Mitogen-activated Protein Kinase Kinase Is Required for the Mos-induced Metaphase Arrest*

The product of the c-mos proto-oncogene functions not only as an initiator of oocyte maturation but also as a component of cytostatic factor that causes the natural arrest of the unfertilized egg at the second meiotic metaphase. It has been shown that Mos can phosphorylate and activate mitogen-activated protein (MAP) kinase kinase (MAPKK) in vitro, leading to activation of MAP ki- nase. In this study, by using an anti-MAPKK antibody that can specifically inhibit Xenopus MAPKK activity, we have shown that MAPKK mediates the cytostatic factor activity of Mos. Coinjection of this anti-MAPKK an- tibody with the bacterially expressed Mos protein into a two-cell embryo prevented the Mos-induced cleavage ar- rest as well as the Mos-induced MAP kinase activation. The analysis of individual embryos indicated that the degree of the cleavage arrest was correlated with the extent of the MAP kinase activation in the Mos- and the Moslantibody-iqjected embryos. These observations suggest the involvement of a signal transmission path- way consisting of Mos, MAPKK, and MAP kinase in the metaphase arrest.

The product of the c-mos proto-oncogene, a 39-kDa germ-cellspecific serinelthreonine protein kinase, is an active component of CSF (7). The Mos protein is absent from immature oocytes, is synthesized from maternal mRNA in response to progesterone, and then is degraded after fertilization (8,9). Translation of Mos is necessary for progesterone-induced oocyte maturation (lo), and the Mos protein alone can initiate maturation without any hormonal stimulation, as shown by injecting Mos mRNAor bacterially-expressed Mos protein into oocytes (8,11). The Mos mRNA or the Mos protein injected into one blastomere of a two-cell embryo mimics CSF and arrests cleavage of the injected half of the embryo at metaphase (7,11). Moreover, the CSF activity in the unfertilized egg cytoplasm is neutralized with anti-Mos antibody (7). Thus, Mos is thought to play crucial roles in both the release from prophase arrest in immature oocytes and the induction of meiotic metaphase arrest in the unfertilized egg.
Mitogen-activated protein ( M A P ) kinase is also activated during Xenopus oocyte maturation (12)(13)(14). This serine/ threonine protein kinase is highly conserved throughout evolution and is activated commonly by various extracellular stimuli (15)(16)(17)(18). Activation of MAP kinase requires phosphorylation on both tyrosine and threonine residues (19,20). A 45-kDa protein that can induce phosphorylation and activation of kinase in vitro was purified from Xenopus unfertilized eggs (21) and from mammalian somatic cells (22)(23)(24)(25). This MAP kinase activating factor (for review, see Ref. 26) can undergo autophosphorylation on serine, threonine, and tyrosine residues (23,271 and phosphorylate the kinase-deficient mutant of MAP kinase on the regulatory tyrosine and threonine residues (22,24,(28)(29)(30). Therefore, this factor is a dual specificity protein kinase and has been called MAP kinase kinase ( MAP=, also called MEK). cDNA cloning and sequencing of MAPKK (28,3 1-34) revealed that vertebrate MAPKK shows high similarities to several yeast protein kinases functioning in various signal transduction pathways, suggesting that the "AP kinase cascade functions universally in eukaryotes (35, 36). (21,37), and activation of MAPKK in cells is accompanied by its phosphorylation on serine and threonine residues (27,38). Therefore, MAPKK itself is thought to be activated by phosphorylation catalyzed by an upstream serinelthreonine protein kinase, M A P = kinase (39, 40). Several MAPKK kinases have been reported, including , MEKK (441,and Mos (45,46). Bacterially expressed Mos protein can activate MAP kinase when injected into immature oocytes or added to cell-free extracts ofXenopus oocytes (4648) and can directly phosphorylate and activate MAPKK in vitro (45,46).

M A P M is inactivated by protein phosphatase 2A treatment in vitro
Mos, MAPKK, and MAP kinase are activated, like MPF, during Xenopus oocyte maturation and are inactivated shortly after fertilization (8,9,(12)(13)(14)21). MPF is reactivated in subsequent mitotic cell cycles, although neither Mos nor MAP kinase seems to be reactivated markedly during early embryogenesis. These suggest that the MAPKWMAP kinase cascade may mediate some of the actions of Mos. Recently, we produced an anti-Xenopus MAPKK antibody that can specifically inhibit Xenopus MAPKK activity in vitro (49). Microinjection of this neutralizing antibody against MAPKK into immature oocytes prevented progesterone-or Mos-induced activation of MAP kinase (49). Moreover, progesterone-or Mos-induced activation of MPF was also blocked, as evidenced by inhibition of both germinal vesicle breakdown and histone H1 kinase activation (49).
This suggested that MAPKK plays a significant role in mediating the maturation-initiating function of Mos. In this report, we have examined whether MAPKK mediates another function of Mos, a function as a component of CSF. Injection of bacterially expressed Mos protein into a two-cell embryo induced MAP kinase activation and cleavage arrest, while coinjection of the neutralizing antibody against MAPKK with the Mos protein prevented the cleavage arrest as well as the MAP kinase activation. These results indicate that MAPKK is required for the Mos-induced metaphase arrest.

MATERIALS AND METHODS
Reagents-The neutralizing antibody against Xenopus MAPKK was prepared by injecting mice with bacterially expressed glutathione Stransferase-MAPKK as described previously (49). Recombinant MAPKK was prepared by cleaving glutathione S-transferase-MAPKK with Factor Xa (28) and by removing glutathione S-transferase with a glutathione-agarose column. Recombinant mulE-mos protein (Xenopus c-mos fused downstream of the maltose-binding protein of Escherichia coli) was expressed in and purified from E. coli as described previously (11,47,49).
Preparation of Cell-free Extructs-Concentrated cell-free extracts were prepared from Xenopus immature oocytes essentially as described by Shibuya et ul. (51). Dissected ovaries were treated with 2 mg/ml of collagenase in MBS (10 m HEPES pH 7.5,88 m M NaCl, 1 mM KC], 2.4 m NaHCO,, 0.3 mM Ca(NO,),, 0.41 m M CaCl,, 0.82 m MgSO,) for 2 h a t 18 "C and then extensively washed with MBS. Stage VI oocytes were sorted by hand, washed twice in EB (20 m M HEPES pH 7.2, 0.25 M sucrose, 0.1 M NaCI, 2.5 m M MgCI,) and transferred to a 1.5-ml tube containing EB with aprotinin, pepstatin, chymostatin, and leupeptin (each 10 pg/ml) and with cytochalasin B (50 &mU. Excess EB was removed, and oocytes were crushed by centrifugation a t 15,000 x g for 15 min at 2 "C. The supernatant between the lipid cap and packed yolk was collected and centrifuged again. Aliquots were frozen in liquid nitrogen and stored a t -80 "C. Microinjections-Ovulated eggs were fertilized in vitro with fresh minced testis and dejellied with 2% cysteine (pH 7.8). Dejellied embryos were washed several times with 0.3 x MBS and then cultured in 0.3 x MBS containing Ficoll 400 (Pharmacia Biotech Inc.) at 22 "C. 90-105 min after fertilization, 50 nl of samples were microinjected into one blastomere of the two-cell embryo using an IM-1 microinjection apparatus (Narishige, Tokyo). Embryos were cultured for 3 h under the same conditions as above and scored for the extent of cleavage arrest. For assays of MAP kinase, groups of 10 embryos or individual embryos were homogenized in 200 or 50 pl of XB (20 m M Tris-CI (pH 7.5), 60 m P-glycerophosphate, 10 mM MgCI,, 10 m M EGTA, 2 m dithiothreitol, 1 mM Na,VO,, 1 m M phenylmethylsulfonyl fluoride, 20 pg/ml aprotinin), respectively. Homogenates were clarified by centrifugation at 15,000 x g for 15 min at 2 "C.

RESULTS
The Neutralizing Antibody against MAPKK Prevents Mosinduced Activation of MAP Kinase in Cell-free Extracts of Xenopus Oocytes-We prepared a neutralizing antibody that can specifically inhibit the activity ofXenopus MAPKK as described previously (49). Immunoblot analysis revealed that this anti-MAPKK antibody (mouse polyclonal antibody) reacted specifically with the 45-kDa MAPKK band in total extracts of embryos at the blastula stage ( Fig. 1, lanes 2 and 5). This antibody reacted strongly with bacterially expressed recombinant MAPKK (Fig. 1, lanes 3 and 6) but not at all with bacterially expressed Mos (malE-mos protein, a fusion protein between the E. coli maltose-binding protein and the Xenopus c-mos protein kinase; Fig. 1, lanes 4 and 7). The recombinant MAPKK contains 11 extra amino acids (GIPGNSALTPN) on the authentic N-terminus of MAPKK (28).
To examine whether this anti-MAPKK antibody prevents Mos-induced activation of MAP kinase in a cell-free system, we prepared concentrated cell-free extracts from immature oocytes (51). ARer incubation with control mouse IgG or the same amount of the anti-MAPKK antibody at 0 "C, malE-nos was added to the extracts, and the incubation was performed at 22 "C. The malE-mos induced full activation of MAP kinase within 1 h in the control IgG-incubated extracts, as judged by the shift in electrophoretic mobility of the 42-kDa polypeptide recognized by anti-MAP kinase antibody (Fig. 2, lanes 5-8). In contrast, the activation of MAP kinase was markedly diminished in the anti-MAPKK antibody-incubated extracts (Fig. 2,  lanes 9-12). When this anti-MAPKK antibody had been preincubated with the recombinant MAPKK, MAP kinase activation induced by malE-mos was restored (Fig. 2, lanes 13-16). Thus, the neutralizing antibody against MAPKK blocked the malEmos-induced activation of MAP kinase by inhibiting the MAPKK activity in a cell-free system.
Mos-induced Metaphase Arrest Is Inhibited in the Embryos Injected with the Neutralizing Antibody against MAPKK-It has been shown that microinjection of malE-mos into one blastomere of a two-cell embryo results in cleavage arrest at mitotic metaphase (11) and in rapid activation of MAP kinase (46). To examine the effect of the neutralizing antibody against MAPKK on the Mos-induced metaphase arrest, we microinjected malE-mos with the control IgG or with the anti-MAPKK antibody into one of the blastomeres of two-cell embryos. Microinjection of mulE-nos with the control IgG resulted in cleavage arrest at the injected half of embryos and caused hemiblas- tulation (Fig. a). There often existed embryos in which part or whole of the uninjected half also ceased cleavage probably because of the leak of the injected samples (Fig. 3, B and C). In contrast, microinjection of malE-mos with the anti-MAPKK antibody induced no cleavage arrest (Fig. 30) or the arrest in the limited area at the injected half (Fig. 3, E andF). The result is represented by the histogram (Fig. 41, in which results obtained from four independent experiments are summarized. The result indicated that the Mos-induced metaphase arrest could be canceled by the neutralizing antibody against MAPKK (Figs. 3 and 4).
Microinjection of malE-mos with the control IgG induced marked activation of MAP kinase, as evidenced by the appearance of the electrophoretically retarded MAP kinase band on SDS-polyacrylamide gel electrophoresis (Fig. 5, lane 3, upper) and the corresponding anti-phosphotyrosine-positive band (Fig. 5, lane 3, lower). In contrast, microinjection of malE-mos with the anti-MAPKK antibody induced almost no appearance of the electrophoretically retarded, tyrosine-phosphorylated form of MAP kinase that represents the activated form of the kinase (Fig. 5, lane 4, upper and lower). Thus, the neutralizing antibody against MAPKK inhibited the Mos-induced metaphase arrest and the Mos-induced MAP kinase activation.
To confirm that the effect of the anti-MAPKK antibody was specifically due to the inhibition of MAPKK activity, we microinjected malE-mos and the anti-MAPKK antibody together with or without the recombinant MAPKK into one of the blastomeres of two-cell embryos. The embryos injected with malEmos, the anti-MAPKK antibody, and the recombinant MAPKK showed cleavage arrest (Fig. 6 A , lower) and MAP kinase activation (Fig. 6B, lane 4 ) to greater extent than did the embryos injected with malE-mos and the anti-MAPKK antibody ( Fig.   6 A , upper and Fig. 6B, lane 3). Thus, the effect of the neutralizing antibody against MAPKK was antagonized by the addition of the recombinant MAPKK. The result that the extent of cleavage arrest in Fig. 6A was less than that in Fig. 4 is due to a lesser amount of injected malE-nos in the experiment shown in Fig. 6.
We then examined in more detail the correlation between the cleavage arrest and the MAP kinase activation by analyzing embryos individually. We arbitrarily selected 16 embryos that had been injected with malE-mos and the control IgG or with malE-mos and the anti-MAPKK antibody. After observing the extent of cleavage arrest of each embryo, that embryo was lysed, and the lysate was analyzed by immunoblotting with anti-MAP kinase antibody. Nearly full activation of MAP kinase had occurred in the embryos of which whole area showed cleavage arrest (Fig. 7, lanes 1 and 2). In the embryos of which mal€-mos + IgG mal€-mos + aMAPKK  (0.34 pg). Embryos in which one of the blastomeres (the right side of the embryos) was injected at the two-cell stage were cultured for 3 h at 22 "C before being fixed in 1% glutaraldehyde. over the half area showed cleavage arrest, 2 0 4 0 % of total MAP kinase molecules were activated even in the embryos injected with malE-mos and the anti-MAPKK antibody (Fig. 7,  Frc. 4. Effect of the anti-MAPKK neutralizing antibody on the maZE-mos-induced cleavage arrest. One blastomere of a two-cell embryo was injected with 50 nl of either malE-mos (3.4 ng) plus control IgG (0.34 pg) or malE-mos (3.4 ng) plus anti-MAPKK antibody (0.34 pg). Injected embryos were cultured for 3 h a t 22 "C and scored for the extent of cleavage arrest, which was classified into four types. Type I includes embryos showing no cleavage arrest (typically Fig. 30). Type I1 includes embryos in which below a quarter area shows cleavage arrest (typically Fig. 3, E and F ) . Type I11 includes embryos in which over the half but not the whole area shows cleavage arrest (typically Fig. 3, A and B ) . Type IV includes embryos in which the whole area shows cleavage arrest (typically Fig. 3C). Results obtained from four independent experiments are summarized as the histogram shown here. In each experiment, injected embryos were derived from eggs laid by one frog.

3-9).
In contrast, the activation of MAP kinase was undetectable in the embryos of which none or below a quarter of the area showed cleavage arrest (Fig. 7, lanes 10-16). Thus, the extent of the cleavage arrest was correlated with that of the MAP kinase activation.    lanes 1-6) or with malE-mos and anti-MAPKK antibody (lanes 7-16) was scored for the extent of cleavage arrest and classified as in Fig. 4 (Z-ZV). Then, each embryo was lysed individually and analyzed by immunoblotting with anti-" kinase antibody.
reported that inhibition of activation of MAP kinase by the neutralizing antibody against MAPKK blocked Mos-induced oocyte maturation (49). In this paper, we have shown that this anti-MAPKK antibody, when introduced into one blastomere of a two-cell embryo, antagonizes the ability of Mos to induce metaphase arrest in cleaving embryos. Thus, MAPKK is thought to play a crucial role in both initiating oocyte maturation and inducing metaphase arrest downstream of Mos. Although MAPKK may have other substrates than MAP kinase and one of these might be the crucial target or functionally redundant with MAP kinase, it is hypothesized that MAP kinase, the only identified substrate for MAPKK, is necessary for inducing metaphase arrest. This, however, remains to be tested directly in the future work. The mechanism by which the same protein kinase participates in these apparently different cellular events is unclear, but MAP kinase may directly or indirectly regulate a factor(s) involved in activation and/or stabilization of MPF.
Recently, Maller and co-workers (52) reported that microinjection of thiophosphorylated MAP kinase into one blastomere of a two-cell embryo induced cleavage arrest similar to that induced by Mos (52), suggesting that active MAP kinase in the unfertilized egg is sufficient for inducing metaphase I1 arrest. Thus, the MAPKWMAP kinase cascade may mediate the CSF activity of Mos. The disappearance of CSF activity upon fertilization may be due to the degradation of Mos and inactivation of W K K and MAP kinase. Interestingly, in clam oocytes that are not arrested at metaphase 11, activation of MAP kinase is not sustained and inactivated prior to germinal vesicle breakdown (53).
The previous report that Ras has the CSF activity like Mos (54) can be explained by Ras-induced activation of the " K K / MAP kinase cascade (51, 55-57). Ras is supposed to induce metaphase arrest independently of Mos probably through Raf-1, another MAPKK kinase (41-43). The MAPKWMAP kinase cascade may function at a convergent point in various signal transduction pathways resulting in metaphase arrest.