis ( 2 ) . In S. pombe, a substantial body of evidence has accumulated to indicate that the product of the cdc2+ gene is critically involved in the G2 to M transition. The product of the cdc2+ gene is a 34-kDa serine/threonine protein kinase, designated ~34'~'', that is regulated in a complex fashion by a number of other cell-cycle components, several of which are also protein kinases (3)(4)(5)(6)(7). A human homolog of cdc2+ has been cloned from a HeLa cell cDNA library by complementation analysis (8), suggesting that the role of p3Pdc2 is sufficiently conserved to support function of the human gene product in yeast.
In Xenopus, biochemical analysis of the cell cycle has been concerned largely with the characterization of a cytoplasmic * This work was supported by National Institutes of Health Grant GM26743. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. activity known as maturation-promoting factor (MPF).' MPF was first described in 1971 as an activity in "phase cytoplasm of maturing Xenopus oocytes that could cause immature oocytes to undergo meiotic maturation without hormonal stimulation or protein synthesis (9,10). Cell-cycle studies showed that MPF activity appears not only in Xenopus oocytes undergoing meiotic maturation but also in maturing oocytes of other species, and during mitosis in mammalian cells and the budding yeast Saccharomyces cerevisiae (11)(12)(13)(14)(15). Despite the apparent fundamental importance of MPF in the cell cycle, until recently little progress had been made in its purification, partly because the oocyte microinjection assay can detect MPF activity only in highly concentrated fractions (16) and because its activity is highly unstable.
Extracts of Rana eggs and Xenopus eggs containing MPF activity have been shown to cause isolated sperm chromatin to form chromosomes in vitro (17,lS). Moreover, the addition of partially purified MPF, which had been detected by the oocyte microinjection assay, caused nuclear envelope breakdown, chromosome condensation, and spindle formation in vitro (18, 19). The induction of these early mitotic events in vitro formed the basis of a new assay for MPF, which allowed the purification of Xenopus MPF to near homogeneity (20). Highly purified MPF, which can induce "phase both in the cell-free system and in microinjected oocytes, consists of two proteins of M, = 34,000 and 45,000. The final preparation expresses a protein kinase activity that can phosphorylate the 45-kDa component, H1 histone, protein phosphatase inhibitor 1, and a-casein (20).
It seemed likely that the 34-kDa component of MPF possessed the protein kinase activity because the 45-kDa component became phosphorylated only in fractions that also contained the 34-kDa component (20). Because of the similar molecular weight of this component of MPF and that of the cdcP gene product, experiments were carried out to determine whether there was any relationship between these two proteins. These studies used antisera raised against a unique stretch of 16 amino acids, termed the PSTAIR sequence, that is perfectly conserved in ~3 4 '~'~ and its homologs from yeast to man, but that is not present in any other members of the protein kinase family so far sequenced (8). This antibody immunoblotted the 34-kDa component of MPF, indicating that it is the product of a Xenopus homolog of the cdc2' gene (21). In addition, both components of MPF could be coimmunoprecipitated, indicating that the 45-kDa protein exists in a complex with the 34-kDa protein kinase (21). Moreover, immunoprecipitated Xenopus ~3 4 '~'~ retains the ability to  H1 kinase and can immunoprecipitate the growth-associated H1 histone kinase activity (28). Moreover, the homologs of ~34'~'' from Xenopus, budding yeast, and fission yeast specifically phosphorylate the same sites in H1 histone as mammalian growth-associated H1 kinase. In contrast, extracts of budding yeast deficient in the CDC28 gene, a homolog of cdc2 (29), are unable to phosphorylate these sites (28). Together these studies suggest that the growth-associated H1 kinase from mammalian cells also contains a ~3 4~~' ' homolog, and that ~3 4 "~" is a protein kinase that functions in "phase control not only in yeasts, but also in Xenopus and mammalian cells. Although these four protein kinases, identified in yeasts, frog, and rat, now appear to be the same or very closely related, little is known about their biochemical properties. We report here characterization of ~3 4 '~' ' kinase activity in the Xenopus MPF complex.

EXPERIMENTAL PROCEDURES
Materials-MPF was purified from unfertilized eggs of X. laeuis as previously described (20) except that the active fractions from the Mono S column were not dialyzed before freezing and storage at -70 "C. products of the reaction were resolved by electrophoresis through a polyacrylamide gel and radioactivity was quantified by liquid scintillation spectroscopy of the excised bands, as described previously (32). %%Labeled proteins were visualized by fluorography with sodium salicylate (33), and radioactivity was quantified by liquid scintillation spectroscopy of the excised bands.

RESULTS AND DISCUSSION
In initial studies MPF was shown to elute as a broad peak of M, = 200,000 upon gel filtration through a TSK 3000 SW column. In subsequent preparations, however, the position of elution of MPF from this column gradually changed to that of a lower molecular weight (20). In all cases elution of H1 kinase activity coincided with that of nuclear breakdown activity and active fractions contained both ~3 4~' ' and ~45"~~"". In order to address the question of its molecular size  Products of the reaction were resolved by polyacrylamide gel electrophoresis, and radioactivity in p45"r"" was determined by liquid scintillation spectrometry. Other experiments showed that incorporation of radiolabel into ~4 5~' ' " was linear for at least 120 s for 20 p l of MPF/reaction. were carried out as described under "Experimental Procedures" except that [T-~*P]GTP (2 cpm/fmol) was present at concentrations from 50 to 500 PM, and the products of the reaction were analyzed by polyacrylamide gel electrophoresis. C, kinase reactions were done as in a different manner, a sample of MPF was subjected to glycerol gradient centrifugation, and fractions were assayed for kinase activity. As shown in Fig. 1, H1 kinase activity sedimented with an apparent M, of 45,000-55,000. The same sedimentation behavior was seen with various MPF preparations regardless of where they had eluted from the TSK column, with or without further purification. The ionic conditions for the two procedures are similar, and the reason for this discrepancy in apparent M , is not clear. It should be noted, however, that the sedimentation behavior shown here is similar to that of the major peak of MPF activity, assayed by microinjection, when crude extracts were analyzed by sucrose gradient centrifugation (34). In other studies, partially purified MPF was reported to elute as a peak of 90-100 kDa from a Sephacryl ,9200 column (35). These results are most consistent with Xenopus MPF existing as a heterodimer, but further work is needed to establish the ratio of subunits in purified Xenopus MPF preparations. MPF isolated from starfish oocytes has been reported to exist has a heterodimer containing one molecule of ~3 4~~" ' and one molecule of p4FYCfiR (36).
Recently, it was found that highly purified MPF expressed kinase activity against MAP-2 also. Fig. 1 illustrates that MAP-2 kinase activity co-sedimented with H1 kinase activity. In addition, kinase activity against the 45-kDa protein, acasein, and inhibitor 1 also co-sedimented with H1 kinase activity (data not shown). These results are consistent with the idea that there is only one kinase present in MPF and that the phosphorylation of the proteins other than H1 is not due to a contaminating kinase(s). Phosphoamino acids detected in the substrates phosphorylated by MPF were as follows: 45-kDa protein, HI, and MAP-2, phosphoserine and phosphothreonine; a-casein, phosphothreonine; inhibitor 1, phosphoserine (data not shown).
As originally reported by Lohka et al. (20), incubation of purified MPF with [Y-~'P]ATP led to phosphorylation of the ~4 5~' " " component. The kinase responsible was most likely ~3 4 '~' ' because the preparation was highly purified and phosphorylation of p45sy'"" occurred only in fractions that also contained ~34"~"'. Since both subunits of MPF exist together in a complex, it seems likely that p45'Y'f'n phosphorylation would occur by an intramolecular mechanism. To assess this possibility directly, the rate of phosphorylation of p4FYc"" in a purified MPF fraction was monitored as a function of MPF concentration in the assay. As shown in Fig. 2, the rate of the reaction increased only slightly over a 20-fold range of MPF concentrations. This result is consistent with p45Q'cfin phosphorylation in the MPF complex occurring by an intramolecular mechanism. Other experiments have shown, however, that purified MPF can also phosphorylate exogenous, bacterially produced cyclin.* In other experiments the ability of highly purified MPF to use ATP, GTP, and ATPyS was investigated. As shown in Fig. 3, for H1 phosphorylation the enzyme was able to use all three nucleotide triphosphates, with apparent K , values of 75, 700, and 250 p M , respectively. Because the amount of p34'd'2 in these assays was too low to be measured, we were unable to determine a reliable specific activity value. However, at saturating nucleotide concentrations, similar amounts of described under "Experimental Procedures" except that [Y-~~S]ATP (314 dpm/pmol) was present at concentrations from 15 to 1000 PM. Products of the reaction were analyzed by polyacrylamide gel electrophoresis, and radiolabel incorporated into HI was determined by liquid scintillation spectrometry. D, kinase reactions were performed as described under "Experimental Procedures" except that the concentration of [y-32P]ATP (3 cpm/fmol) was 500 PM and H1 histone was present in concentrations from 0.0125 to 0.5 mg/ml.  phosphate (70 pmol/lO min) were transferred to H1 histone with ATP and GTP, suggesting that the VmaX values with either nucleotide are similar. Phosphorylation of the 45-kDa component of the MPF complex also occurred with GTP and ATPyS (data not shown). Incubation in the presence of hexokinase and glucose such that >95% of the [y3'P]ATP was abolished (37) had no effect on the incorporation of radioactivity into H1 from [y-3'P]GTP. Therefore the use of GTP by MPF kinase is not the result of contamination of the GTP with ATP or of a contaminating GTP-ADP phosphotransferase activity. In addition, neither ATPase nor GTPase activity was detectable in the MPF preparation. Other studies using ATP as the nucleotide indicated MPF has an apparent K , for H1 of 40 pg/ml (Fig. 3D), similar to the value for mammalian growth-associated H1 kinase?
Some, but not all, protein kinases have been reported to utilize both M$+ and Mn'+, and in S. cereukiae, the CDC28 kinase activity in immunoprecipitates is stimulated in the presence of zinc (38). Consequently, the effects of these divalent cations on p3PdC2 kinase activity were investigated. As shown in Fig. 4 the metal ion, the pH optimum for activity was found to be 7.5 to 8.5 (Fig. 5 A ) . In addition, as illustrated in Fig. 5B, the H1 kinase activity of MPF is very thermolabile. Incubation of the enzyme at temperatures above 30 "C caused a substantial decrease in activity, with incubation at 50 "C for 5 min resulting in a complete loss of activity.
MPF activity is highly unstable in uitro and often cannot be extracted from cells in the absence of @-glycerophosphate and sodium fluoride, compounds that are thought to act by inhibiting protein phosphatase activity (39). In fact, MPF is unable to cause nuclear envelope breakdown and chromosome condensation in oocytes or in the cell-free system in the absence of @-glycerophosphate. Although @-glycerophosphate is also known to have a marked stabilizing effect on phosphorylase b kinase (40), other work in this laboratory has shown that both @-glycerophosphate and NaF are potent inhibitors of ribosomal protein S6 kinase I1 from Xenopus eggs (32). To investigate the effect of these substances on ~34'~'*, assays using H1 as substrate were carried out in the presence of varying concentrations of these compounds. As shown in Fig.  6, A and B , both @-glycerophosphate and NaF inhibited MPF kinase activity, with apparent 1C5o values of 75 and 7 mM, respectively. However, p-nitrophenyl phosphate, another widely used protein phosphatase inhibitor, had a slight stimulatory effect on MPF kinase activity (Fig. 6C). The fact that @-glycerophosphate is required for mitotic effects of MPF in the cell-free system yet inhibits the kinase activity in uitro suggests that the positive effect of @-glycerophosphate is directed at other components in the cell-free system. It also indicates the necessity of using different buffer systems to assay MPF as protein kinase activity uersus nuclear breakdown.
As indicated in the introduction, both MPF and growthassociated H1 kinases contain homologs of ~3 4 '~" ' as one component of their structure. This suggests that the substrate specificity and relative rates of phosphorylation of various substrates should be similar for both enzymes. To evaluate this prediction, the relative rates of phosphorylation of H1 histone, a-casein, inhibitor 1, and MAP-2 were compared for highly purified MPF and partially purified growth-associated H1 kinase. As shown in Table I, the relative activity of the two enzymes was similar against these substrates, with MAP-2 being phosphorylated at 40% of the rate of H1 histone, and the other substrates at levels below 20% of that with HI. Similar results were obtained when GTP was used as the phosphate donor (data not shown).
These results provide an initial characterization of the protein kinase component of MPF, ~34'~''. This kinase is unusual in its highly restricted substrate specificity and in its ability to use both ATP and GTP and also ATPyS. Other than growth-associated H1 kinase, which contains a ~34'~'' homolog, only one other serine/threonine kinase, casein kinase 11, has been found previously to utilize both ATP and GTP (41). However, that enzyme is markedly inhibited by heparin (411, whereas heparin did not inhibit MPF H1 kinase activity at concentrations up to 200 pg/ml (data not shown). The ability of the enzyme to use GTP may facilitate identification of substrates other than H1, since assays with GTP should significantly reduce the level of background phosphorylation by other kinases in partially purified preparations.
Studies with human cells have indicated that only about 5% of ~3 4 "~" is present as active protein kinase at mitosis (42,43). This suggests that ~3 4~~' ' is likely to have other functions besides control of "phase, and indeed, ~3 4 '~' ' is known to have an obligatory role in the G1 phase of the yeast cell cycle (1). These different roles of ~34'~"' suggest that the kinase may associate with different components in different cells or in different phases of the cell cycle. Some evidence for this already exists since different proteins are co-purified or coimmunoprecipitated with ~34'~'' in budding yeast, fission yeast, Xenopus MPF, starfish egg MPF, HeLa cells, and growth-associated H1 kinase (20, 21, 23, 42, 43). At present the function of these associated components is unknown. In some cases these associated components seem to be substrates for ~34'~"', whereas in other cases they do not. Whether they affect the activity of ~34'~'' or its substrate specificity will require analysis of other purified preparations of ~34'~'' complexed with other proteins.