Meiotic Abnormalities of c-mos Knockout Mouse Oocytes: Activation after First Meiosis or Entrance into Third Meiotic Metaphase

Transcription

1 BIOLOGY OF REPRODUCTION 55, (1996) Meiotic Abnormalities of c-mos Knockout Mouse Oocytes: Activation after First Meiosis or Entrance into Third Meiotic Metaphase Kazushi Araki, 3 Kunihiko Naito, 2, 3 Seiki Haraguchi, 3 Rika Suzuki, 4 Minesuke Yokoyama, 4 Maki Inoue, 3 Shinichi Aizawa, 5 Yutaka Toyoda, 6 and Eimei Sato 3 Department of Reproductive and Developmental Biology,3 Institute of Medical Science, The University of Tokyo, Minato-ku, Tokyo 108, Japan Mitsubishi Kasei Institute of Life Sciences, 4 Machida-shi, Tokyo 194, Japan Laboratory of Morphogenesis,5 Institute of Molecular Embryology and Genetics, Kumamoto University School of Medicine, Kumamoto 860, Japan Research Center for Protozoan Molecular Immunology, 6 Obihiro University of Agriculture and Veterinary Medicine, Obihiro, Hokkaido 080, Japan ABSTRACT In Xenopus oocytes, Mos activates the mitogen-activated protein kinase (MAPK) signal transduction cascade and regulates meiosis. In mammalian oocytes, however, the functions of Mos are still unclear. In the present study, we used c-mos knockout mouse oocytes and examined the roles of Mos in mouse oocyte maturation and fertilization, including whether Mos controls MAPK and maturation promoting factor (MPF) activity. The kinetics of germinal vesicle breakdown (GVBD) and the first polar body emission were similar in wild-type, heterozygous mutant, and homozygous mutant mice. Activities of MPF were also not significantly different among the three genotypes until the first polar body emission. In contrast, MAPK activity in c-mos knockout oocytes did not significantly fluctuate throughout maturation, and the oocytes had abnormal diffused spindles and loosely condensed chromosomes, although a clear increase in MAPK activities was observed after GVBD in wild-type and heterozygous mutant oocytes that had normal spindles and chromosomes. After the first polar body emission, 38% of c-mos knockout oocytes formed a pronucleus instead of undergoing second meiosis, indicating the crucial role of Mos in MPF reactivation after first meiosis. When oocytes that reached second metaphase were fertilized or stimulated by ethanol, many c-mos knockout oocytes emitted a second polar body and progressed into third meiotic metaphase instead of interphase, although all fertilized or activated oocytes in the heterozygote progressed to interphase, indicating that Mos deletion leads to compensatory factors that might not be degraded after fertilization or parthenogenetic activation. These results suggest that Mos is located upstream of MAPK in mouse oocytes as in Xenopus oocytes but is independent of MPF activity, and that Mos/MAPK is not necessary for GVBD and first polar body emission. Our results also suggest that Mos plays a crucial role in normal spindle and chromosome morphology and the reactivation of MPF after first meiosis. INTRODUCTION In Xenopus oocytes, the c-mos proto-oncogene product (Mos) is required for activation of the maturation promoting factor (MPF) in G2 arrested oocytes [1, 2], for reactivation of MPF after metaphase I, for transition to meta- Accepted July 31, Received May 21, 'Financial support: Grant-in-aid for scientific research ( ) from the Ministry of Education, Science, Sports and Culture of Japan. 2Correspondence: Kunihiko Naito, Department of Reproductive and Developmental Biology, Institute of Medical Science, The University of Tokyo, Shirokanedai, Minato-ku, Tokyo 108, Japan. FAX: (+81) phase II without DNA replication [3, 4], and for maintaining high MPF activity in metaphase II oocytes [5]. Mos has a serine-threonine kinase domain [6]. In Xenopus oocytes, Mos activates mitogen-activated protein kinase (MAPK) kinase by direct phosphorylation, and subsequently activates MAPK [7-10]. It has been shown that the above-mentioned Mos functions in Xenopus oocytes are mediated by MAPK activity [11, 12]. In contrast, the function of Mos in mouse oocytes is unclear, although Mos has been thought to play important roles in oocyte maturation because of the exclusive transcription in oocytes [13-15]. When anti-mos antibody or antisense oligonucleotide was microinjected into immature mouse oocytes, various results were reported, such as the inhibition of germinal vesicle breakdown (GVBD) [16], the normal induction of GVBD but the inhibition of first polar body emission [17, 18], and the normal induction of first polar body emission but entrance into interphase instead of second meiosis [19,20]. Recently c-mos knockout mice were generated by homologous recombination in embryonic stem cells [21, 22]. These mutant mice have truncated Mos that has no kinase activity. Oocytes obtained from the c-mos knockout mice have been reported to undergo GVBD and mature normally with a frequent spontaneous parthenogenic activation, indicating that that Mos is not essential for oocyte maturation in the mouse [21, 22]. Recently, Mos has been shown to be required for MAPK activation and to be involved in microtubule organization during meiotic maturation in the mouse [23]. In the present study, we used c-mos knockout mouse oocytes and examined the roles of Mos in mouse oocyte maturation and fertilization in more detail, including whether Mos controls MAPK and MPF activity. Our data indicate that Mos is located upstream of MAPK in mouse oocytes as in Xenopus oocytes, but is independent of MPF activity, and that Mos/MAPK is not necessary for GVBD and first polar body emission but is required for normal spindle and chromosome morphology. Our results also suggest that Mos plays a crucial role in the reactivation of MPF after first meiosis, and that Mos deletion produces compensatory factors that might not degrade after fertilization and therefore prevent the oocytes from escaping meiosis. MATERIALS AND METHODS Animals The c-mos-deficient mice were produced by homologous recombination in TT2 embryonic stem cells [21]. Male and

2 1316 ARAKI ET AL. female heterozygous mice were mated; their pups were analyzed for their genotypes by polymerase chain reaction (PCR) using whole blood without any purification according to the method of Mercier et al. [24], with several modifications. Briefly, 0.5 I of whole blood was added to a 50-R1I final volume of PCR buffer (10 mm Tris-HCI [ph 8.3], 50 mm KCI, 1.5 mm MgCl 2, 0.01% gelatin, 0.2 mm dntp) containing the following primers: MOS102 (TCTGTGGCCCCGTTATG-CTTAGGAAAAGCA), MOS105 (TGGGATTGAAGGCAGCCATCTTCAGCCA- TG), and MOS107 (TCGTGCTTTACGGTATCGCC- GCTCCCGATT); this mixture was then incubated three times according to the following sequence: for 10 min at 94 C and 3 min at 52 C. After this incubation cycle, 2.5 U of Taq DNA polymerase (Boehringer Mannheim Biochemica, Mannheim, Germany; 250 U/50 xi) was added. The PCR was cycled 45 times with a programmable thermal controller (PTC-100, MJ Research Inc.). Each cycle consisted of 45 sec at 94 C, 45 sec at 52 C, and 150 sec at 72 C. After this incubation, 15 Rxl of the PCR mixture was electrophoresed in 0.8% agarose gel (Agarose ME, Iwai Chem. Pharm. Inc., Tokyo, Japan). The gel was stained in ethidium bromide (0.5 Rpg/ml) for 30 min. Mice with a 1.3- kilobase (kb) band, a 0.87-kb band, and both bands were considered as wild type, homozygous mutant, and heterozygous mutant, respectively. In Vitro Maturation of Oocytes Three- to eight-week-old mice were used for the experiments. Equine CG (5 IU) was injected i.p. 48 h before the ovaries were removed. The large antral follicles were placed in maturation medium consisting of Waymouth's MB752/1 medium (Gibco Labs., Grand Island, NY) supplemented with 1 IU/ml ecg, 0.23 mm sodium pyruvate, 50 mg/l streptomycin sulfate, 75 mg/l penicillin G (potassium salt), 1 mg/ml undialized fetuin (Deutsch method; Gibco), and 0.2 mm 3-isobutyl-l-methylxanthine (IBMX, Sigma Chemical Company, St. Louis, MO) [25] and punctured; the oocytes completely enclosed by cumulus cells were incubated in IBMX-free maturation medium for 0-16 h at 37 C under 5% CO 2 in air. Homozygous oocytes matured for 16 h were classified as activated oocytes or metaphase oocytes by Hoechst staining as described below. In Vitro Fertilization and Parthenogenetic Activation of Oocytes Adult cycling mice received injections of 5 IU ecg 48 h before injection of 5 IU hcg. Oocytes matured in vivo were obtained h after administration of hcg. Groups of complexes were then placed in 0.2 ml fertilization medium/modified Krebs Ringer bicarbonate solution (TYH medium) [26] containing 4 mg/ml BSA. For some experiments, cumulus cells and the zona pellucida were removed with 0.1% hyaluronidase (Type IV-S; Sigma, St. Louis, MO) in PBS and with acidic Tyrode's solution, ph 2.5 [27], respectively. The first polar body was also removed by gentle pipetting before insemination. Spermatozoa collected from the caudal epididymides, preincubated for h at 37 C under 5% CO 2 in air, were used for insemination at concentrations of 60 cells/pll. For parthenogenetic activation, oocytes were treated with 8% ethanol in TYH medium for 6.5 min and cultured in Whitten's medium [28] at 37 C in 5% CO 2. Examination of Oocytes Oocytes were mounted on a glass slide and fixed with 0.5% glutaraldehyde in PBS (ph 7.4) and then with 10% neutral formalin in PBS overnight. The fixed oocytes were stained with 0.25% acetolacmoid and examined under a Nomarski interference microscope (Axiophoto, Zeiss, Basel, Switzerland). Some oocytes were stained with 2.5 mg/ml Hoechst (Calbiochem-Novabiochem, San Diego, CA) for 10 min at 37 C in a CO 2 incubator. The oocytes were examined under an epifluorescence microscope. Immunocytochemical Staining for Alpha-Tubulin Morphology of oocytes matured for 8 h was confirmed as first metaphase under a fluorescent microscope using Hoechst-33342, and the zona pellucida was removed. Zonae were fixed for 15 min with 3.7% paraformaldehyde in PBS and with 2% Triton X-100 (Boehringer Mannheim) in PBS for 30 min. A rat monoclonal antibody (YL 1/2; BIS 077b, BIOSYS, Compiegne, France), specific for tyrosinated alpha-tubulin [29], and a rhodamine-labeled anti-rat antibody were used for immunostaining. Samples were observed with a Bio-Rad (Richmond, CA) MRC-600 Confocal Laser Scanning Microscope. Protein Kinase Assays Histone H1 kinase and myelin basic protein (MBP) kinase activity was assayed according to the method of Pelech et al. [30], with several modifications [31]. Briefly, oocyte lysates were prepared by freezing oocytes (10 oocytes/tube) in 2 I kinase assay buffer (15 mm EGTA, 60 1 mm sodium -glycerophosphate, 30 mm p-nitrophenylphosphate, 25 mm 3-[n-morpholino]propanesulfonic acid [MOPS], 15 mm MgCl 2, 0.2 mm Na 3 VO 4, 2 ug/ml leupeptin, 2,ug/ml aprotinin, 1 Rpg/ml pepstatin, 1 mm PMSF, 50 pim para-aminobenzoic acid [PABA], and 0.5 RM protein kinase A [PKA] inhibitor [TTYADFIASGRTGR- RNAIHD; Sigma]. Lysates were stored at -70 C until use. Each assay tube contained 10 oocytes, 1% Nonidet P40, 20 mm [y- 3 2 P]ATP (Amersham International plc; Amersham, Bucks., UK), 1 mg/ml histone H1 (type III-S, Sigma) or 0.4 mg/ml MBP (Sigma) in a final volume of 12.5,ul. Assay tubes were incubated for 20 min at 30 0 C. Assays were terminated by the addition of 0.4 ml 20% trichloroacetic acid solution (TCA) and 0.1 ml 1% BSA as a carrier protein for precipitation. The assay suspensions were centrifuged at x g for 5 min. The precipitates were washed once with 0.4 ml 20% TCA and dissolved in 0.2 ml 1 N NaOH. The radioactivity was counted, after the addition of 0.5 ml scintillation fluid (ACS II; Amersham), in a liquid scintillation counter (LSC-1000; Aloka, Tokyo, Japan). The value of blank tubes containing all materials except the cytosol preparation was subtracted from each value. Histone H1 kinase activity and MBP kinase activity were assayed at least three times. Immunoblotting for MAPK MAPK was detected by immunoblotting with the anti- MAPK antibody (Zymed Laboratory Inc., South San Francisco, CA). Immunoblotting was performed as described previously [32]. Oocytes were collected at the germinal vesicle (GV) stage (0 h after in vitro culture)and at metaphase I (8 h after in vitro culture). Seventy-five oocytes collected in 5 ul1 medium were transferred into 15 p. 1 of extraction buffer, and immediately 20 R.1 of double-strength

3 MEIOTIC ABNORMALITY OF c-mos KNOCKOUT MOUSE OOCYTES 1317 (A) 100' Laemmli sample buffer was added. After SDS-PAGE, proteins were transferred to a nitrocellulose membrane and probed with the antibody by means of a blotting detection kit in which the streptavidin-alkaline phosphatase conjugate was used as the signal-generating sy'stem (Amersham). Statistical Analysis Chi-square analysis and Student' s t-test were used for statistical evaluation of the results. IDifferences at a probato be statistically sig- bility of p < 0.05 were considered nificant. (B) RESULTS Progression of Oocyte Maturation 80' Almost all oocytes from homozygous (86.1%), heterozygous (91.4%), and wild-type (87.4%) females underwent 0 GVBD within 4 h after the initiation of in vitro maturation 40' (Fig. 1A). No conspicuous differences were found in the first polar body emission kinetics among the three genotypes, and the rates reached more than 80% at 14 h after 20' m 0* the initiation of maturation in all genotypes (Fig. B). Clonr ennlnenatlnn ana arrantrempnt nf rhrnmncenme nn culture period (hours) cuu8 l0 1erod the equatorial plane at the first and the second metaphase culture period(hours) were observed in the wild-type (Fig. 2) and heterozygous FIG. 1. Time course of GVBD (A) and first polar body emission (B) of (data not shown) oocytes. In homozygous mutant oocytes, cultured homozygote (circles), heterozygote (triangles), and wild-type the morphology at the GV stage was identical to that in mouse (squares) oocytes. Oocytes were isolalted in a medium containing heterozygous mutant and wild-type oocytes (Fig. 3, A and 0.2 mm IBMX and transferred to IBMX-free medium. Results were ob- D), but many oocytes had abnormal metaphase chromosomes such as loose condensation and disorder on the equa- tained from at least three independent experi torial plane (Fig. 3, C, F, H, and K), although a few had relatively normal ones (Fig. 3, B and E). Moreover, homozygous oocytes had a large first polar body that con- tained metaphase-like chromosomes (Fig. 3, G and J) or a nucleus-like structure. Morphologically, the alpha-tubulin in homozygous mutant oocytes showed signs of diffuseness and weakness of detection strength (Fig. 4B) in comparison with that of wild-type oocytes (Fig. 4A). Entrance into Interphase from First Metaphase Caused by Loss of Mos The maturation stages of oocytes cultured for h in vitro are shown in Table 1. In heterozygous mutant and wild- FIG. 2. Morphology of oocyte during in vitro maturation in wild-type mice. A-C) Oocytes were mounted on glass slides and stained with 0.25% acetolacmoid after fixation. D-F) Oocytes were stained with Hoechst (2.5 g/ml) for 10 min without fixation. A, D) germinal vesicle stage; B, E) metaphase I stage; C, F) metaphase II stage. gv, germinal vesicle; ch, chromosomes at metaphase stage; pb, first polar body. x400.

4 1318 ARAKI ET AL. FIG. 3. Morphology of oocytes during in vitro maturation in homozygous mutants. A-C, G-) Oocytes were mounted on glass slides and stained with 0.25% acetolacmoid after fixation. D-F, -L) Oocytes were stained with Hoechst (2.5 Ilg/ml) for 10 min without fixation. A, D) germinal vesicle stage; (B, C, E, F) metaphase I stage; (GC, H,, K) metaphase II stage; (I, L) activated oocyte (type I). gv, germinal vesicle; ch, chromosomes at metaphase stage; pb, first polar body; pn, pronucleus. x400. type oocytes, 96% and 99% of oocytes, respectively, were at the second metaphase. In homozygous mutant oocytes, however, only 54% of oocytes were at the second metaphase, and 38% of oocytes formed a pronucleus with a polar body or 2 pronuclei without a polar body, indicating that the oocyte had been activated directly from the first metaphase (type I activated oocyte). These oocytes were clearly distinguishable from the oocytes activated from the second metaphase, which had 2 polar bodies and a pronucleus or 1 polar body and 2 pronuclei (type II activated oocyte).

5 MEIOTIC ABNORMALITY OF c-mos KNOCKOUT MOUSE OOCYTES 1319 TABLE 1. Maturation stages of oocytes cultured for h in vitro. Genotype Number of oocytes (%)a Total Type I oocytes Anaphase I activated examined Telophase I oocytesb Type II activated Metaphase 11 oocytes c -/ (2) 202 (39) 289 (56) 13 (2) +/ (4) 0 (0) 124 (96) 0 (0) +/ (1) 0 (0) 85 (99) 0 (0) a Numbers in parentheses indicate percentage of oocytes examined. b Activated oocytes from metaphase I. c Activated oocytes from metaphase II. FIG. 4. Microtubules of the oocyte stained with the anti-a-tubulin antibody (YL1/2) at metaphase I viewed by confocal microscopy. A) Wildtype mouse, (B) c-mos knockout mouse. Bars = 35 Rm. Oocytes at metaphase I were harvested at 8-9 h after initiation of in vitro maturation culture and fixed by paraformaldehyde in PBS. Arrowheads indicate the liberated structure of microtubules. MAPK Activity during Oocyte Maturation The fluctuations in MAPK activity in the three genotype oocytes is shown in Figure 5A. MAPK activities in immature oocytes were at the basal level in all genotypes. The activity of wild-type and heterozygous mutant oocytes was significantly increased at 4 h of culture, when almost all oocytes underwent GVBD. During 4-16 h of culture, the activity of wild-type oocytes was maintained at about 5 to 7 times the initial activity and was highest among the three genotypes. In heterozygous mutant oocytes, the activity was maintained at about 4 times the initial activity after GVBD and in the middle range between that of wild-type and ho- FIG. 5. A) Activities of MBP kinase in the oocytes of c-mos-deficient (circles), heterozygous (open triangles), and wild-type (squares) mice. Oocytes were isolated in medium containing 0.2 mm IBMX and transferred to IBMX-free medium. Solid inverted triangles represent the activity of activated oocytes of homozygous mice. Kinase activity is expressed as the fold increase over the initial level detected in oocytes before in vitro maturation culture. Experiments were repeated at least three times. *Significantly lower than wild-type values; a, significantly higher than the initial value. B) Detection of 42- and 44-kDa MAPKs by immunoblotting. Oocytes were collected at 0 and 8 h after in vitro maturation culture from wild-type (+/+), heterozygous (+/-), and homozygous (-/-) mutant mice.

6 1320 ARAKI ET AL. 19 _ IL j a r 2- a u I I I I culture period (hour) FIG. 6. Activities of histone H1 kinase in c-mos-deficient (circles), heterozygote (open triangles), and wild-type (squares) mouse oocytes during meiotic maturation in vitro. Oocytes were isolated in medium containing 0.2 mm IBMX and transferred to IBMX-free medium. A solid inverted triangle indicates the activity of activated oocytes of homozygous mice. Kinase activity is expressed as the fold increase over the initial level detected in oocytes before in vitro maturation culture. Experiments were repeated at least three times. *Significantly lower than the second metaphase values. All values except for that indicated by a closed inverted triangle were significantly higher than the initial value. mozygous mutant oocytes. In contrast, MAPK activity in homozygous mutant oocytes did not show significant fluctuation throughout maturation and was significantly lower than that of wild-type oocytes after 4 h of culture. Immunoblotting detected two bands at 42 and 44 kda in all genotypes at both 0 and 8 h of culture (Fig. 5B). Both bands in wild-type and heterozygous mutant oocytes shifted upwards at 8 h of culture, when almost all oocytes were at the first metaphase and had high MAPK activity. In c-mos knockout mouse oocytes, however, no band shift was detected at 8 h of culture (Fig. 5B). Histone H1 Kinase Activity during Oocyte Maturation The fluctuation in histone H1 kinase activity in the three genotype oocytes is shown in Figure 6. The activity in GV oocytes was low in all three genotypes and significantly increased after GVBD at 4 h after the start of culture. The activity further increased in all groups at 8 h of culture to about 8 to 10 times the initial activity. Then the activity decreased transiently at the first polar body emission. There were no significant differences in histone H1 kinase activity among the three genotype oocytes until this time. At 16 h, the activity of the wild-type and heterozygous mutant oocytes, and of the homozygous mutant oocytes that arrested at the second metaphase, increased again. On the other hand, the activity of type I activated oocytes in c-mos knockout mice further decreased to the basal level and became significantly lower than that of the second metaphase oocytes. Third Metaphase Oocytes Observed in c-mos Knockout Mice The typical morphology of fertilized oocytes in heterozygous and homozygous mice at 6-7 h after insemination is shown in Figure 7. The fertilized oocytes in heterozygous mice had two pronuclei and a penetrated sperm tail (Fig. 7A). There were no differences between heterozygous mice and wild-type mice in the morphology of fertilized oocytes (data not shown). In the homozygous mice, 69% of fertil- FIG. 7. Morphological profile of fertilized eggs. Fertilized eggs at 6-7 h after insemination were fixed by 2.5% glutaraldehyde in PBS. A) Fertilized egg of the heterozygote with a second polar body, two pronuclei, and penetrated sperm tail. B, C) Fertilized egg of the homozygote. B) Egg with polar bodies, two pronuclei, and penetrated sperm tail. C) Egg with two polar bodies, chromosome at metaphase, and penetrated sperm tail. Arrowheads indicate penetrated sperm tails. pb, polar body; ch, chromosome; pn, pronucleus. x400. ized oocytes had a pronucleus or pronuclei and a penetrated sperm tail (Fig. 7B), and 31% of fertilized oocytes had a metaphase-like chromosome structure in spite of having a penetrated sperm tail (Fig. 7C, Table 2). To determine whether these metaphase-like chromosomes were in the second metaphase or not, we removed the zona pellucida and first polar body, and then inseminated the oocytes and observed the emission of the second polar body as in Ku-

7 MEIOTIC ABNORMALITY OF c-mos KNOCKOUT MOUSE OOCYTES 1321 TABLE 2. Morphology of fertilized oocytes at 6-7 h after insemination. Total oocytes Number of oocytes (%)d Genotype examined PN + STb CH + STc -/ (69) 24 (31) +/ (100) 0 (0) a Numbers in parentheses indicate percentage of total oocytes examined. b Pronucleus and fertilized sperm tail. c Chromosome and fertilized sperm tail. biak [33]. As shown in Table 3, in homozygous mice, only 47% of fertilized oocytes formed a pronucleus or pronuclei, and 53% of fertilized oocytes had a second polar body and metaphase-like chromosomes with a penetrated sperm tail, indicating the so-called metaphase III (Fig. 8), although 100% of fertilized heterozygous oocytes formed pronuclei. In metaphase III oocytes, sperm nuclei did not form pronuclei, but were transformed into clumps of highly condensed chromatin (Fig. 8). To verify that chromosome structures in the metaphase III oocytes did not originate from sperm nuclei, experiments on parthenogenetic activation by means of ethanol stimulation were conducted. As shown in Table 4, in homozygous mice, 60% of the oocytes examined were activated, and 25% of them formed one polar body and metaphase III chromosomes (Fig. 9, A and B), although not all activated heterozygous oocytes forming a pronucleus and metaphase III oocytes were observed. DISCUSSION In the present study, it was clear that Mos is not essential for GVBD and first polar body emission in murine oocytes, since the loss of Mos did not affect rates and time course of these processes. The present data indicate that the mechanism initiating oocyte maturation in the mouse differs from that in Xenopus, where Mos synthesis is essential for undergoing GVBD [1, 2]. The high rate of first polar body emission in homozygous oocytes contradicted previous reports, which demonstrated that in mouse oocytes injected with c-mos antisense oligonucleotides [17] or anti c-mos antibody [18], first polar body emission was prevented. Although the reason for this is unknown, the antibodies and oligonucleotides used, or the microinjection process, might have had some deleterious effects on the oocytes. In sharp contrast to the lack of effect of Mos on the progression of first meiosis in the mouse, the abnormalities of chromosomes and alpha-tubulin morphologies in the metaphases of homozygous mutant oocytes indicate that Mos participates in chromosome condensation and microtubule reorganization. Zhao et al. [16, 18] have reported that about 90% of oocytes that received an antibody to Mos did not assemble a meiotic spindle. Mos has been implicated in the reorganization of the microtubules, which leads TABLE 3. Morphology of fertilized oocytes without a zona pellucida and a first polar body. Total oocytes Number of oocytes (%) Genotype examined PN + STb Metaphase IIIc -/ (47) 10 (53) +/ (100) 0 (0) anumbers in parentheses indicate percentage of oocytes examined. b Pronucleus and fertilized sperm tail. c Chromosomes with a second polar body and sperm tail. FIG. 8. Fertilized egg at 6 h after insemination obtained from c-mos knockout mouse. The zona pellucida and first polar body were removed before insemination. Photographs show the second polar body (A) and the cytoplasm (B). Note chromosome at metaphase III stage (arrow) stained with Hoechst (C). x500. TABLE 4. Morphology of ethanol-activated oocytes without a zona pellucida and a first polar body. Number of oocytes Total oocytes Metaphase III Genotype examined Activated (%)a PN stage (%)b (%)c -/ (60) 18 (75) 6 (25) +/ (84) 164 (100) 0 (0) a Number in parentheses indicates percentage of oocytes examined. b Pronucleus stage; number in parentheses indicates percentage of activated oocytes. c Chromosomes with a second polar body; number in parentheses indicates percentage of activated oocytes.

8 1322 ARAKI ET AL. FIG. 9. Metaphase III oocyte obtained from c-mos knockout mouse. Zona pellucida and first polar body were removed, and then the oocytes were treated with ethanol (8%, 6.5 min). Three hours after treatment with ethanol, the oocytes were stained with Hoechst and photographs were taken under a Nomarski differential interference microscope (A) and fluorescent microscope (B). x400. to formation of the spindle and the spindle pole [34, 35]. Mos overexpression in somatic cells induced meiotic-like alterations in the mitotic spindle [36]. Abnormalities in the organization of the microtubules and chromatin were recently reported in c-mos knockout mouse oocytes [23]. The morphological abnormalities induced by the loss of Mos in the present study are consistent with these previous reports and confirm that Mos plays an important role in the reorganization of microtubules and chromosome condensation. In Xenopus oocytes, Mos activates the MAPK cascade through the specific activation of MAPK kinase [7-10]. In the present study, MAPK activity was assayed throughout mouse oocyte maturation in the three genotypes. In heterozygous mutant and wild-type oocytes, the fluctuation patterns were in close agreement with those in previous reports [37, 38]. On the other hand, MAPK activity of homozygous mutant oocytes did not significantly fluctuate throughout maturation and was clearly lower than that of wild-type oocytes. It has been shown by SDS-PAGE that mouse MAPK was present as 42- and 44-kDa bands, and the migration rate was decreased when MAPK was activated by phosphorylation [23, 37, 39]. The two bands at 42 and 44 kda were detected in all three genotypes in the present study, and a band shift was also observed at 8 h of maturation when MAPK activity was high in heterozygous mutant and wild-type oocytes. In homozygous mutant oocytes, however, no decrease in the migration rate was detected at 8 h of culture when the oocytes were at the first metaphase. Recently, the same result has been reported in a different strain of c-mos knockout mouse oocytes [23]. These results indicate that Mos physiologically stimulates MAPK during maturation of murine oocytes as in Xenopus oocytes. MAPK activation is a prerequisite for GVBD in Xenopus oocytes [12]. The present results suggest, however, that in murine oocytes, MAPK activation is not essential for GVBD and first polar body emission. In mouse oocytes, MAPK is localized in microtubule-organizing centers [37, 38]. Verlhac et al. [38] reported that microtubule and chromatin behavior was controlled by MAPK activity during meiosis in mouse oocytes. We therefore considered that the morphological abnormalities in homozygous mutant oocytes referred to above contributed to this low MAPK activity caused by the loss of Mos. In the present study, MPF activity was low in the G2 arresting oocytes and high in the first and the second metaphase oocytes, with a transient decrease at first polar body emission. This pattern closely agrees with previous reports [32, 40]. There was almost no difference among the three genotype oocytes until first polar body emission, indicating that Mos does not stimulate MPF activity directly and that MPF and MAPK activities are regulated independently. These data suggest that the normal fluctuation in MPF activity can cause the normal process of oocyte maturation in spite of the loss of Mos and MAPK activity, confirming the importance of MPF in oocyte maturation in the mouse. One of the most drastic abnormalities in c-mos knockout mouse oocytes was their entrance into the interphase instead of second meiosis after first polar body emission. In these oocytes, MPF was inactivated to the basal level instead of being reactivated after the decrease in polar body emission. Furuno et al. [4] have reported that suppression of DNA replication during meiotic divisions in Xenopus oocytes is accomplished by the Mos-mediated premature reactivation of cdc2 kinase. Oocytes injected with antisense c-mos oligonucleotides completed the first meiotic division but failed to initiate second meiosis and reformed a nucleus [19, 20]. The present results are consistent with these reports, and suggest also that in the mouse Mos plays a crucial role in the reactivation of MPF after the first polar body emission. In the present study, however, 56% of oocytes in the homozygous mutant were not activated after first meiosis and reached the second metaphase in spite of the loss of Mos. This indicates that c-mos deletion induces some compensatory factors that reactivate MPF after first meiosis, although there are wide variations in compensatory efficiency of the oocytes. Verlhac et al. [23] reported that their c-mos knockout oocytes did not require Mos for MPF reactivation after the first meiosis, prompting us to think of the high compensatory activity in their mice. Details of these factors are still unknown, but the present results indicate that a compensatory action other than the MAPK cascade may be at work, since MAPK in homozygous mutant oocytes was maintained in an inactive form throughout maturation. When matured c-mos knockout oocytes were activated by fertilization or ethanol stimulation, some oocytes were transformed into metaphase III instead of interphase. The phenomenon of metaphase III has already been reported by Kubiak [33]. Kubiak demonstrated that metaphase III frequently appeared at h after hcg injection, due to ethanol stimulation and fertilization, but our experimental conditions were different from his, because all heterozygous oocytes entered the interphase after activation. Under

9 MEIOTIC ABNORMALITY OF c-mos KNOCKOUT MOUSE OOCYTES 1323 normal conditions in Xenopus oocytes, the stimulus of fertilization destroys cyclin B by the ubiquitin pathway [41] and Mos by calmodulin-dependent protein kinase II [42] or the N-terminal proline-dependent ubiquitin pathway [43], so that the MPF activity is decreased and the cell cycle progresses into the interphase. Under the conditions of our study, however, the MPF activity was maintained not by Mos but presumably by compensatory factors. We therefore present here the following hypothesis to account for metaphase III in c-mos knockout mice. When the metaphase II arrested oocytes were penetrated by sperm, degradation of cyclin B and a transient decrease in MPF activity occurred and the cell cycle progressed into the anaphase-telophase. In these oocytes, however, since unknown compensatory factors other than Mos may not be degraded even if sperm penetrate the oocyte, the decreased MPF activity may be restored and the cell cycle then progresses into metaphase III. Verlhac et al. [23] also reported the appearance of metaphase HI in their c-mos knockout mice, confirming the presence of Mos-compensatory factors that were not degraded after metaphase II in the c-mos knockout oocytes. In summary, the results of our present study indicate that Mos is located upstream of MAPK in mouse oocytes as well as in Xenopus oocytes, but independent of MPF activity, and that Mos/MAPK is not essential for GVBD and first polar body emission, but is essential for maintaining condensation of chromosomes and a normal spindle by increasing MAPK activity. In addition, one of the most important roles of Mos may be to maintain high MPF activity after the first polar body emission to enable oocytes to progress from first meiosis to second meiosis and to correctly degrade after fertilization in order to deactivate MPE Finally, unknown compensatory factors other than Mos may maintain the MPF activity in the c-mos knockout mouse and prevent the escape from meiosis. ACKNOWLEDGMENTS We thank Dr. Naohiro Hashimoto for the generous supply of c-mos knockout mice; Dr. Hiroshi Imahie for his valuable discussion; and Ms. Takako Saito, Dr. Takashi Tsuji, and Mr. Shin-ichi Kamijo (Mitsubishi Kasei Institute of Life Sciences) for their assistance in this work. REFERENCES 1. Sagata N, Oskarsson M, Copeland T, Brumbaugh J, Vande Woude GE Function of c-mos proto-oncogene product in meiotic maturation in Xenopus oocytes. Nature 1988; 335: Yew N, Mellini ML, Vande Woude GE Meiotic initiation by the mos protein in Xenopus. Nature 1922; 355: Kanki JP, Donoghue DJ. Progression from meiosis I to meiosis II in Xenopus oocytes requires de novo translation of the mosxe protooncogene. Proc Natl Acad Sci USA 1991; 88: Furuno N, Nishizawa M, Okazaki K, Tanaka H, Iwashita J, Nakajo N, Ogawa Y, Sagata N. Suppression of DNA replication via Mos function during meiotic divisions in Xenopus oocytes. EMBO J 1994; 13: Sagata N, Watanebe N, Vande Woude GF, Ikawa Y. The c-mos protooncogene product is a cytostatic factor responsible for meiotic arrest in vertebrate eggs. Nature 1989; 342: Hanks SK, Quinn AM, Hunter T. The protein kinase family: conserved features and deduced phylogeny of the catalytic domains. Science 1988; 241: Posada J, Nelson Y, Ahn NG, Vande Woude GF, Cooper JA. Mos stimulates MAP kinase in Xenopus oocytes and activates a MAP kinase in vitro. Mol Cell Biol 1993; 13: Nebreda AR, Hunt T. The c-mos proto-oncogene protein kinase turns on and maintains the activity of MAP kinase, but not MPF, in cellfree extracts of Xenopus oocytes and eggs. EMBO J 1993; 12: Nebrada AR, Hill C, Gomez N, Cohen P, Hunt T. The protein kinase mos activates MAP kinase in vitro and stimulates the MAP kinase pathway in mammalian somatic cells in vivo. FEBS Lett 1993; 333: Shibuya EK, Ruderman JV. Mos induces the in vitro activation of mitogen-activated protein kinases in lysates of frog oocytes and mammalian somatic cells. Mol Biol Cell 1993; 4: Haccard O, Sarcevic B, Lewellyn A, Hartley R, Roy L, Izumi T, Eleanor E, Mallar JL. Induction of metaphase arrest in cleaving Xenopus embryos by MAP kinase. Science 1993; 262: Kosako H, Gotoh Y, Nishida E. Requirement for the MAP kinase kinase/map kinase cascade in Xenopus oocyte maturation. EMBO J 1994; 13: Propst F, Vande Woude GE Expression of c-mos proto-oncogene transcripts in mouse tissues. Nature 1985; 315: Goldman DS, Kiessling AA, Millette CF, Cooper GM. Expression of c-mos RNA in germ cells of male and female mice. Proc Natl Acad Sci USA 1987; 84: Mutter GL, Wolgemuth DJ. Distinct developmental patterns of c-mos protooncogene expression in female and male mouse germ cells. Proc Natl Acad Sci USA 1987; 84: Zhao X, Batten B, Singh B, Arlinghaus RB. Requirement of the c-mos protein kinase for murine meiotic maturation. Oncogene 1990; 5: Paules RS, Buccione R, Moschel RC, Vande Woude GE Eppig JJ. Mouse Mos protooncogene product is present and functions during oogenesis. Proc Natl Acad Sci USA 1989; 86: Zhao X, Singh B, Batten BE. The role of c-mos proto-oncoprotein in mammalian meiotic maturation. Oncogene 1991; 6: O'keefe SJ, Wolfes H, Kiessling AA, Cooper GM. Microinjection of antisense c-mos oligonucleotides prevents meiosis II in the maturing mouse egg. Proc Natl Acad Sci USA 1989; 86: O'Keefe SJ, Kiessling AA, Cooper GM. The c-mos gene product is required for cyclin B accumulation during meiosis of mouse eggs. Proc Natl Acad Sci USA 1991; 88: Hashimoto N, Watanabe N, Furuta Y, Tamemoto H, Sagata N, Yokoyama M, Okazaki K, Nagayoshi M, Takeda N, Ikawa Y, Aizawa S. Parthenogenetic activation of oocytes in c-mos-deficient mice. Nature 1994; 370: Colledge WH, Carlton MBL, Udy GB, Evans MJ. Disruption of c-mos causes parthenogenetic development of unfertilized mouse eggs. Nature 1994; 370: Verlhac M-H, Kubiak JZ, Weber M, Geraud G, Colledge WH, Evans MJ, Maro B. Mos is required for MAP kinase activation and is involved in microtubule organization during meiotic maturation in the mouse. Development 1996; 122: Mercier B, Gaucher C, Feugeas O, Mazurier C. Direct PCR from whole blood, without DNA extraction. Nucleic Acids Res 1990; 18: Eppig JJ, Telfer EE. Isolation and culture of oocytes. Methods Enzymol 1993; 255: Toyoda Y, Yokoyama M, Hoshi T. Studies on the fertilization of mouse eggs in vitro. I. In vitro fertilization of eggs by fresh epididymal sperm. Jpn J Anim Reprod 1971; 16: Nicolson GL, Yanagimachi R, Yanagimachi H. Ultrastructural localization of lectin binding sites on the zonae pellucidae and plasma membranes of mammalian eggs. J Cell Biol 1975; 66: Whitten WK. Nutrient requirements for culture of preimplantation embryos in vitro. Adv Biosci 1971; 6: Kilmartin JV, Wright B, Milstein C. Rat monoclonal antitubulin antibodies derived by using a new non secreting rat cell line. J Cell Biol 1982; 93: Pelech SL, Meijer L, Krebs EG. Characterization of maturation-activated histone Hl and ribosomal S6 kinases in Sea Star oocytes. Biochemistry 1987; 26: Naito K, Toyoda Y. Function of histone HI kinase activity during meiotic maturation in porcine oocytes. J Reprod Fertil 1991; 93: Choi T, Aoki F Mori M, Yamashita M, Nagahama Y, Kohmoto K. Activation of p34cdc2 protein kinase activity in meiotic and mitotic cell cycles in mouse oocytes and embryos. Development 1991; 113: Kubiak JZ. Mouse oocytes gradually develop the capacity for activation during the metaphase II arrest. Dev Biol 1989; 136: Zhou R, Oskarsson M, Paules RS, Schulz N, Cleveland D, Vande Woude GE Ability of the c-mos product to associate with and phosphorylate tubulin. Science 1991; 251: Fukasawa K, Murakami MS, Blair DG, Kuriyama R, Hunt T, Fis-

10 1324 ARAKI ET AL. chinger P, Vande Woude GE Similarities between somatic cells overexpressing the mos oncogene and oocytes during meiotic interphase. Cell Growth Differ 1994; 5: Fukasawa K, Vande Woude GE Mos overexpression in Swiss 3T3 cells induces meiotic-like alterations of the mitotic spindle. Proc Natl Acad Sci USA 1995; 92: Sobajima T, Aoki F, Kohmoto K. Activation of mitogen-activated protein kinase during meiotic maturation in mouse oocytes. J Reprod Fertil 1993; 97: Verlhac MH, Kubiak JZ, Clarke HJ, Maro B. Microtubule and chromatin behavior follow MAP kinase activity but not MPF activity during meiosis in mouse oocytes. Development 1994; 120: Verlhac MH, Pennart HD, Maro B, Cobb MH, Clarke HJ. MAP kinase becomes stably activated at metaphase and is associated with microtubule-organizing centers during meiotic maturation of mouse oocytes. Dev Biol 1993; 158: Hashimoto N, Kishimoto T. Regulation of meiotic metaphase by a cytoplasmic maturation-promoting factor during mouse oocyte maturation. Dev Biol 1988; 126: Glotzer M, Murray AW, Kirschner MW. Cyclin is degraded by the ubiquitin pathway. Nature 1991; 349: Lorca T, Cruzalegui FH, Fesquet D, Cavadore JC, Mery J, Means A, Doree M. Calmodulin-dependent protein kinase II mediates inactivation of MPF and CSF upon fertilization of Xenopus eggs. Nature 1993; 366: Nishizawa M, Furuno N, Okazaki K, Tanaka H, Ogawa Y, Sagata N. Degradation of Mos by the N-terminal proline (Pro 2 )-dependent ubiquitin pathway on fertilization of Xenopus eggs: possible significance of natural selection for Pro 2 in Mos. EMBO J 1993; 12: