Purification and Characterization of a Maturation-activated Basic Protein Kinase from Sea Star Oocytes*

A meiosis-activated myelin basic protein (MBP) kinase was purified approximately 8700-fold from sol- uble post-germinal vesicle breakdown extracts from maturing oocytes of the sea star Pisaster ochraceus. Purification to apparent homogeneity was achieved by sequential chromatography on DEAE-cellulose, hydroxylapatite, phosphocellulose, phenyl-Sepharose, heparin-Sepharose, polylysine-Sepharose, and Mono-&. The final product exhibited an apparent molecular mass of -42 kDa by both native gradient and soduim dodecyl sulfate-polyacrylamide gel electrophoresis, and this precisely correlated with the chromatographic behavior of the recovered MBP kinase activity on a Superose 6/12 column. The kinase utilized the MBP as the major substrate with little or no phosphorylation of histones (Hl, H2A, or H2B), casein, phosvitin, prot- amine, or 40 S ribosomal proteins. The purified enzyme was

as Mn'+, Zn'+, and Ca". The true K,,, values for ATP and myelin basic protein were determined to be 58 and 25 pM, respectively, using double-reciprocal plots. The purified enzyme was unable to utilize GTP in place of ATP. The enzyme was shown to rapidly undergo autophosphorylation.
The autophosphorylation was sensitive to alkali treatment implying that phosphate was incorporated on serinelthreonine residues. The properties of this MBP kinase are reminiscent of a protein kinase that is also activated in a cyclic fashion at Mphase during the early cell divisions of sea star and sea urchin embryos (Pelech, S. L., Tombe, R., Meijer, L., and Krebs, E. G. (1988) Dev. Biol. 130,26-36).
Meiotic maturation of oocytes features many parallels with mitotic cell division, including disintegration of the nuclear envelope and chromosome condensation. Just prior to these events in maturing echinoderm (I) and amphibian (2) oocytes, there is a large burst in the phosphorylation of numerous proteins, implicating the concerted action of a cascade of protein kinases in regulating these and other processes required for cell division. We have previously identified at least five distinct protein kinases that become activated when sea star (3) and Xenopus laeuis (4)  Recent studies have demonstrated that homogeneous preparations of one of these, a histone Hl kinase, are sufficient to trigger a precocious maturation response when microinjected into immature prophase-arrested sea star (5) and Xenopus (6) oocytes. By a number of criteria, including physicochemical, immunological, substrate specificity, and affinity interaction with the 13-kDa product of the yeast sucl gene, these purified preparations of the oocyte histone Hl kinases were shown to be species homologs of the 34-kDa product of the schizosaccharornyces pombe cdc-2 and Saccharomyces cerevisiae cdc-28 genes of yeast (7-10). Genetic studies with S. pombe have established that the p34'*'-' kinase plays a pivot01 role in mitosis (11,X!) and that it is under the negative control of another protein kinase, ~112~"' (13), which in turn is negatively regulated by a third protein kinase, P50"'"' (14). Following fertilization of sea star oocytes (3, 15), sea urchin eggs (3, 16), and Xenopus eggs (17), the histone Hl kinase undergoes a cyclic activation during each of the early mitotic cell divisions. Similarly, in dividing mammalian somatic cells, p34'*'-' kinase activity becomes transiently increased during M-phase (18-20). Phosphorylation of histone Hl at multiple sites, which are also phosphorylated by Xenopus egg p34cdcV2 in vitro (20), correlates with chromosome condensation during mitosis (21-23). Another downstream consequence of p34'*'-' kinase activation in maturing Xenopus oocytes is increased phosphorylation of ribosomal protein S6 at multiple seryl residues (24, 25). This can be attributed to the activation of two Xenopus oocyte S6 kinases (4) that are also detectable in maturing sea star oocytes (3,26). One of these S6 kinases has been purified from Xenopus eggs (27,28) and was shown to be encoded by two genes that specify 74-and 83-kDa proteins (29). The purified Xenopus egg S6 kinase could also phosphorylate nuclear lamins (30, 31), which are phosphorylated during oocyte maturation (31) and mitosis (32). Lamins are thought to serve as an architectural framework for the nuclear envelope, and phosphorylation has been proposed to promote its disassembly (33).
Activation of the S6 kinases during sea star (26, 34) and Xenopus (35) oocyte maturation appears to be mediated through its phosphorylation by an unidentified protein kinase. The purified Xenopus egg S6 kinase II can be activated following phosphorylation at a threonyl residue and an additional site by a 42-kDa protein kinase that is itself phosphorylated on tyrosyl and threonyl residues in mouse 3T3-Ll cells treated with insulin and other mitogens (36, 37). Exposure of Xenopus oocytes to insulin can induce, within 10 min, an acute 2-3-fold activation and, at the onset of germinal vesicle Purification of an M-phase-activated MBP Kinase breakdown (GVBD),' a delayed "r-lo-fold increase in S6 kinase II activity (38, 39). Activation of Xenopus egg S6 kinase II can be reversed by purified protein phosphatases 1 and 2A (36). While microinjection of the histone Hl kinase into Xenopus oocytes can elicit the indirect activation of the S6 kinase (4,39), the frog S6 kinase is not a substrate for purified P34'd' L from the same source.
One of the M-phase-regulated protein kinases that we de-Lane 53 MBP Kinase Activity (pm01 per min per ml) scribed in previous studies (3, 4) exhibited a time course of activation in maturing sea star oocytes that closely correlated with that of the S6 kinases and lagged behind peak stimulation of the histone Hl kinase. This -45kDa protein kinase phosphorylated myelin basic protein (MBP), with little or no activity toward histone Hl or ribosomal protein S6 (3, 4). Although the MBP activity of this kinase was high in sea urchin eggs, it rapidly declined within 5 min following fertilization, and it became transiently elevated at M-phase in concert with the histone Hl kinase in each of the early cell divisions (3). We now described the purification and characterization of this novel M-phase-activated protein kinase from maturing sea star oocytes.
Activation of MBP Kinase(s,J during Oocyte Maturation-Pisaster ochraceus oocytes typically undergo GVBD in -80 min when induced to mature at 14 "C with the hormone lmethyladenine. We have previously described the appearance of an MBP kinase that becomes maximally activated following GVBD at loo-140 min after initial exposure to l-methyladenine (3). Cytosolic extracts from immature oocytes and cells treated for 2 h with l-methyladenine were separately fractionated on hydroxylapatite columns and the kinase activity toward MBP determined (Fig. lA, Miniprint Section). The majority of the maturation-stimulated MBP phosphorylating activity eluted as a broad peak with a conductivity between 10 and 40 mmho, which corresponded to 40-170 mM potassium phosphate. The MBP phosphorylating activity was increased greater than lo-fold during oocyte maturation.
A histone Hl phosphorylating activity was eluted from the hydroxylapatite at higher conductivities between 60 and 80 mmho (Fig. 1B). This histone Hl kinase activity was also stimulated over 20-fold during oocyte maturation and may correspond to sea star p3Ydc-' (7-10).
Electrophoretic and Gel Filtration Analysis of Purified MBP Kinase-The MBP kinase was purified -8700.fold from cytosolic extracts of maturing sea star oocytes (see Miniprint Section). The final Mono-Q preparation essentially contained a single band with an apparent size of -42 kDa. The other higher molecular mass species in the Mono-Q preparation were contaminants from the /3-mercaptoethanol in the sample buffer and appeared throughout the gel. There was a strong correlation between the activity of the MBP kinase in the Mono-Q fractions and the intensity of the silver-stained 4% kDa protein band (Fig. 3). The other lower molecular mass bands present were contaminants, because they also appeared in fractions that lacked MBP kinase activity (data not shown). Thus, the MBP kinase possessed a single subunit of -42 kDa by SDS-PAGE.
To determine the subunit composition of the MBP kinase, native gradient PAGE was performed on the purified enzyme. Silver staining of the gradient polyacrylamide gel also revealed a single protein band migrating with an apparent molecular mass of -42 kDa (data not shown). On a native non-gradient polyacrylamide gel, MBP kinase activity was found to coincide with the major silver-stained protein band, providing strong evidence that the band corresponded to the MBP kinase (Fig. 4).
The purified MBP kinase was subjected to gel filtration on Superose 6 and 12 columns linked in series (Fig. 5). The MBP kinase activity chromatographed as a single peak with an apparent size of -40-45 kDa. Thus, by several criteria, it may be concluded that the native enzyme was composed of a single subunit of -42 kDa.
Kinetic Analysis of Purified MBP Kinase-The protein substrate specficity of the purified MBP kinase was investigated (Fig. 6). MBP was determined to be the major substrate for the kinase with a 20-30-fold higher phosphate incorpora- tion than into other substrates examined. Protamine was not a substrate for the purified kinase, whereas phosphorylation of histones Hl, H2A, and H2B were only 13% of the phosphorylating activity toward MBP. Phosphorylation of casein and phosvitin was -5% of the phosphorylating activity toward MBP. The kinase failed to phosphorylate ribosomal protein S6 in rat liver 40 S ribosomes (Fig. 6). Therefore, with the range of potential substrates tested, the isolated kinase seems to be relatively specific for MBP as a substrate.
The true K,,, values for ATP and MBP were determined by double-reciprocal plots (data not shown). The plots were linear in both cases. The K,,, for ATP was 58 PM with 1 mg MBP/ml.
With 50 pM ATP, the K,,, for MBP was 25 PM. Both plots intersected at the same V,,,,, value. The maximal velocity obtained for the enzyme preparation used in the experiment was 4124 pmol. min-' . ml-'. This specific activity was similar to the specific activity of 4062 pmol.min-'. ml-' obtained by routine assay procedure. This indicates the routine assay was optimal and that substrate inhibition did not occur at the concentrations used in the assay. The MBP kinase was unable to utilize GTP in place of ATP (data not shown).
The concentrations of various compounds that inhibited MBP kinase activity by 50% are shown on Table II. MBP kinase was only inhibited at very high concentrations of pglycerol phosphate, EGTA, NaCl, NaF, heparin, and dithiothreitol. However, the kinase was very sensitive to inhibition by metal ions such as Mn'+, Cat+, and Zn'+. The purified MBP kinase activity was unaffected by 1000 nM CAMPdependent protein kinase inhibitor peptide, 2 pM CAMP, and 1000 units of calmodulin/ml. Autophosphorylation of Purified MBP Kinase-Since a majority of protein kinases have the inherent property of undergoing autophosphorylation, the purified MBP kinase was tested for its ability to self-phosphorylate.
It is evident that MBP kinase does undergo phosphorylation ( Fig. 7B).
Furthermore, the degree of autophosphorylation correlated with the activity of the MBP kinase in the various Mono-Q fractions (Fig. 7, A and B), again providing strong evidence that the 42-kDa band was a kinase. To determine whether the autophosphorylation was on serinelthreonine residues or tyrosine residues, the gel was treated with alkali. This treatment selectively hydrolyzes phosphate from serine/threonine residues but not tyrosine (46). Alkali treatment removed the radiolabeling of the 42-kDa band, indicating that serine/ threonine residues are autophosphorylated on MBP kinase (Fig. 7C). DISCUSSION The -8700-fold purification of the major M-phase-activated MBP kinase from maturing oocytes of the sea star P. ochraceus yielded a single -42-kDa silver-stained protein band on SDS-polyacrylamide gels and gradient buffer gels. Assignment of this protein band as the kinase was supported by its ability to undergo autophosphorylation, the recovery of MBP kinase activity from nondenaturing polyacrylamide gels Purification of an M-phase-activated MBP Kinase 55 only in the position of the band, and the behavior of the purified MBP kinase as a 42kDa protein on Superose. These data indicated that the MBP kinase exists as a monomeric protein in the maturing oocyte. It is possible that it may correspond to the 45-47-kDa phosphoprotein that is detectable in maturing Xenopus (47, 48) and sea star (49) oocytes. The calculated 4% recovery of MBP kinase activity with the developed purification protocol probably represents an underestimate of the actual recovery, because the total MBP kinase activity in the homogenate also included contributions from other MBP kinases, such as a p3Ydc.' (3, 4).
The 42-kDa MBP kinase was clearly distinct from the 34-kDa histone Hl kinase, which does not autophosphorylate (6). The two kinases were well resolved by chromatography on hydroxylapatite (Fig. l), analytical DEAE-Sephacel (3), and phosphocellulose, which retains the histone Hl kinase, but not the MBP kinase (9). Neither kinase bound to heparin-Sepharose (10). Like purified sea star oocyte p34cdc-2 (lo), the MBP kinase could phosphorylate histone Hl and casein, but only 20-fold less effectively than MBP, whereas ~34'~'-' apparently phosphorylated histone Hl and MBP at similar rates ( Fig. 1  (3), and the stimulation of oocyte S6 kinase activity is thought to be a consequence of its phosphorylation by another protein kinase (26,36,37,40), we were intrigued by the prospects that the MBP kinase may function in this capacity. In preliminary studies, we have noted that the reconstitution of cytosol from immature sea star oocytes with the purified MBP kinase, 40 S ribosomes, and ATP does not facilitate an increase in S6 phosphorylation in these extracts.3 However, future studies with purified S6 kinase will be necessary to establish whether there may be direct phosphorylation of the S6 kinase by the MBP kinase. Furthermore, it will be important to test whether purified p34cdc-2 histone Hl kinase can phosphorylate and activate the MBP kinase, as optimal activation of the latter in maturing sea star oocytes is preceded by peak stimulation of the former (3). The only protein kinase that has been demonstrated thus far to phosphorylate and activate Xenopus egg S6 kinase II is a mouse 3T3-Ll cell enzyme, which also phosphorylates itself and microtubule-associated protein 2 (MAP-2) (36, 37). Both the MBP kinase and the MAP-2 kinase fail to bind to phosphocellulose and are adsorbed to phenyl-Sepharose (50, 51). Both kinases did not utilize GTP and poorly phosphorylated, if at all, histones, protamine, casein, phosvitin, and ribosomal protein S6 (50). Partially purified 3T3-Ll MAP-2 kinase can phosphorylate MBP.4 We have found that purified MBP kinase can phosphorylate, in porcine brain microtubule preparations, a high molecular mass protein that resembles MAP-2 (data not shown). However, the sea star MBP kinase appears to differ from the MAP-2 kinase in some respects. The 'J. S. Sanghera, H. B. Paddon, and S. L. Pelech, unpublished observations. 4 T. Sturgill, personal communication.
The 35-kDa MAP-2 kinase has recently been detected in a number of mammalian tissues, including insulin-treated rat hepatocytes (52). In this regard, it is significant that a 36-kDa MBP kinase, which was purified from pig liver, was shown to undergo a Mg-ATP-mediated activation via intramolecular activation (53). This autophosphorylation event was stimulated P-fold by heparin (53). The purification protocol for this liver MBP kinase was quite different from the MAP-2 kinase from mouse 3T3-Ll cell (51) and the MBP kinase in our study, thus rendering it difficult to compare these kinases. The eventual cloning and sequencing of the mammalian MAP-2 kinase and the echinoderm MBP kinase will resolve the question of how closely related these enzymes are. 16 (Arg-Arg-La-Se,-Se,-Leu-Arg-Ala) and CAMP dcpcndcnt protein kinase inbibimr pcplide (Thr-Thr-Tyr-Ala-Asp-Phe-Jle~Ala-Se,-Gly-A,g-Th,-G~y-A,g-A,g-A~~-Al~-ll~-His-Asp) In some crpcrimcmr, the MBP was "placed by 1 "g hirmnc U-A or JJ-S or Ill-S/ml. 1 "g cscin/"l. 1 "g p,ota"ine/"l. I "g phawilinlml.