Biochemical and Functional Characterization of Protein Kinase CK2 in Ascidian Ciona intestinalis Oocytes at Fertilization CLONING AND SEQUENCE ANALYSIS OF cDNA FOR (cid:1) AND (cid:2) SUBUNITS*

The ubiquitous and pleiotropic dual specificity protein kinase CK2 has been studied and characterized in many organisms, from yeast to mammals. Generally, the enzyme is composed of two catalytic ( (cid:1) and/or (cid:1) (cid:1) ) and two regulatory ( (cid:2) ) subunits, forming a differently as-sembled tetramer. Although prone to controversial in-terpretation, the function of CK2 has been associated with fundamental biological processes such as signal transduction, cell cycle progression, cell growth, apo-ptosis, and transcription. Less known is the role of CK2 during meiosis and the early phase of embryogenesis. In this work, we studied CK2 activity during oocyte activation, a process occurring at the end of oocyte maturation and triggered by fertilization. In ascidian Ciona intestinalis , an organism whose complete genome has been published recently, CK2 was constitutively active in unfertilized and fertilized oocytes. The enzymatic activity oscillated through meiosis showing three major peaks: soon after fertilization (metaphase I exit), before metaphase II, and at the exit from metaphase II. Biochemical analysis of CK2 subunit composition in activated

The ubiquitous and pleiotropic dual specificity protein kinase CK2 has been studied and characterized in many organisms, from yeast to mammals. Generally, the enzyme is composed of two catalytic (␣ and/or ␣) and two regulatory (␤) subunits, forming a differently assembled tetramer. Although prone to controversial interpretation, the function of CK2 has been associated with fundamental biological processes such as signal transduction, cell cycle progression, cell growth, apoptosis, and transcription. Less known is the role of CK2 during meiosis and the early phase of embryogenesis. In this work, we studied CK2 activity during oocyte activation, a process occurring at the end of oocyte maturation and triggered by fertilization. In ascidian Ciona intestinalis, an organism whose complete genome has been published recently, CK2 was constitutively active in unfertilized and fertilized oocytes. The enzymatic activity oscillated through meiosis showing three major peaks: soon after fertilization (metaphase I exit), before metaphase II, and at the exit from metaphase II. Biochemical analysis of CK2 subunit composition in activated oocytes indicated that CK2-␣ was catalytically active as a monomer, independently from its regulatory subunit ␤; however, CK2-␤ was only detectable in unfertilized oocytes where it was associated with a bona fide identified ascidian mitogen-activated protein kinase. After fertilization, CK2-␤ was undetectable, suggesting its rapid degradation. Protein sequence analysis of CK2-␣ and -␤ cDNA indicated a high identity compared with vertebrate homologs. In addition, the absence of putative phosphorylation sites for Cdc2 kinase on both ␣ and ␤ subunits suggested an important role for CK2 in regulating meiotic cell cycle in C. intestinalis oocytes.
CK2 has been associated with the regulation of several biological processes. High levels of CK2 activity have been described in different types of cancer (for review, see Ref. 8), including leukemia in cattle (9) and lymphoma and mammary tumorigenesis in transgenic mice expressing CK2-␣ (10,11). Compelling evidence indicates a role for CK2 in cell growth and, more recently, in cell death. Yeast CK2-␣ subunit is essential for cell viability and is required for progression through the cell division cycle (for review, see Ref. 12). In addition, the role of CK2 in mitosis is supported by two bits of circumstantial evidence: 1) CK2-␣ and -␤ subunits are phosphorylated in a cell cycle-dependent manner (13,14); 2) CK2 phosphorylates important cell cycle regulatory proteins, such as Cdc2/Cdc28 (15)(16)(17). More recently, a novel function of CK2 in regulating apoptotic cell death has emerged (3,4,18,19).
More controversial and poorly understood is the role of CK2 during meiosis. In Xenopus oocyte nuclei, CK2 is probably the major phosphorylating protein, and its activity can be enhanced by polyamines (spermine or spermidine) and inhibited by heparin (20,21). Besides, in the ovary of Xenopus, the mRNA amount for both the CK2-␣ and -␤ subunits is higher than many other mRNAs and increased during oogenesis, in parallel with the increment of enzymatic activity (22). In Rana temporaria during oocyte maturation, CK2 activity increases 7 h after progesterone administration and at the final stage of maturation, suggesting that CK2 participates in the translation control mechanisms during maturation of frog oocytes (23). In contrast to these data, microinjection into Xenopus fullgrown oocytes of highly purified CK2 inhibits meiosis progression induced by progesterone (24). In addition, CK2 becomes activated before germinal vesicle breakdown in maturing echinoderm oocytes (25,26). During oocyte meiotic maturation in Xenopus, Cdc2, a component of maturation-promoting factor (MPF), phosphorylates several factors including CK2, supporting the view that CK2 is involved in the phosphorylation cascade originated by Cdc2 (27). The same authors also reported that Xenopus Cdc2 phosphorylates CK2-␤ in vitro, increasing * 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.
The the activity of the holoenzyme (28). Finally, several CK2 substrates are expressed in oocytes (for review, see Ref. 7).
Strong evidences suggest a role for CK2-␤ independent from the catalytic subunit in meiotic maturation. In unfertilized Xenopus oocytes, CK2-␤ binds and inhibits the activity of Mos (29), a component of cytostatic factor (30). The docking activity of CK2-␤ on Mos negatively regulates oocyte maturation by inhibiting the Mos-dependent activation of MAP kinase (MAPK) (31). The finding of a biological activity of CK2-␤ distinct from its role as a CK2 regulatory subunit is not novel because CK2-␤ might have its own function in binding and recruiting a potential substrate for CK2 holoenzyme (for review, see Ref. 3), or it might sequester partners not necessarily phosphorylated by CK2-␣ (32,33).
In the present work, we investigated the role of CK2 in activated oocytes, a process that takes place in mature oocytes, is dependent on fertilizing stimuli, and leads to meiosis completion (34 -36). We used oocytes from ascidian Ciona intestinalis, an organism studied intensively in developmental biology (37) and, more recently, proposed as a model to study meiotic regulation (38,39). In addition, in 2002 the draft copy of the C. intestinalis genome became publicly available, providing new insights into origin and evolution of chordates (40,41).
Our results indicate a novel role of CK2 after oocyte activation in C. intestinalis. The enzyme activity of CK2-␣ oscillates during meiosis completion after fertilization independently from the activity the ␤ subunit, engaged in forming complexes in unfertilized oocytes. Protein sequence analysis of both CK2-␣ and -␤ leads to the suggestion that the loss of Cdc2 putative phosphorylation sites on both subunits is possibly related to the meiotic arrest in metaphase I observed in C. intestinalis as well as in other invertebrates.

EXPERIMENTAL PROCEDURES
Collection and Fertilization of Oocytes in Vitro-Metaphase I-arrested oocytes were collected from the oviducts of the ascidian C. intestinalis from the Bay of Naples and washed several times in seawater. The chorion and follicle cells were removed manually using sharp steel needles or treatment with 0.1% trypsin as described previously (42). Dechorionated oocytes were kept in 0.1% gelatin-formaldehyde-coated Petri dishes to prevent lysis of nude oocytes (43). Spermatozoa were collected with a fine Pasteur pipette and diluted 1,000-fold in seawater immediately before insemination. Because of the variability of the in vitro fertilization in ascidian oocytes, the population of synchronous fertilized oocytes during meiosis ranged between 70 and 80%.
Immunoblotting, Immunoprecipitation, and SDS-PAGE-For immunoblotting of CK2-␤ and MAPK, unfertilized and fertilized oocytes were lysed in lysis buffer (LB) containing 10 mM Tris-HCl, pH 7.4, 150 mM NaCl, 1 mM EDTA, 1 mM EGTA, pH 8, 0.5% Nonidet P-40, phosphatase inhibitors (20 mM 4-nitrophenyl phosphate, 1 mM sodium vanadate, 40 mM sodium fluoride, 1 mM sodium pyrophosphate), and a mixture of protease inhibitors (Complete by Roche Applied Science), plus 0.2 mM phenylmethylsulfonyl fluoride. Subsequently, cell lysates were sonicated for 20 s, centrifuged at 13,000 rpm for 10 min, and measured for protein concentration (44). Total protein lysates (0.05-0.1 mg) were added with 2ϫ Laemmli loading buffer, heated at 95°C for 5 min, and loaded on a 12% SDS-polyacrylamide gel (45). Subsequently, proteins were blotted onto polyvinylidene difluoride membrane and incubated for 1 h in blocking solution made by 5% nonfat dry milk in 0.1% Tween, 25 mM Tris, pH 8, 137 mM NaCl, 2.69 mM KCl. Primary antibody solutions reported in the text and in figure legends were incubated overnight at 4°C. The membranes were washed and incubated with horseradish peroxidase-linked secondary antibody (Amersham Biosciences) at room temperature for 2 h. Bands were visualized by the enhanced chemiluminescence method (ECL Plus system by Amersham Biosciences).
Whole C. intestinalis cell lysates (0.5 mg of total protein) derived from unfertilized and fertilized oocytes were immunoprecipitated essentially as reported previously (46) employing 2.5 l (1:40 v/v dilution) of an anti-pan-ERK/MAPK monoclonal antibody able to recognize phosphorylated and unphosphorylated forms of MAPK (BD Transduction Laboratories, San Diego). After a 2-h incubation at 4°C with rocking agitation, immunocomplexes were immunoprecipitated by adding 60 l of protein A/G beads (Santa Cruz Biotechnology) and incubation continued for a further 16 h. At the end of the incubation, supernatants were removed, immunocomplexes were washed twice in LB, and LDS sample preparation buffer was added (Invitrogen). Subsequently, immunocomplexes and supernatants were separated on NuPage BisTris 12% precast gel in MES buffer (Invitrogen), transferred to nitrocellulose before immunoblotting that was performed as reported above using as primary antibody Ab278, raised against a peptide spanning from residue Met 140 to Gly 162 of human CK2-␤ (46,47), or anti-ERK/MAPK antibody (New England Biolabs; Beverly, MA).
CK2 and MAPK Assays-CK2 activity was determined on whole cell extracts deriving from unfertilized and fertilized C. intestinalis oocytes. The reaction mixture contained in a total volume of 50 l, 50 -90 M [␥-32 P]ATP (1,500 -3,000 cpm/pmol ATP) (PerkinElmer Life Sciences), 0.5 mM peptide substrate ETE (RRREEETEEE, amino acid sequence one letter code; Promega) in CK2 kinase buffer (50 mM MOPS, pH 7.0, 10 mM MgCl 2 , 10 mM NaCl, 60 mM ␤-glycerophosphate), as reported previously (46). Reactions were incubated for 30 -40 min at 30°C in the presence of 5-10 g of crude extract and terminated by transferring the supernatants to P81 paper (Whatman) to determine phosphate incorporation on ETE peptide as described (48). When [␥-32 P]GTP was used as the phosphate donor instead of [␥-32 P]ATP, the final concentration in the assay was 90 M.
MAPK activity was assayed using a kit commercially available (Amersham Biosciences) based on a peptide derived from epidermal growth factor receptor and containing a highly specific MAPK phosphorylation site. The assays were performed following the manufacturer's instructions with some modifications. Briefly, 5 l of cell lysate (corresponding to 10 g of total proteins) was added to a reaction mixture (30-l final volume) containing 10 l of substrate buffer and 15 l of [␥-32 P]ATP (50 M; 5,000 cpm/pmol). The reaction was incubated for 60 min at 30°C and stopped by adding 10 l of stop reagent. The amount of phosphorylated peptide was determined as described above for the CK2 kinase assay.
Purification of C. intestinalis CK2 and MAPK on Gel Filtration Chromatography-Unfertilized oocytes or samples at different times from fertilization (0, 5, 20, 30 min) were lysed in LB and analyzed on a SMART system (Amersham Biosciences) equipped with a Superose 12 gel filtration column (PC 3.2/30; Amersham Biosciences). The column was equilibrated in LB and calibrated using the following molecular mass markers (Serva): cytochrome c (12.400; 3 g; eluted in fraction 14), ovalbumin (45.400; 4 g; eluted in fraction 10), bovine serum albumin (67.000; 4 g; eluted in fraction 8), and ferritin (450.000; 1 g; eluted in fraction 4). Subsequently, 0.36 mg of cell lysate in 200 l of LB was loaded on Superose 12 column, and proteins were eluted in the same buffer at a flow rate of 50 l/min. Fractions (100 l) were collected and analyzed by SDS-PAGE and immunoblotting to detect the presence of C. intestinalis CK2-␤ and MAPK (CiCK2-␤ and CiMAPK). Aliquots of each fraction (10 l) were used to measure kinase activity as reported above.
Isolation and Sequencing of cDNA Clones-For the cloning of CK2-␣ and -␤, total RNA was isolated from C. intestinalis ovary using the guanidine isothiocyanate/acid-phenol method (49) and used to synthesize the cDNA using avian myeloblastosis virus reverse transcriptase (50). An aliquot of cDNA was amplified in the PCR using the degenerate oligonucleotides reported in Table I, derived from conserved regions of CK2A (coding for CK2-␣) and CK2B (coding for CK2-␤) genes. The PCR incubations were carried out for 36 cycles of denaturation (94°C for 1 min), annealing (46°C for 1 min), and elongation (72°C for 2 min) using in the reaction mixture containing 500 pmol of each oligonucleotide, 200 M dNTP, and a reaction buffer provided with the Taq DNA polymerase (Hoffmann-La Roche). PCR products were isolated from the agarose gel and cloned directly into the pMOS cloning vector (Amersham Biosciences) following the manufacturer's instructions. After transformation, a limited number of positive clones were selected and sequenced to identify the PCR products corresponding to CK2 fragments.
To determine sequences at the 5Ј-and 3Ј-ends of the CiCK2A and CiCK2B mRNAs, rapid amplification of cDNA ends (RACE) assays were carried out using the 3Ј-and 5Ј-RACE system (Invitrogen). For CiCK2B, amplification of the 3Ј-end was carried out using an oligo(dT) containing adapter primer for first strand cDNA synthesis of poly(A) ϩ RNA. The first round of cDNA amplification was performed using 50 pmol each of an abridged universal amplification primer (AUAP) and a gene-specific primer ( Table I). The PCR incubations were carried out for 30 cycles as follows: denaturation, 94°C for 1 min, annealing 57°C for 1 min, and elongation 72°C for 1 min. The prominent band on an agarose gel was subcloned and sequenced.
To amplify the CiCK2B 5Ј-end, 2.5 pmol of the oligonucleotide re-ported in Table I was used for first strand cDNA synthesis of poly(A) ϩ RNA. After tailing, as reported in the kit instructions, the purified single strand cDNA was amplified by the PCR using 10 pmol each of the anchor primer and the nested gene-specific oligonucleotide (Table I).
PCR was performed exactly as described above for the 3Ј-RACE except for the annealing temperature, which was set at 50°C. The PCR product was subcloned and sequenced. To obtain the full-length CiCK2B the two specific oligonucleotides reported in Table I were designed and used to amplify a cDNA synthesized from ovary RNA. The final size of CiCK2B cDNA was 648 bp. The RACE protocol was also applied to isolate CiCK2A. cDNA from the ovary was prepared as described above and used as template in a PCR employing 500 and 380 pmol of sense and antisense degenerate primers, respectively (Table I). The reaction was carried out with 30 cycles of denaturation (1.5 min at 94°C), annealing (2 min at 37°C), and elongation (2 min at 72°C). The amplified fragments were cloned into pMOS PCR vector and sequenced. For CiCK2A, amplification of the 3Ј-end was carried out using an oligo(dT) containing adapter primer for first strand cDNA synthesis of poly(A) ϩ RNA. The first round of cDNA amplification was performed using 50 pmol each of an AUAP and a gene-specific primer (Table I). A second nested gene-specific primer (Table I) was used in a second amplification reaction in conjunction with AUAP to increase the specificity of the procedure. The first PCR incubations were carried out for 30 cycles as follows: denaturation, 94°C for 1 min, annealing 57°C for 1 min, and elongation 72°C for 1 min. In the nested PCR the number of cycles was reduced to 20, and the annealing temperature was set at 52°C. The prominent band on an agarose gel was subcloned and sequenced.
To amplify the CiCK2A 5Ј-end, 2.5 pmol of the oligonucleotide reported in Table I was used for first strand cDNA synthesis of poly(A) ϩ RNA. After tailing, as reported in the kit instructions, the purified single strand cDNA was amplified by the PCR using 10 pmol each of the anchor primer and the nested gene-specific oligonucleotide (GSP2, Table I). To generate enough specific product, the original PCR was diluted (0.1 to 0.5%) and reamplified using the AUAP and a nested primer (GSP3, Table I) located at the 5Ј-end of the GSP2 oligonucleotide. Both PCRs were performed for 30 cycles with an annealing temperature set at 55°C. The PCR product was subcloned and sequenced. To obtain the full-length CiCK2A the two specific oligonucleotides reported in Table I were designed and used to amplify a cDNA synthesized from ovary RNA. The final size of CiCK2A cDNA was 1,158 bp.
cDNA translation was obtained using the PrettySeq program in the EMBOSS software package available on the Web (www.hgmp.mrc. ac.uk/Software/EMBOSS/index.html); sequence alignments were obtained using ClustalW (multiple sequence alignment) and BOXSHADE (Multiple alignments designer) software available at the Web address www.ch.embnet.org.

RESULTS AND DISCUSSION
Biochemical Characterization of CK2 in C. intestinalis Oocytes-CK2 activity was present and easily detectable in a crude extract prepared from dechorionated and unfertilized C. intestinalis oocytes. As reported in Fig. 1A, CK2 activity was linear between 1 and 12 g of crude cell extract in an enzymatic assay performed in the presence of 90 M ATP, similar to the kinase activity measured in a crude extract derived from rapidly proliferating tissues (51) and cell lines (46). To prove that the measured activity was imputable to CK2, we performed the kinase assay in the presence of GTP and quercetin. CK2 is among the few kinases able to utilize both ATP and GTP as phosphate donor (52). As expected, GTP was recognized as substrate by CiCK2, as well as ATP, in the enzymatic assay (Fig. 1B); the apparent K m for GTP was comparable with that measured for ATP (data not shown). In addition, quercetin, a well known CK2 inhibitor (53,54), at a concentration of 10 M, inhibited CiCK2 activity with an IC 50 comparable with that determined for the pure enzyme from different sources (Fig.  1C) (53). It is worthwhile to note that two CK2 inhibitors, one specific such as TBB (55), the other less specific, apigenin (56,57), applied to the assay at the same concentration used for quercetin (10 M), were unable to inhibit CiCK2, suggesting possible structural differences at the level of the catalytic subunit, or the presence of potential interfering factors in oocyte lysates. This observation deserves further studies in the near future.
To analyze the expression of CiCK2 in unfertilized oocytes, we tested several antibodies that were either available commercially or provided by other laboratories. We did not detect any cross-reacting band corresponding to CK2-␣ or ␣Ј subunit; Degenerate primers for CiCK2B CiCK2B CiCK2B RACE 5Ј-end This paper 5Ј-GGCCACGCGTCGACTAGTACGGGIIGGGIIGGGIIG-3Ј Abridged anchor primer Commercially available c CiCK2B full-length This paper 3Ј-primer: 5Ј-TGCTACAAATTCGCCATTGA-3Ј CiCK2B full-length This paper Degenerate primers for CiCK2A CiCK2A

CK2 in Ascidian Ciona intestinalis
on the contrary, an antibody raised against a peptide corresponding to human CK2-␤ (region Met 140 -Gly 162 ; Refs. 46 and 47) recognized a band at a molecular mass of 25 kDa, which we attributed to bona fide CiCK2-␤ (Fig. 1D).
Oscillation of CK2 Activity during Meiosis Completion in C. intestinalis-In an initial attempt to investigate the role of CK2 in oocyte activation and meiosis completion, we measured the enzymatic activity at different times from fertilization. Fig.  2 shows that CK2 activity oscillates during the meiosis completion. Setting an initial 5-min time point, the enzymatic activity increased 5 min after fertilization and was followed by two more peaks of activity at 15 and 25 min. Morphologically, the first peak of CK2 activity preceded the extrusion of the first polar body and corresponded to anaphase I-telophase I (Fig. 2,  top panels), whereas the increase in CK2 activity observed at 25 min from fertilization peaked with the appearance of the second polar body (Fig. 2, top panels), slightly after metaphase II, generally observed around 20 -30 min from fertilization. Finally, the activity detected at 15 min corresponded to the interval elapsing between telophase I and metaphase II.
Changes in MPF kinase activity have been largely studied in ascidian oocytes during meiosis completion (42,58,59). In addition, although partially controversial, these changes in MPF activity possibly correlate with sperm-triggered calcium oscillations (42,58). Two peaks of MPF kinase activity have been described in ascidians, corresponding to metaphase I (unfertilized oocytes) and metaphase II (20 -30 min from fertilization) (42,58). On the other hand, several works reported a mutual interplay between CK2 and Cdc2 kinase (the catalytic subunit of MPF) during mitotic cell division cycle (13-17, 28, 62, 63). Here, we observed two peaks of CK2 activity, at 5 and 15 min from fertilization, when MPF kinase activity is low following the exit from metaphase I (42,58). Similarly, the second peak of MPF activity, observed in metaphase II (42,58), probably precedes the third oscillation of CK2 activity (25 min). Although still preliminary, the data presented suggest that the increase in CiCK2 activity might correlate with a decrease in MPF activity, and vice versa. This observation is not totally novel because a similar oscillation in the activity of both kinases has been reported during HeLa cell division cycle (64). Further work will be devoted to clarify this issue.
Since 1987, a consistent literature has accumulated demonstrating the dramatic activation of CK2 following hormone and growth factor stimulation. Although a large part of these studies has been critically reevaluated and reappraised (65), the present work goes in the direction of the potential regulation of CK2 activity by an external stimulus, the fertilizing spermatozoon, which causes an oscillation of the kinase activity depend- ent on the different phase of the meiotic cell cycle. To our best knowledge, this is the first time that changes in CK2 activity have been clearly demonstrated during meiosis completion after oocyte activation.
Role of CiCK2-␤ Subunit during Meiosis Completion in C. intestinalis-Limited and circumstantial information is available on the role of CK2 during oocyte maturation and fertilization (see the Introduction). Among these, recent data support an important function of CK2-␤ in inhibiting Xenopus oocyte maturation via a physical association between CK2-␤ and Mos kinase with a consequent inhibition of MAPK activation and arrest of progesterone-induced meiotic maturation (29,31). Based on these data, we analyzed the ability of CiCK2-␤ to form complexes with CK2-␣ and/or other partners during meiosis completion.
Analyses have been performed after protein fractionation of oocytes lysed at different times from fertilization on Superose 12, a gel filtration column able to resolve in the range of 1 ϫ 10 3 -3 ϫ 10 5 Da. Surprisingly, we observed that, in unfertilized oocytes, CiCK2-␤ was expressed maximally in fraction 8, peaking at 67 kDa ( Fig. 3B and Table II), a molecular mass very different from 25-28 kDa, representing the average molecular mass of CK2-␤ from different organisms, including C. intestinalis (Fig. 1D). No signals were observed in fractions 12-13 corresponding to 25-30 kDa, suggesting that in unfertilized oocytes, CiCK2-␤ was not present as a monomer but in a complex involving other factors. Although CiCK2-␣ represented the best candidate to bind CiCK2-␤, it was not totally associated with the ␤ subunit because CK2 kinase activity was maximally present in fraction 10, corresponding to 45 kDa (Fig.  3A and Table II), a molecular mass in good agreement with that expected for CiCK2-␣ and/or ␣Ј, based on the homology existing among CK2 catalytic subunits from different species (5). Five min after fertilization, CiCK2-␣ activity remained detectable in fraction 10 corresponding to the putative molecular mass of the catalytic subunit as a monomer, although with an activity 4-fold increased with respect to the analysis performed in un-fertilized oocytes (Fig. 3C), in good agreement with the upregulation of the enzymatic activity observed after 5 min from fertilization in the crude cell lysate (Fig. 2). Immunoblotting of the same fractions probed with Ab278 (anti-human CK2-␤) did not evidenced any positive signal corresponding to CiCK2-␤, either in fraction 8 or in fractions 12-13, suggesting its possible degradation (Fig. 3D). We performed further fractionations of fertilized oocytes (e.g. 20 and 30 min after fertilization) without detecting any significant difference either in the profile of CiCK2 kinase activity or in the immunoblotting performed to verify the presence of CiCK2-␤ in the eluted fractions (data not shown). It is possible that the canonical CiCK2 holoenzyme formed by two catalytic (␣ and/or ␣Ј) and two regulatory (␤) subunits was eluted in fractions 4 -6, as confirmed by the presence of an immunoreactive CK2-␤ band (Fig. 3, B and D, and Table II), and a residual CK2-␣ activity (Fig. 3, A and C, and Table II). Moreover, we cannot exclude the possibility that both CK2-␣ and -␤ were involved in a high molecular mass complex, eluting around 300 kDa, the void volume of the column. In both cases, CK2 detectable in fractions 4 -6 represented only a limited amount of total CiCK2 activity (less than 20%) present in unfertilized oocytes. In addition, we noted that after gel filtration (Fig. 3A), CK2 activity was significantly higher compared with its activation in the total cell extract ( Fig. 2; 5-min time point). This might be explained by suggesting that gel filtration chromatography removed putative inhibitor(s) of CK2 activity.
Because the ability of CiCK2-␤ to interact and inhibit Mos kinase and inactivate MAPK cascade in Xenopus (29, 31), we measured MAPK activity in C. intestinalis oocytes after fertilization. As reported in Fig. 4, the enzymatic activity was high in unfertilized oocytes, peaked at 5 min from the addition of spermatozoa, as reported previously (42), decreased rapidly, and remained low during meiosis completion and extrusion of the second polar body (Fig. 4A). Changes in MAPK activity were also confirmed by the reactivity of an antibody raised against phospho-MAPK and able to recognize the double phos- phorylated form of human MAPK. This antibody cross-reacted against CiMAPK showing a unique band at a molecular mass of 42-44 kDa, whose change of intensity reflected the activation of the enzyme that was maximal at 5 min from fertilization (Fig. 4B), confirming the result of the enzymatic assay (Fig. 4A).
Based on these results, we analyzed the molecular mass of potential complexes involving CiMAPK by means of the gel filtration method reported above. As reported in Fig. 5 and Table II, in unfertilized oocytes, the large amount of CiMAPK activity was present in fraction 8, corresponding to a molecular mass of 67 kDa, significantly higher than that of the CiMAPK monomer (Fig. 4B), suggesting the presence of the kinase as a complex. A residual activity was also present in fraction 4 to indicate the presence of a higher molecular mass complex including CiMAPK. The immunoblotting of the same fractions reported in Fig. 5B confirmed the presence of a highly active enzyme in fraction 8 and a lower activity in fraction 4. In addition, the absence of any significant cross-reacting band in fractions 12-16 seems to indicate that in unfertilized oocytes, the large amount of CiMAPK is present in a bound form to reach a molecular mass in the range of 70 kDa. However, considering the kinase activity detected in fraction 10 (Fig. 5A), we cannot totally exclude that free CiMAPK was present in unfertilized oocytes. If this was the case, it represented only a   4. MAPK activity in Ciona oocytes. Unfertilized and dechorionated oocytes were fertilized, harvested at the indicated time points, and lysed. A, MAPK activity was determined as reported under "Experimental Procedures" employing 15 g of cell lysate as the enzymatic source. Error bars represent the range of triplicate assays, each of which was analyzed twice by scintillation counting. B, 50-g aliquots of cell lysates prepared at the indicated times (numbers on top) from fertilization were separated on SDS-PAGE, transferred to nitrocellulose, and incubated with an antibody raised against the phosphorylated and active form of human ERK/MAPK. Bands in the immunoblot were revealed using a chemoluminescence method. minor amount of the total enzymatic activity detectable. Five min after fertilization, CiMAPK kinase activity increased of about 8-fold, as expected, and, surprisingly, all the enzymatic activity was detected in fraction 10, corresponding to the molecular mass of CiMAPK monomer (Fig. 5C and Table II). The immunoblotting showed in Fig. 5D confirmed the result of the in vitro kinase assay, showing the maximal CiMAPK activation in fraction 10. Further analysis performed at 20 and 30 min from fertilization showed only a residual CiMAPK activity only in fraction 10, with a maximal activity in the peak fraction ranging between 20,000 and 30,000 cpm, about one-fifth to one-third of the activity reported in Fig. 5A, as expected from the result showed in Fig. 4 (data not shown).
Based on these data, we suggest a possible functional interaction between CK2-␤ and MAPK in C. intestinalis oocytes involving a direct physical interaction between these two molecules. In fact, in unfertilized oocytes, both CiCK2-␤ and Ci-MAPK eluted in fraction 8, at a molecular mass of 67 kDa, corresponding, approximately, to the sum of their molecular mass (25 plus 42-44, respectively). Therefore, we immunoprecipitated CiMAPK from unfertilized oocytes and probed the nitrocellulose membrane with Ab278, reacting against human CK2-␤ and CiCK2-␤, as described above (Figs. 1D and 3B). The immunoblotting reported in Fig. 6A (lane 3) reveals the presence of a weak but detectable band corresponding to the molecular mass of CiCK2-␤. The band was absent when the immunoprecipitation was performed on a cell lysate obtained from oocytes within 5 min from fertilization (Fig. 6A, lane 2). No commercially available antibody tested was able to immunoprecipitate CiCK2-␤ (data not shown). The same antibody employed in Fig. 6A immunoprecipitated a weak band recognized on the immunoblot by a different anti-MAPK antibody (Fig. 6B, lane 4). In addition, an enzymatic assay performed on immunoprecipitated CiMAPK (Fig. 6B, lane 4) indicated the presence of low but detectable enzymatic activity, to confirm that the antibody was able to immunoprecipitate CiMAPK (data not shown). Probably the lack of specific antibodies able to immunoprecipitate C. intestinalis CK2-␤ and MAPK allowed us to determine only qualitatively, but not quantitatively, the stoichiometric ratio between CiCK2-␤ and CiMAPK in the immunocomplex. In fact, a large amount of CiCK2-␤ (Fig. 6A, lane  1) and CiMAPK (Fig. 6B, lane 5) polypeptides remained in the supernatants after immunoprecipitation.
The gel filtration approach used in the present paper allowed us to postulate an interaction between CiCK2-␤ and CiMAPK in unfertilized oocytes (summarized in Table II). Two lines of evidence support this conclusion: 1) both proteins eluted in the same fraction, corresponding to a molecular mass of about 67 kDa; and 2) an antibody able to immunoprecipitate CiMAPK also pulled down CiCK2-␤. Because C. intestinalis oocytes represent a relatively new model to study cell cycle regulation, only a few specific reagents are available to study this process. The recent sequencing of the C. intestinalis genome will certainly accelerate the synthesis of specific antibody and reagents necessary to confirm and extend the observations reported here. Although still speculative, we can postulate that the docking activity of CiCK2-␤ maintains CiMAPK in a minimal active state, probably sufficient to ensure the metaphase I block. At fertilization, the cascade reaction triggered by sperm entry, induces inactivation of MPF activity (42,58), degradation of CiCK2-␤, and a transient activation of CiMAPK (Fig. 4). All of these events are completed rapidly within 5-10 min from fertilization, when the oocytes are in telophase I. In metaphase II, when MPF activity increases again (42,58), MAPK activity is no longer required, justifying, perhaps, the absence of CiCK2-␤.
Another major novelty resulting from the present study regards the separate behavior of CiCK2-␣ with respect to its putative, regulatory subunit, ␤. In fact, at different times from fertilization, the catalytic activity of CK2 was always detectable as a monomer, around 45 kDa, in agreement with the putative molecular mass expected for CK2-␣, but very far from that supposed for the ␣ 2 ␤ 2 tetramer (about 130 -140 kDa). In addition, our findings suggest two important biochemical and functional differences of CiCK2-␣ with respect to its vertebrate homologs: 1) a different response to inhibitors (Fig. 1); and 2) the presence of a catalytically activity ␣ subunit always unbound to its putative regulatory partner, ␤ (Fig. 3).
Isolation of CiCK2A and CiCK2B cDNAs and Protein Sequence Analyses-To understand better the function-structure relationship existing in C. intestinalis CK2-␣ and -␤, CiCK2A and CiCK2B cDNA clones were isolated by reverse transcription-PCR from ovarian RNA using degenerate primers (67,68) and RACE protocols. Nucleotide and predicted amino acid sequences of CiCK2B and CiCK2A cDNA are reported in Figs. 7 and 8, respectively. During the course of this study, the cDNA clones for CiCK2B and CiCK2A have been deposited in Gen-Bank under the accession numbers AF360544 and AY092081, respectively. Later, the draft genome of Ciona became available (40,41). Performing a search on the Ciona Genome Web sites (genome.jgi-psf.org/ciona4/ciona4.home.html; ghost.zool.kyotou.ac.jp/indexr1.html) using the Smith-Waterman algorithm (69), we found that protein ci0100132142 perfectly matched with CiCK2-␤ (AF360544; Fig. 9). DNA sequence analysis per-formed by the BLAST 2 Sequences tool (70) showed 100% identity (data not shown). On the contrary, protein ci0100136361, also derived from the Ciona Genome Web site and matching CiCK2-␣ (AY092081), presented the insertion of an arginine between amino acids Gly 354 and Met 355 of CiCK2-␣ (see Fig. 11). Therefore ci0100136361 possesses one extra residue with respect to AY092081. Obviously, this discrepancy was also reflected at the level of the nucleotide sequence with the insertion of triplet G 1060 GT into ci0100136361. We do not have a final explanation for this discrepancy probably because of a mistake in sequencing or contaminations.
Amino acid sequences derived from CiCK2B cDNA were used to perform a PSI-BLAST analysis of all nonredundant Gen-Bank CDS translations/PDB/SwissProt/PIR/PRF data bases at the National Center for Biotechnology Information (71). CiCK2-␤ is very well conserved among vertebrates because it shows 91% identity and 94% similarity among amphibian (Xenopus), mammals (Homo), with a slight decreased homology compared with Drosophila CK2-␤ (identity 83%; positivity 91%). Fig. 9 reports the multiple sequence alignment of FIG. 7. Nucleotide and predicted amino acid sequences of C. intestinalis CK2-␤. The amino acid sequence is shown in single-letter code. The underlined amino acid sequences were used to design the PCR degenerate primers (see Table I). The asterisk denotes the termination of the coding region. The GenBank accession number for CiCK2B is AF360544. C. intestinalis CK2-␤ (CiCK2-␤ and CiCK2-␤-JGI) toward different low vertebrate species such as amphibians (Xenopus) and fish (Danio and Takifugu), insects (Drosophila), and mammals (Homo), according to the phylogeny of urochordata subphylum (ascidians) posed, from an evolutionary point of view, between the origins of the chordates and the origins of the vertebrates (40). CiCK2-␤ maintains all of the main structural features observed in its closest homologs (for review, see Refs. 2, 3, and 5): 1) the autophosphorylation site including S2, S3, and S4; 2) the destruction box originally identified by Allende and Allende (residues Arg 47 -Asp 55 ) (2) which might account for the rapid disappearance of the protein after fertilization (Fig.  3), although it is not clear whether the destruction box acts in concert with other post-translational modification to regulate CK2-␤ stability (72); 3) the acidic region responsible for the polyamine binding loop (Asp 55 -Asp 64 ); 4) the zinc finger region including four cysteine residues (109, 114, 137, 140); 5) the regions involved in ␤-␤ dimerization and ␣-␤ binding (Asp 155 -Leu 167 and Pro 172 -Ala 196 , respectively). It is worthwhile to note that peptide Met 140 -Gly 162 of human CK2-␤, used to synthesize Ab278, is highly conserved in CiCK2-␤, confirming the immu-noreactivity of this antibody against the ascidian protein (Fig. 1D).
Residue Ser 209 deserves a more detailed analysis because it is phosphorylated in vivo and in vitro by Cdc2 kinase (13,62,73). As reported in Figs. 9 and 10, Ser 209 is one of the few amino acids not conserved in Ciona and Drosophila CK2-␤ with respect to other vertebrates. Ser 209 is also missing in yeasts Saccharomyces cerevisiae and Schizosaccharomyces pombe and in plant Aradopsis thaliana, as observed previously by others (5). In contrast with vertebrate oocytes, where maturation ends with metaphase II arrest, mature oocytes from many invertebrates arrest at metaphase I (34). Therefore, we analyzed in more detail the Ser 209 consensus region in all animal species whose CK2-␤ gene has been cloned. We observed that Ser 209 was absent in all invertebrates having a bona fide meiotic block in metaphase of the first meiotic division, such as Insecta (Drosophila, Anopheles, and Spadoptera) and Platyhelminthes (Schistosoma) (34) (Fig. 10). In Caenorhabditis (Nematoda), the phosphorylation site is conserved (Thr 210 ), but the amino-terminal basic residue (generally a lysine) is lost, making Thr 210 an unfavorable site for Cdc2 phosphorylation (60). In summary, protein sequence analysis of CiCK2-␤ evidenced the following points: 1) the consensus site for Cdc2 phosphorylation is perfectly conserved among vertebrates, all having mature oocytes arrested in metaphase II; 2) H. sapiens and C. intestinalis CK2-␤ are highly homologous except for few residues, including Ser 209 ; 3) in all invertebrates, including C. intestinalis, Ser 209 is missing, although the remaining part of the consensus site for Cdc2 phosphorylation is conserved; 4) many invertebrates present a meiotic block in prophase-metaphase I. Taken together, all these data support the hypothesis that Ser 209 might be an important requirement to ensure a metaphase II arrest in vertebrates, or, in general, be involved in regulating cytostatic factor activity in oocytes. This hypothesis, at the moment partially speculative, will be experimentally tested in the near future.
The primary structure of CiCK2-␣ is highly related to its vertebrate homologs (Fig. 11). Running PSI-BLAST (71), the protein showed 84% identity to human CK2-␣, with the more unrelated region localized at the carboxyl-terminal tail of the protein. Apparently, the twin sister of CK2-␣, namely CK2-␣Ј, was not identified in C. intestinalis, neither in this work nor in the Ciona Genome Project. CK2 belongs to the socalled CMGC group of protein kinases which includes Cdc2/ Cdks, MAPKs, and GSK-3 (61). All of the main structural motifs characterizing vertebrate CK2-␣ (2, 5) are conserved in CiCK2-␣. These include: 1) the glycine loop responsible for the phosphate anchor (residues Gly 44 -Val 51 ); 2) the basic stretch responsible for substrate binding and probably involved in ␤ subunit binding (residues Lys 61 -Ile 67 ); 3) the catalytic loop (residues His 153 -Asp 164 ); 4) the activation loop that, unlike other CMGC kinases, does not require phosphorylation to express a maximal enzymatic activity (4) (residue Pro 183 -Pro 199 ); 5) the minor and major insert (residues Gly 234 -Tyr 238 and Pro 266 -Leu 291 , respectively), a hallmark of all CMGC kinases (61).
In mammalian dividing cells, CK2-␣ is phosphorylated at four sites (Thr 344 , Thr 360 , Ser 362 , and Ser 370 ; numbers refer to the amino acid position in the human protein) that result in phosphorylation in vitro by Cdc2 (14,63). In addition, a putative autophosphorylation site has been recently identified on CK2-␣ corresponding to Tyr 182 and located within the activation subdomain (66). In CiCK2-␣ the putative Thr 182 autophosphorylation site is conserved (Thr 181 in Ciona), whereas three of four Cdc2 consensus sites are lost (Fig. 8). Residue Thr 360 in human CK2-␣ is mutated to isoleucine (Ile 360 in Ciona), whereas residues Thr 344 and Ser 362 are conserved, but the adjacent carboxyl-terminal proline, a determinant essential for Cdc2 phosphorylation (60), replaced by a glutamine and a threonine, respectively. The only consensus site for Cdk1 phosphorylation perfectly conserved in CiCK2-␣ is Ser 370 . It is intriguing to note that all four sites are lost in CK2-␣ from invertebrates, including sea urchin, the most closely related to ascidians in the phylogenic tree (Fig. 11).
All the data presented in the present work suggest a new and unexplored function of CiCK2 during meiosis. To our best knowledge, this is the first time that CK2 function has been investigated after oocyte activation because previous and controversial studies were limited to the maturation process. This new CK2 function increases the broad range of biological processes where CK2 is, in some way, involved. The publication of Ciona genome will certainly accelerate studies on early phase of embryogenesis of this organism, and will probably clarify the role of CK2 in meiosis completion.