Phosphorylation of the p34cdc2 target site on goldfish germinal vesicle lamin B3 before oocyte maturation

ARTICLE IN PRESS

European Journal of Cell Biology 85 (2006) 501–517

0171-9335/$ - se

doi:10.1016/j.ej

�CorrespondE-mail addr

www.elsevier.de/ejcb

Phosphorylation of the p34cdc2

target site on goldfish germinal vesicle

lamin B3 before oocyte maturation

Akihiko Yamaguchia,�, Yoshinao Katsub, Michiya Matsuyamaa,Michiyasu Yoshikunic, Yoshitaka Nagahamac,d

aLaboratory of Marine Biology, Department of Animal and Marine Bioresource Sciences, Faculty of Agriculture,

Kyushu University, Hakozaki 6-10-1, Higashi-Ku, Fukuoka 812-8581, JapanbNational Institute for Basic Biology, Department of Bioenvironmental Research Center for Integrative Bioscience, Okazaki, JapancLaboratory of Reproductive Biology, National Institute for Basic Biology, Department of Developmental Biology, Okazaki, JapandCREST, Japan Science and Technology Corporation, Kawaguchi, Saitama, Japan

Received 24 November 2005; received in revised form 7 February 2006; accepted 8 February 2006

Abstract

The nuclear membranes surrounding fish and frog oocyte germinal vesicles (GVs) are supported by the lamina, aninternal, mesh-like structure that consists of the protein lamin B3. The mechanisms by which lamin B3 is transportedinto GVs and is assembled to form the nuclear lamina are not well understood. In this study, we developed aheterogeneous microinjection system in which wild-type or mutated goldfish GV lamin B3 (GFLB3) was expressed inEscherichia coli, biotinylated, and microinjected into Xenopus oocytes. The localization of the biotinylated GFLB3 wasvisualized by fluorescence confocal microscopy. The results of these experiments indicated that the N-terminal domainplays important roles in both nuclear transport and assembly of lamin B3 to form the nuclear lamina. The N-terminaldomain includes a major consensus phosphoacceptor site for the p34cdc2 kinase at amino acid residue Ser-28. Toinvestigate nuclear lamin phosphorylation, we generated a monoclonal antibody (C7B8D) against Ser-28-phosphorylated GFLB3. Two-dimensional (2-D) electrophoresis of GV protein revealed two major spots of laminB3 with different isoelectric points (5.9 and 6.1). The C7B8D antibody recognized the pI-5.9 spot but not the pI-6.1spot. The former spot disappeared when the native lamina was incubated with lambda phage protein phosphatase(l-PP), indicating that a portion of the lamin protein was already phosphorylated in the goldfish GV-stage oocytes.GFLB3 that had been microinjected into Xenopus oocytes was also phosphorylated in Xenopus GV lamina, as judgedby Western blotting with C7B8D. Thus, lamin phosphorylation appears to occur prior to oocyte maturation in vivo inboth these species. Taken together, our results suggest that the balance between phosphorylation by interphase laminkinases and dephosphorylation by phosphatases regulates the conformational changes in the lamin B3 N-terminalhead domain that in turn regulates the continual in vivo rearrangement and remodeling of the oocyte lamina.r 2006 Elsevier GmbH. All rights reserved.

Keywords: Germinal vesicle; Intermediate filament; Nuclear matrix; Protein kinase; Protein phosphatase

e front matter r 2006 Elsevier GmbH. All rights reserved.

cb.2006.02.002

ing author. Tel./fax: +81 92 642 2888.

ess: [email protected] (A. Yamaguchi).

www.elsevier.de/ejcb

ARTICLE IN PRESSA. Yamaguchi et al. / European Journal of Cell Biology 85 (2006) 501–517502

Introduction

The nucleus is a complex organelle in which a numberof important functions, such as DNA replication,transcription, and RNA processing, are carried out in atemporally and spatially defined manner (Cox and Laskey,1991; Ma et al., 1998; Mills et al., 1989; Misteli andSpector, 1998; Xing and Lawrence, 1991). The mechan-isms by which cells achieve spatial ordering of their nuclearfunctions are unclear, however. Although recent observa-tions suggest that the nuclear architecture plays animportant role in establishing spatial organization in thenucleus, the molecules involved in this process have notbeen identified, with the exception of the nuclear lamins(Baskin, 1995; Berezney and Wei, 1998; Berezney et al.,1995; Hoffman, 1993; Nickerson et al., 1995; Spector,1993; Van Driel et al., 1995). The lamins are the majorconstituents of the nuclear lamina (Aebi et al., 1986;Belmont et al., 1993; Gerace and Burke, 1988), a mesh-likestructure that lies next to the inner nuclear membrane, andthey are also present in the inner nuclear matrix (Bridgeret al., 1993; Hozak et al., 1995; Moir et al., 1994).

Sequencing of lamin cDNA clones from differentspecies has revealed that nuclear lamins are members ofthe intermediate filament (IF) protein family (e.g.,McKeon et al., 1986). IFs are present in all animalcytoplasms and nuclei, and they have a self-polymerizingfunction (for reviews see Fuchs and Weber, 1994; Parryand Steinert, 1995; Stuurman et al., 1998). Site-specificphosphorylation and dephosphorylation of IF proteinsalters the filament structure and organization (for a review,see Inagaki et al., 1996). Lamins also exhibit dynamicbehavior by assembling and disassembling during the cellcycle, as is the case with other cytoplasmic IF proteins (forreviews, see Nigg, 1992; Skalli and Goldman, 1991).

Lamins contain several conserved phosphoacceptor sitesthat lie within the regions flanking either end of an a-helical-rod domain. Studies have revealed that specific phosphor-ylation of lamins at conserved Ser or Thr residues occursonly in M phase, and that it triggers lamina disassembly invitro (Eggert et al., 1991; Peter et al., 1990; Ward andKirschner, 1990). Furthermore, mutations within thephosphoacceptor sites interfere with M-phase-specificlamina disassembly in vivo (Heald and McKeon, 1990).The cdc2 kinase is a prominent candidate for the M-phase-specific kinase, since either purified p34cdc2 kinase ormaturation-promoting factor can induce lamin disassemblyin isolated nuclei (Dessev et al., 1991; Peter et al., 1990),nuclear ghosts (Molloy and Little, 1992), and bacteriallyproduced, assembled lamins (Heitlinger et al., 1991; Peteret al., 1991; Ward and Kirschner, 1990).

In addition to the cdc2 kinase phosphoacceptor sites,other M-phase-specific phosphorylation sites that regulatelamin assembly/disassembly have been identified inseveral species. For example, phosphorylation of a Serthat lies close to the cdc2 site inhibits head-to-tail

polymerization of lamin, as does phosphorylation of thecdc2 kinase phosphoacceptor site itself (Stuurman, 1997),and M-phase disassembly of the nuclear lamina can beinduced by protein kinase C (PKC)-mediated phosphor-ylation (Collas, 1999; Collas et al., 1997; Hocevar et al.,1993). These findings suggest that various kinases areinvolved in M-phase lamin disassembly, with distinctkinases functioning in different cell types.

In several species and cell lines, nuclear lamins areknown to be phosphorylated in the interphase nucleus(Eggert et al., 1991; Ottaviano and Gerace, 1985;Schneider et al., 1999), but the function of interphasephosphorylation is not as clearly defined as that of mitoticphosphorylation. Interphase phosphorylation by PKC ata site close to the nuclear localization signal (NLS) ofchick lamin B2 inhibits nuclear transport (Hennekes et al.,1993), suggesting that lamin phosphorylation regulates itsspatial localization. Considerable evidence also exists for astructural role for lamins in DNA replication. Anunidentified site in human lamin B2 is phosphorylatedto a significant extent during S-phase (Kill and Hutchison,1995), suggesting that S-phase phosphorylation of laminsmay be involved in DNA replication. A few candidatesfor the interphase lamin kinase have been described.Ca2+- and cAMP-independent lamin kinases have beenshown to bind tightly to the lamin-enriched fraction ofEhrlich ascites tumor cells (Dessev et al., 1988). PKCphosphorylates nuclear lamin B2 during interphase, bothin vivo and in vitro (Fields et al., 1988), and casein kinaseII is associated with the nuclear matrix in rat liver nuclei(Tawfic and Ahmed, 1994). However, the biologicalsignificance of these observations is unclear.

An oocyte is a specialized cell that undergoes meiosisand passes on the maternal genetic material to the nextgeneration. Oocytes differ from somatic cells in theirappearance. Each oocyte contains a germinal vesicle(GV), a large, highly specialized nucleus that is arrested atthe meiotic diplotene stage in almost all animals. The GVcontains abundant nuclear material inside a nuclearmembrane that is supported mechanically by a mesh-likestructure called the nuclear lamina. In frogs and fish, theprotein lamin B3 is the major component of the GVlamina (Benavente et al., 1985; Hofemeister et al., 2002;Krohne et al., 1981; Stick, 1988; Stick and Hausen, 1985;Yamaguchi and Nagahama, 2001; Yamaguchi et al.,2001), but the mechanism by which lamin B3 istransported into and assembled inside the GV is unclear.The fish and Xenopus GV lamin B3 proteins contain anN-terminal consensus site for phosphorylation by p34cdc2

kinase, which induces lamina disassembly, in much thesame manner as the somatic lamins (Hofemeister et al,2002; Yamaguchi et al., 2001).

Recently, anti-phosphopeptide antibodies have beenused to analyze the phosphorylation status of IFs in vivo(Inagaki, 1994; Sekimata et al., 1996). In the present study,we exploited the large nuclear Xenopus GV in examining

ARTICLE IN PRESSA. Yamaguchi et al. / European Journal of Cell Biology 85 (2006) 501–517 503

lamin assembly/disassembly in vivo, using a novel hetero-geneous microinjection system. We investigated the role ofphosphorylation of amino acid residue Ser-28 of goldfishlamin B3 (GFLB3) in lamina assembly in GV-stageoocytes, since Ser-28 is a prominent consensus phospho-acceptor site for p34cdc2 kinase, using monoclonalantibodies (mAbs) directed against the phosphorylatedand non-phosphorylated forms of Ser-28.

Materials and methods

Animals and oocytes

Female goldfish (Carassius auratus) were bought fromdealers (Yamato-Koriyama, Japan) and raised in thelaboratory (Biotron Institute, Kyushu University) at16 1C. Ovaries were removed and dissected into smallpieces in goldfish Ringer’s solution (Kagawa et al.,1984). Fully grown, immature oocytes were collectedwith wide-tip Pasteur pipettes.

Female Xenopus laevis were bought from dealers andkept at 18 1C. Follice-free immature oocytes wereisolated by collagenase (Sigma, type 1A, St. Louis,MO; 0.2mg/ml,) treatment in calcium free-MBSH(Cyert et al., 1988) at 18 1C overnight. Fully grownimmature oocytes (stage VI) were discerned using abinocular microscope and used for microinjection.

Constructs of site-directed mutagenic and deletion

mutants

Site-directed mutagenic goldfish lamin B3 cDNAswere prepared by PCR-directed mutagenesis using aMutan-Super Express mutagenesis kit (Takara, Tokyo,Japan). Briefly, PCR-amplified full-length lamin B3which contained EcoRI and NdeI sites upstream thestart codon, BamHI site downstream of the stop codonwas constructed, then ligated into pKF18k (Takara)which had been double digested by EcoRI and BamHI.This construct was used as a template for PCR-directedmutagenesis. PCR was carried out in 50ml total volumecontaining 10 ng template, a mutagenic oligonucleotideforward primer that had been phosphorylated at the50-end with T4 protein kinase (Takara), a selectionreverse primer (5 pmol each), and 2.5U Takara LA Taqpolymerase. Cycling conditions were as follows: 30cycles 1min at 94 1C; 1min at 55 1C; 4min at 72 1C.Mutagenic primers for deletion or substitution of singleamino acid were designed as following (the underlinednucleotide means the changed one for substitution of anamino acid). S25A (TCC-GCC) 50-CGCGTCTCCGG-CCGGCGTCAG-30; S28A (AGC-GCC) 50-GTCTC-CGTCCGCGTCGCCCCGCGCGTCTGACG-30; T30A(ACG-GCG) 50-CGTCAGCCCGGCGCGTCTGAC-30;

T33A (ACG-GCG) 50-GACGCGTCTGGCGCGCCT-GCAG-30; S396A (TCT-GCT) 50-GACTGAATCTG-GCTCCGAGTCC-30; S398A (AGT-GCT) 50-CTGAA-TCTGTCTCCGGCTCCGAGTCC-30; S404A (TCT-GCT) 50-CCAGCAGGCGGCTGTCTCTCG-30; S406A(TCT-GCT) 50-GGCGTCTGTCGCTCGCACACA-3’;T408A (ACA-GCA) 50-GTCTCTCGCGCACCGCT-C-30. Similarly, mutagenic primers were designed to producean imitated phosphorylation form at S28. S28A-D(GCC-GAC) 50-GTCCGGCGTCGACCCGACGCGT-C-30; S28A-E (GCC-GAG) 50-GTCCGGCGTCGA-GCCGACGCGTC-30. The C-terminal CaaX box (CVVM)mutants (substituted M for C, MVVM; substituted S for C,SVVM) were also constructed. PCR primers were designedas follows. MVVM 50-CTCGGATCCGACTCACAT-GACGACCATGCTGGCGTCCTGATTG-30; SVVM 50-CTCGGATCCTCACATGACGACGCTGCTGGCGTC-CTG-30 . PCR was carried out using the combination of aforward primer (50-GGGAATTCCATATGATCACCTC-CACCCCG-30) and a reverse primer described above. PCRproducts were ethanol precipitated and dissolved in distilledwater, then transformed into E. coli strain MV1184 thathad been made competent with CaCl2.

Three N-terminal deletion mutants (ND6, ND21 andND30) and two C-terminal CaaX box deletion mutantswere constructed. PCR primers were designed asfollows. ND6, 50-GGGAATTCCATATGGAGAGCC-GCGCCTCCACCG-30; ND21, 50-GGGAATTCCA-TATGGCGTCTCCGTCCGGCGTCAGC-30; ND30,50-GGGAATTCCATATGCGTCTGACGCGCCTGC-AGGAG-30; DCVVM, 50-CTCGGATCCTCAGCTGG-CGTCCTGATTGTGTC-30; DVVM, 50-CTCGGATC-CTCAGCAGCTGGCGTCCTGATTGTG-30 (italicTC: additional nucleotides introducing the stop cadon).PCR was carried out using the combination of aforward primer (N-terminus) described above and areverse primer 50-CTCGGATCCATCCTGAGACTCA-CATGACGACGC-30 or reverse primer (C-terminus)described above and a forward primer (50-GGGAA-TTCCATATGATCACCTCCACCCCG-30). PCR pro-ducts were subcloned into the pGEM-T easy TA cloningvector (Promega, Madison, WI, USA).

Plasmids from each transformant were checked byDNA sequencing. A correct mutagenic (pKF18K) anddeletion mutant plasmid (pGEM-T easy) was digestedwith NdeI and BamHI, then ligated to pET21a vector(Novagen, Madison, WI, USA) that had been doubledigested with the same enzyme.

Purification and biotinylation of recombinant

goldfish lamin B3

E. coli-produced lamins were purified from IPTG(0.4mM)-induced culture extracts (500ml) using ion-exchange chromatography according to Heitlinger et al.


(1991) with some modification. Bacterial cell pellets wereharvested, then suspended in 30ml lysis buffer [20mMMES 2-(N-morpholino)ethanesulfonic acid, pH 6.5,150mM NaCl, 1mM DTT, 100 mM (p-amidinophenyl)methanesulfonyl fluoride hydrochloride (p-APMSF)].Twelve milligrams lysozyme (Seikagaku Corp., Tokyo,Japan) was mixed with the bacterial suspension andstood still on ice for 1 h to promote lysis. Bacterial cellpellets were sonicated and inclusion bodies weresedimented by centrifugation at 10,000g for 30min at4 1C. The crude inclusion body preparation wassonicated and washed with the lysis buffer three times.The inclusion bodies were dissolved in 30ml solubliza-tion buffer (20mM Tris, pH 9, 6M urea, 1mM DTT,100 mM p-APMSF), then recentifuged at 12,000g for30min at 4 1C. The supernatant that contains laminproteins was further purified using column chromat-ography. All recombinant lamin B3 proteins used in thisstudy were eluted from a Mono S column (Cationexchange; Amersham-Pharmacia Biotech, Buckingham-shire, UK) with a linear 0–250mM KCl in solublizationbuffer (100ml). Peak fractions eluting with 150mM KClwere collected. Pooled lamin was highly purified judgedby SDS-PAGE. All lamin preparations were stored at�80 1C until use. Purified lamin in the urea solution(about 2mg/ml, 1ml) was dialyzed against 0.1Mbicarbonate buffer (pH 8.3) to assemble lamin. NHS-LC-biotin (Pierce, Rockford, IL, USA; 1mg/ml) indimethyl formamide was added to paracrystals ofpurified lamin B3 to give a final concentration of75 mg/ml and incubated on ice for 2 h. After washingwith bicarbonate buffer containing p-APMSF (50 mM)three times, biotinylated pellets were dissolved in 8Murea (1ml) at room temperature. Soluble biotinylatedlamin was dialyzed against MES buffer (20mM MES,pH 6.5, 150mM NaCl, 1mM DTT, 10 mM p-APMSF)following to 20mM Tris, pH 8.8, 150mM NaCl, 1mMDTT, 10 mM p-APMSF to assemble lamin again.Finally, assembled biotinylated lamin was dissolved in8M urea and then diluted with 20mM Tris (pH 9), 6Murea, 2mM EGTA. Samples were divided into 20-mlaliquots and stored at �80 1C until use. Before micro-injection, the protein (20 ml) was mixed with 50mMsodium carbonate buffer including 500mM NaCl (noadjusting pH, 40 ml), then dialysed against the samebuffer in a microdialyzer overnight (over 14 h) at 4 1C.Samples were centrifuged at 10,000g for 15min, then thesupernatant was used for microinjection.

Microinjection and GV isolation

Microinjection was performed with an ultra micro-pump in conjugation with microcontroller commands(World Precision Inst., Sarasota, FL, USA). BiotinylatedE. coli-produced dialysed lamins (about 1mg/ml; 5 nl/s)

was microinjected into the cytoplasm below the GV of theanimal-hemisphere of Xenopus oocytes (15–20nl peroocyte; 10 oocyes per each sample). Injected oocytes werecultured in MBS-H for 3 h at room temperature. Dead orabnormal oocytes were removed, intact oocytes were keptin MBS-H overnight. Some oocytes were induced to startmaturation by adding progesterone (Sigma, final concen-tration 1mg/ml). Oocytes were fixed and processed forlaser confocal microscopy.

GVs (n ¼ 40) were isolated from microinjectedXemopus oocytes (n ¼ 120) in GV isolation buffer(10mM Tris-HCl, pH 7.4, 33mM NaCl, 7mM KCl,1mM spermin, 0.2% NP-40) at 0 h (initial) 8 h and 16 h.Isolated GVs (n ¼ 20) were then transferred directly toSDS sample buffer to extract total GV protein. OtherGVs (n ¼ 20) were suspended in high salt buffer (50mMphosphate buffer, pH 7.0, 1M NaCl, 1% NP-40) on icefor 1 h, then centrifuged at 10,000g for 20min. Thenuclear lamina/matrix was collected as pellets anddissolved in SDS sample buffer. Samples equivalent toten GV were electrophoresed.

Confocal microscope observation

Cultured oocytes were fixed at the appropriate time(see legends of the figures after microinjection). Methodsof oocyte fixation and processing for confocal micro-scope observation were according to Gard and Kropf(1993) with some modification. Microinjected oocyteswere fixed in 4% formaldehyde, prepared from para-formaldehyde and 0.15% glutaraldehyde (Polyscience,Warrington, UK) diluted in 0.1M MOPS (pH 7.2),2mM EGTA, 1mM MgCl2, 0.2% Nonidet P-40 at roomtemperature. Oocytes were post-fixed with 100%methanolovernight and stored at �30 1C until processing. Forconfocal microscopy, fixed oocytes were rehydrated withPBS, then trimmed by a sharp knife to remain a half-animal hemisphere including germinal vesicle. Afterincubation in 0.1M sodium tetrahydroborate in PBS for4–6 h at room temperature, oocyte blocks were rinsed inPBS containing 0.2% NP-40 (PBSN) for 1 h (exchangingevery 15min), then incubated with Cy3- or Cy5-conjugated streptavidin (KPL, Gaithersburg, MD,USA; 5 mg/ml) in PBSN containing 2% BSA for 2 h atroom temperature or overnight at 4 1C. After rinsingwith PBSN (3 h overnight), oocyte blocks were dehy-drated in 100% methanol for 1 h (exchanging every15min) and stored at �20 1C until use. After washing inmethanol, oocytes were incubated in clearing solution(benzyl-alcohol: benzyl-benzoate 1:2) and mounted inthe same solution. All steps were carried out in glasssample vials (5ml). Images were obtained with a laserconfocal scanning microscope (Zeiss LSM510, Ger-many) equipped with epifluorescence optics, appropriatefilter sets and argon/krypton laser.


Extraction of proteins

Bulk GVs were isolated from immature goldfishoocytes and a lamin-rich nuclear matrix was prepared(Yamaguchi et al., 2001). Nuclear lamina/matrix pro-teins were extracted from GVs (2� 103) by homogeniz-ing the nuclear matrix in lysis buffer (LB: 9M urea, 2%NP-40, 2% b-mercaptoethanol, and 0.8% ampholine3–10) and centrifuging at 10,000g for 30min. Proteinswere recovered from surgically removed tissues (brain,gill, heart, intestine, kidney, spleen, and testis) usingIsogen (Nippon gene, Tokyo, Japan) according to theprotocol supplied. Briefly, frozen tissues (1 g) wereground in liquid nitrogen, mixed with Isogen (10ml),homogenized, and stood at room temperature for 5min.Chloroform (2ml) was added and the mixture shakedvigorously. After a spin at 12,000g for 15min, theinterphase and organic phase were used for the isolationof proteins. DNA was removed by ethanol precipitation(2000g for 5min). Proteins were precipitated byisopropanol (15ml) at room temperature then collected(12,000g for 10min). Precipitates were washed in 0.3Mguanidine hydrochloride in 95% ethanol, centrifuged at7500g for 5min three times, then washed with ethanoland dried briefly with a vacuum pump. Proteins weredissolved in 1% SDS/6M urea with shaking at 15 1C andused for immunoblotting.

Antibodies

Mouse anti-goldfish lamin B3 polyclonal antibodieswere prepared against insoluble GV laminae. Twofemale (6-week-old) BALB/C mice were immunizedweekly. Lamina from 4� 104 isolated GVs wereemulsified in Freund’s adjuvant (Difco, Detroit, MI)for the initial injection and in incomplete Freund’sadjuvant for subsequent inoculations. The final immu-nization was without adjuvant. About 5ml of ascitesfluid was collected. The polyclonal anti-GV-lamin anti-body was purified using a protein A-conjugated column,Am-pure (Amersham-Pharmacia). The ascites wasdiluted 1000-fold with TBS (20mM Tris-HCl, pH 7.5,150mM NaCl) for use in Western blotting.

Anti-goldfish lamin B3-specific monoclonal anti-bodies were prepared against recombinant full-lengthlamin B3. While 100 mg of recombinant full-length laminB3 was used for the 1st injection, 250 mg was used for the2nd, 3rd and booster injections. The 1st screening wascarried out on ELISA plates, Xenobind, (XenoporeCorp., Hawthorne, NJ, USA) conjugated with purifiedlamin B3 (10 mg/well). Eventually, hybridomas produc-ing anti-goldfish lamin B3 monoclonal antibodies(O3E6, Q10G1) were selected by Western blottinganalysis of crude GV extracts. Hybridoma supernatantswere used for subsequent Western blotting analysis.

Anti-phosphoSer monoclonal antibodies were raisedagainst a BSA-coupled phosphopeptide (GASPS-GVSPTRLTRLQEK-C, phosphoSer underlined) de-signed from the goldfish lamin B3 cDNA sequence(Accession number AB034197 for DDBJ/EMBL/Gen-Bank). The procedure for immunization and screeningwas as described above. The 1st screening was carriedout on ELISA plates conjugated with phosphopeptide(250 nM). Finally, hybridomas producing anti-lamin B3monoclonal antibodies (C7B8D, L2C1G) were selectedby Western blotting analysis of crude GV extracts andpurified recombinant mutant lamin B3 protein in whichglutamic acid had been substituted for serine 28 (S28E).Ascites fluid was obtained by injection of hybridomacells (clones O3E6, C7B8D and L2C1G).

Anti-fish somatic lamin monoclonal antibody (L-200)was used for immunodetection of Xenopus lamin B3.This antibody did not recognize goldfish lamin B3 butdid recognize Xenopus lamin B3 (Yamaguchi andNagahama, 2001).

Electrophoresis and immunoblotting

For SDS-PAGE, BSA-coupled peptides, the extractsor immunoprecipitated protein A-conjugated beadswere mixed with an equal volume of 2� SDS samplebuffer and boiled for 1min. Proteins were separated bySDS-PAGE (7.5% or 9% polyacrylamide). Two-dimen-sional (2-D) gel electrophoresis (IEF/SDS-PAGE) wascarried out utilizing the Multiphor 2 two-dimensionalgel system (Amersham-Pharmacia Biotech.) as describedby Yamaguchi et al. (2001). Proteins were stained withsilver (2-D). Alternatively, proteins were electro-trans-ferred to immobilon membranes (Millipore, Bedford,MA). Blotted membranes were blocked with 5% non-fatdry milk in TTBS (0.1% Tween 20 containing TBS) andincubated with primary antibodies (anti-lamin B3polyclonal antibody, 1:1000 and mAbs C7B8D, O3E6hybridoma supernatants) overnight at 4 1C. After beingwashed twice with TTBS and once with TBS for 5mineach, membranes were incubated with alkaline phos-phatase- or peroxidase-conjugated goat anti-mouseIgA+G+M antibody (ZYMED, San Francisco, CA;diluted 1:1000 with TBS) for 1 h at room temperature,alternatively. Following TTBS (1� ) and TBS (2� )washes, peroxidase-conjugated products were visualizedon X-ray film with a chemiluminescence reagent (NENLifescience Products Inc., Boston, MA). Alternatively,alkaline phosphatase-conjugated goat anti-mouse IgG(ZYMED) was used to detect lamin peptides with theNBT/BCIP system. The same membrane was reprobedafter blocking following incubation in denaturationbuffer (62.5mM Tris-HCl, pH 6.8, 2% SDS, 100mM2-mercaptoethanol). Images of bands and spots on theX-ray film or 2-D gel film were obtained by scanning.


Baculovirus expression

The cDNA coding regions of the goldfish (Carassius

auratus) b-actin (fused with GST tag), cyclin A (fusedwith GST tag) and p34cdc2 (fused with His6 tag) wereligated into the multicloning site of pFASTBAC(Invitrogen, Carlsbad, CA). Recombinant baculoviruswere harvested after the transfection of Sf21 cells withBacmid DNA. Approximately 4� 105 Sf21 cells per well(35mm diameter) were subcultured in Grace’s medium(Invitrogen) every 4 days. The next day of thesubculture, Sf21cells were co-infected with baculovirusparticles His-cdc2 and GST-cycA or GST-actin alone.Cells were harvested 48 h later, and then sonicatedbriefly with MPF extraction buffer (Yamashita et al.,1992) without ATP and a-naphthylphosphate. Thesupernatants were collected by spinning and stored asa cocktail at �80 1C until use for the lamin kinase assay.Accession numbers of (C. auratus) cyclin A and p34cdc2

kinase are genbank:S79215 and D17758 for DDBJ/EMBL/GenBank, respectively.

Lamin kinase assay

The frozen cocktail (Sf21 cell lysate) was diluted 20-foldwith dilution buffer (DB: 100mM b-glycerophosphates,15mM MgCl2, 5mM EGTA, and 20mM HEPES, pH7.5) before the measurement of kinase activity. The laminkinase assay was started by mixing 20ml of lamin kinasebuffer (50mM Tris-HCl, pH 7.5, 10mM MgCl2,5mM EGTA, 10mM ATP) in the presence of purifiedE. coli-produced goldfish lamin B3 (GFLB3, 0.2mg/ml,final concentration) and 1mCi [g-32P]ATP (Amersham-Pharmacia Biotech.; 110TBq/mmol) with 5ml of Sf21 cellextract in a microtube at 22 1C. After 20min incubation,the reaction was terminated by adding 25ml of 2� SDSsample buffer. Ten microliters (1/5 of the sample) wassubjected to 12% SDS-PAGE. Gels were exposed to animaging plate for 24h and analyzed using an imaginganalyzer (Fuji Photo Film Co. Ltd., Tokyo, Japan). Thelamin kinase assay without radioisotopes was started bymixing 20ml of a mixture of lamin kinase buffer (50mMTris-HCl, pH 7.5, 10mM MgCl2, 5mM EGTA, 500mMATP) and 5ml of GFLB3 (0.2mg/ml, final concentration).After 1h incubation at 22 1C, the reaction was terminatedby adding 25ml of 2� SDS sample buffer. Five ml (1/10vol of the sample) was subjected to 7.5% SDS-PAGE,then Western blotting was carried out using anti-phosphoSer antibody (C7B8D) followed by peroxidase-conjugated secondary IgG.

Phosphatase treatment

The lamina fraction (from bulk-isolated 1� 104GV)was separated using 2-D gel electrophoresis (IEF/SDS-

PAGE) with agarose gel (pI 5–8) in the first dimension(ATTO, Tokyo, Japan). Lamin spots (pI 5.9 and pI 6.1)were excised from eight gels after negative staining(ATTO). Lamin proteins were recovered from the gelsusing a concentrator, and then dialyzed in bicarbonatebuffer (pH 11). Recombinant phosphatases were boughtfrom Upstate Com. (NY, USA). Fifty units of eachphosphatase (l-PPase, human PP1 and PP2A) wereincubated with isolated GV laminae (1� 104) or purifiedlamins from each of the excised spots in reaction buffer(50mM HEPES, pH 7.5, 2mM MnCl2, 100 mM EDTA,5mM DTT) for 1 h at 37 1C. The reaction was stoppedwith SDS sample buffer (purified lamin) or extractedwith 9M urea after brief rinses (native lamina). High-sensitive Coomassie staining (SeePico CBB stain kit,benebiosis Co., Ltd, Seoul, Korea) was carried out to seethe shift of the spot after 2D-PAGE.

Results

Microinjection of biotinylated, wild-type GFLB3

into Xenopus oocytes

In amphibians and fish, the major lamin of the oocytenucleus (GV) is lamin B3, which has a typical IFstructure (Fig. 1). The role of lamin B3 phosphorylationin assembly/disassembly in vivo remains unclear. Toinvestigate the role of lamin B3 phosphorylation, wedeveloped a heterogeneous microinjection system inwhich biotinylated wild-type GFLB3 protein was micro-injected into Xenopus oocytes. No apparent nucleartransport was observed at 2 h after microinjection(Fig. 2A), but after 4 h, nuclear transport was observed(data not shown). The biotinylated GFLB3 completelyaccumulated in the GV within 6 h (Fig. 2B), as shown bybright staining of the GV by fluorescent streptavidin.The accumulated GFLB3 assembled into the nuclearlamina during overnight culture (Fig. 2C). No aggre-gated or intranuclear spots were observed during thelamina assembly process.

In Xenopus and goldfish oocytes, lamin B3 disassem-bles and mixes with the cytoplasm when germinal vesiclebreakdown (GVBD) occurs during oocyte maturationafter hormonal stimulation. If the biotinylated GFLB3polymerizes properly in the Xenopus GV lamina, itshould disassemble when GVBD occurs. To examinedisassembly, Xenopus oocytes were cultured for 24 hafter microinjection of biotinylated GFLB3 to allowassembly of the nuclear lamina, and then maturationwas initiated by adding progesterone (1 mg/ml). TheGFLB3 depolymerized and mixed with the cytoplasmcoincident with GVBD (Fig. 2D). Therefore, GFLB3underwent nuclear transport, assembly, and disassemblyin Xenopus oocytes in morphologically identifiablesteps.

ARTICLE IN PRESS

Fig. 1. Schematic representation of mutations made in goldfish lamin B3 (GFLB3) and comparison of nuclear lamin N-terminal

amino-acid sequences. (Upper panel) GFLB3 is divided into head and tail domains flanking a central coiled-coil a-helix structure

(1a, 1b, 2a, 2b). Mutations were introduced into the N- and C-terminal flanking regions and the C-terminal CaaX-box domain. The

positions of the amino acid substitutions (Ser-25, -28, -396, -398, -404, and -406, and Thr-30 and -33) are underlined. Bracketed

letters indicate the newly introduced amino acids. The truncation positions of the N-terminal deletion mutants (ND6, ND21, andND30) are indicated in the alignments of the N-terminal head domains of vertebrate lamins. (Lower panel) B-type lamins from

goldfish (GF), Xenopus (Xe), mouse (Mu), chicken (Ch), Drosophila (Dr), sea urchin (Su), and C. elegans (Ce) are shown. All

vertebrate lamins have a conserved target Ser phosphoacceptor site for p34cdc2 kinase in the consensus motif SPTR (I/L) found in

the head domain (square bracket). This motif is not as well conserved in invertebrates: In the Drosophila lamins Dmo and C, the

(I/L) in the motif SPTR (I/L) is replaced by Thr and His, respectively, and in the sea urchin lamin, TR is replaced by AK. The C.

elegans lamin does not possess the conserved motif at all. Arrows indicate the position of the consensus Ser in the motif. The amino

acid sequence used as the synthetic peptide antigen WTpep for raising monoclonal, GFLB3-specific, anti-phosphoSer antibodies is

underlined. In GFLB3, Ser-28 is a consensus phosphorylation site for M-phase-specific p34cdc2 kinase. NLS, nuclear localization

signal.

A. Yamaguchi et al. / European Journal of Cell Biology 85 (2006) 501–517 507

Microinjection of mutated, biotinylated GFLB3 into

Xenopus oocytes

The N-terminal ‘‘head’’ domain of all vertebratelamins contains a putative phosphorylation site forp34cdc2 kinase within a conserved Ser-Pro-Thr-Arg-(Ile/Leu) motif (Fig. 1). In GFLB3, the putative p34cdc2

kinase phosphoacceptor site is amino acid residue Ser-28.To examine the role of Ser-28 phosphorylation, wemutated Ser-28 to Asp and Glu to mimic phosphoryla-tion of Ser. The resulting mutant GFLB3 proteins,S28D and S28E, were expressed in E. coli, biotinylated,and then microinjected into Xenopus oocytes. S28E waspoorly assembled and was dispersed in the nucleoplasmin vivo (Fig. 3F), whereas wild-type GFLB3 waslocalized in the nuclear lamina (Fig. 3B). S28Dassembled partially in the nuclear lamina to form afragile network (Fig. 3E).

Other conserved phosphorylation sites that flank theends of the longitudinal a-helical-rod domain (Fig. 1)were also investigated. We mutated nine Ser and Thrresidues of GFLB3 to Ala, expressed the mutantproteins in E. coli, and microinjected the biotinylatedmutant proteins into Xenopus oocytes. Four proteinshad mutations in the N-terminal-domain (S25A, S28A,T30A, and T33A), and five proteins had mutations nearthe nuclear localization signal (NLS) (S396A, S398A,S404A, S406A, and T408A). All mutant proteins weretransported to the nucleus except S28A (Fig. 3C, D,G–L), which aggregated as dots at the cytoplasmicperipheral region of the nuclear membrane. No signalsfrom S28A were detected inside the GV (Fig. 3D), evenafter more than 24 h in culture. The T30A mutant waspoorly assembled in the GV lamina and somewhatdispersed in the nucleoplasm compared to the othernuclear-transported mutants (Fig. 3G). None of the

ARTICLE IN PRESS

Fig. 2. Behavior of wild-type GFLB3 after microinjection into

Xenopus oocytes. When wild-type GFLB3 was microinjected

into Xenopus oocytes, it was not transported into the nucleus

within 2 h (A). Within 6 h, it was transported into and

accumulated in the nucleus (B), and within 16 h, it was

assembled in the lamina (C). In vitro oocyte maturation was

initiated by progesterone treatment after 16 h in culture.

Oocytes were transferred to fixing solution after formation of

white spots (D). Bars ¼ 100 mm.

A. Yamaguchi et al. / European Journal of Cell Biology 85 (2006) 501–517508

mutations at the phosphoacceptor sites inhibited GVBDduring oocyte maturation, as judged by white spotformation and visualization of fixed sections, orprevented GFLB3 depolymerization in the cytoplasm.

B-type nuclear lamins have another conservedmotif, a CaaX box, at their C-terminus (where ‘‘C’’ isCys, ‘‘a’’ is any aliphatic amino acid, and ‘‘X’’ is anyamino acid). This motif is the site of several types ofpost-translational modification, including isoprenyla-tion, proteolytic trimming, and carboxyl methylation.Modification of the Cys is necessary for nuclearmembrane targeting. To study the role of the CaaXbox in GFLB3 transport and assembly, we made a seriesof mutant proteins (CVVM mutants) in which the lastthree amino acids were deleted, or the Cys mutated toSer and Met, or the CaaX box itself was deleted(DVVM, SVVM, MVVM, and DCVVM, respectively).Upon microinjection, all of these mutant proteins weretransported to the nucleus, but none of them assembledinto the lamina (Fig. 3M–P). Since a portion of the S28Emutant assembled at the rim of the lamina (Fig. 3F),these results indicate that post-translational modifica-tion of the CaaX box is the most important factordetermining nuclear membrane targeting of lamin, afterwhich the phosphorylation status of Ser-28 regulateslamin polymerization at the nuclear lamina.

Anti-phosphoSer GFLB3 mAbs

To investigate the phosphorylation at the p34cdc2 site(Ser-28) of GFLB3, a series of phosphopeptides thatcontained variants of this motif were synthesized andused as antigens to produce specific anti-phosphoSermAbs. Since the S28E GFLB3 mutant appeared to bestmimic Ser-28-phosphorylated GFLB3, antibody screen-ing was carried out by Western blotting against E. coli-produced S28E following the conventional ELISAscreening of the phosphopeptides. Two positive hybri-domas, C7B8D and L2C1G, were isolated.

The C7B8D and L2C1G antibodies were initiallycharacterized against three BSA-coupled peptides.These peptides were: (1) WTpep, a non-phosphorylatedpeptide with the wild-type GFLB3 sequence; (2)pS28pep, Ser-28-phosphorylated WTpep; and (3)pT30pep, Thr-30-phosphorylated WTpep (Fig. 4A).C7B8D recognized only pS28pep-coupled BSA, andL2C1G recognized both the pS28pep- and WTpep-coupled BSA. Neither antibody recognized pT30pep-coupled BSA. The antibodies were then characterized byWestern blotting against the full-length mutant laminsS28A, S28D, S28E, SVVM, and MVVM, which wereproduced in E. coli (Fig. 4B, left panel).

The characteristics of all the mutant GFLB3 proteinsand antibodies used in these experiments are summar-ized in Table 1. As expected, the two anti-phosphoSermAbs (C7B8D and L2C1G) did not recognize wild-typeGFLB3 or the SVVM or MVVM mutant proteins, sincethese proteins were not phosphorylated at Ser-28.Although the CaaX box mutants have a wild-typep34cdc2 site at Ser-28, they were unable to polymerize inthe lamina because they lack the proper post-transla-tional modification of Cys (Fig. 3). In contrast, bothmAbs recognized the pseudo-phosphorylated mutantsS28D and S28E, and they did not recognize S28A.Therefore, both anti-phosphoSer mAbs recognized thespecific structure of the N-terminal domain that isinduced by phosphorylation at Ser-28.

Another mAb, O3E6, was raised against E. coli-produced GFLB3 protein and was also used in thisanalysis. Western blotting revealed that the specificitiesof O3E6 and the anti-phosphoSer mAbs (C7B8D andL2C1G) were different (Fig. 4B, left panel). O3E6recognized wild-type GFLB3 and the CaaX boxmutants, which represent the Ser-28-dephosphorylatedform of lamin B3, but did not recognize WTpep (thepeptide equivalent to amino acids 20–38 of the wild-typeprotein) (Fig. 4A). N-terminal deletion mutants ofGFLB3 (ND6, ND21, and ND30) were used todetermine the epitope that is recognized by mAb O3E6(Fig. 4C). ND21, but not ND30, was recognized byO3E6, indicating that O3E6 recognizes a complexepitope that contains amino acids 21–30 and overlapswith the epitopes recognized by the anti-phosphoSer

ARTICLE IN PRESS

Fig. 3. Behavior of GFLB3 mutated in the N- or C-terminal flanking region or CaaX box. Various GFLB3 mutants were

constructed and microinjected into Xenopus oocytes. (A) No injection, (B) wild-type, (C) S25A, (D) S28A, (E) S28D, (F) S28E, (G)

T30A, (H) T33A, (I) S396A, (J) S398A, (K) T404A, (L) T406A, (M) SVVM, (N) MVVM, (O) DVVM, (P) DCVVM. S28A remained

in the cytoplasm near the GV, but the other mutants were transported into the nucleus. Some of the transported S28D was detected

in the ‘‘rim’’ of the GV, whereas S28E was strongly detected in the nucleoplasm. All CaaX box mutants were transported into the

nucleus but remained in the nucleoplasm without assembly into the lamina. All oocytes were fixed 16 h after injection, and

fluorescence was observed using a laser-confocal microscope. Bars ¼ 100mm.


antibodies. The anti-phosphoSer mAb C7B8D, whichwas used as a negative control, did not recognize any ofthe N-terminal deletion mutants. In contrast, the mAbQ10G1 and polyclonal anti-GFLB3 ascites fluid, whichwere used as positive controls, recognized all of theN-terminal deletion mutants (Fig. 4C).

Western blotting with mAbs O3E6, C7B8D, L2C1G,and Q10G1 was carried out against proteins from isolated

goldfish GVs and nuclear lamina. Surprisingly, the anti-phosphoSer mAbs recognized a 67-kDa protein (GFLB3)in the total GV protein and lamina protein fractions (Fig.4B, right panel), indicating that the consensus Ser (Ser-28)for M-phase-specific p34cdc2 kinase is phosphorylated inthe late G2 stage in immature oocytes. In general,phosphatase treatment is a requisite control experimentin the characterization of antibodies directed

ARTICLE IN PRESS

Fig. 4. Characterization of anti-GFLB3mAbs by Western blotting. (A) BSA was coupled with three different synthetic peptides,

separated by SDS-PAGE (9%), and stained with Coomassie brilliant blue (CBB). Wild-type (unphosphorylated) WTpep ¼ 21-

GASPSGVSPTRLTRLQEK-C; pS28pep ¼ 21-GASPSGVSPTRLTRLQEK-C, phosphoSer indicated by underlining; and

pT30pep ¼ 21-GASPSGVSPTRLTRLQEK-C, phosphoThr indicated by underlining). Lane 1, BSA (�1mg) before coupling

(treated with EMCS, N-(e-maleimidocaproyloxy)succinimide)); lane 2, BSA coupled with WTpep; lane 3, BSA-pS28pep; lane 4,

BSA-pT30pep. Alternatively, Western blotting was carried out using anti-GFLB3mAbs (C7B8D, L2C1G, and O3E6). As a control,

the electroblotted membrane was incubated only with the secondary IgG; under these conditions, slight staining of BSA-coupled

pS28pep occurred (lane 3, 2nd IgG). (B) Purified E. coli-produced recombinant GFLB3 proteins (wild type, S28A, S28D, S28E,

SVVM, MVVM) (left panel) and protein samples extracted from isolated GVs and GV lamina (right panel) were separated by SDS-

PAGE (7.5%). Western blotting was then performed with anti-GFLB3mAbs (C7B8D, L2C1G, O3E6, and Q10G1). The anti-

phosphoSer mAbs C7B8D and L2C1G, which were raised against phosphopeptides, immunoreacted with pseudo-phosphorylated

forms of GFLB3 (S28D, S28E) but not with dephosphorylated GFLB3 (wild type, S28A, SVVM, MVVM). In contrast, mAb O3E6,

which was raised against E. coli-produced full-length recombinant GFLB3, did not recognize the pseudo-phosphorylated forms

(S28D, S28E) or the S28A mutant. The mAb Q10G1, which was also raised against E. coli-produced full-length recombinant

GFLB3, recognized all the recombinant proteins. All four mAbs recognized GFLB3 (67 kDa) in isolated GV or GV lamina extracts.

(C) Purified, E. coli-produced, wild-type GFLB3 or N-terminal deletion mutants (ND6, ND21, ND30) were separated by SDS-

PAGE (7.5%), and analyzed by Western blotting with anti-GFLB3mAbs (O3E6, C7B8D, Q10G1) and mouse polyclonal ascites

fluid (Polyclo) raised against isolated goldfish GV lamina. (D) Native goldfish nuclear lamina (bulk-isolated, 5� 103) were incubated

in the presence (+) or absence (�) of phosphatases l-PP, human PP1, and human PP2A. After urea extraction, proteins (from

500GV laminae) were separated with SDS-PAGE (8%), and analyzed by Western blotting with anti-GFLB3mAbs (O3E6, C7B8D)

and Polyclo. n.d., not done. (E) Phosphorylation of purified recombinant GFLB3 by lysates from Sf21 cells expressing recombinant

goldfish proteins in vitro. Left lane, no extract; middle lane, Sf21 goldfish actin extract; right lane Sf21 goldfish p34cdc2 kinase/cyclin

A extract. Incorporation of 32P into GFLB3 was investigated by autoradiography (upper panel) or Western blotting using mAbs

specific for dephosphorylated (middle panel) and phosphorylated Ser-28 (lower panel) (O3E6, and C7B8D, respectively). No

significant difference in kinase activity was detected by autoradiography, but Western blotting with C7B8D showed that GFLB3

was phosphorylated by the cdc2/cyclinA complex. Lamin-antibody complexes were detected by chemiluminescence (A, C, D, E) or

alkaline-phosphatase activity (B).


against phosphorylated amino acids. The prokaryotic andeukaryotic phosphatases commercially available (frombacteria and plants, and human PP1 and PP2A) failed to

dephosphorylate GFLB3 (Fig. 4D, right rectangle).However, when bacteriophage lambda protein phospha-tase (l-PP) was incubated with native lamina protein,

ARTICLE IN PRESS

Table

1.

Characterizationofanti-G

FLB3antibodiesincluded

inthisanalysis

Antibodies

Antigen

Type

Peptide

Wt

S28A

S28D

S28E

SVVM

MVVM

ND6

ND21

ND30

GV

Ln

Ln(2D)

a/b

MII

(I.P).

WT

pS28

C7B8D

Peptide(p.S28)

IgG1

�+

��

++

��

��

�+

++

/�+

L2C1G

Peptide(p.S28)

IgG1

(+/�

)+

��

++

��

n.d.

n.d.

n.d.

++

+/(+

/�)

+

O3E6

E.

coliB3

IgG1

��

+�

��

++

++

�+

+�/+

�

Q10G1

E.

coliB3

IgG1

��

++

++

++

++

++

++

/+�

Polyclonal

GV

lamina

��

+n.d.

n.d.

n.d.

n.d.

n.d.

++

++

++

/++

Invivo

location

ALn

AcCyto

A/D

Ln*

DNp

DNp

DNp

A/D

Ln*

ACyto

ACyto

Invitro

AA

AD

AA

AA

A

Thereactivityoftheanti-G

FLB3antibodiesusedin

this

studyis

summarized.Thestrength

oftheim

munoreactivityis

shownas–(negative),(+

/�)(positive),or+

(stronger

than(+

/�)).Ln

(Lamina)2-D

indicatestheintensity

ofthespots

withpI5.9

[spot(a)]andpI6.1

[spot(b)]by2-D

Western

blottinganalysis.Theim

munoprecipitation(I.P)ofsolublelamin

B3from

metaphase

II

(MII)extractsisshownas–(notprecipitated)or+

(precipitated).In

vivoassem

bly

wasmonitoredthroughthebehaviorofbiotinylatedrecombinantgoldfish

lamin

B3in

Xen

op

usoocytes.The

microinjected

lamin

proteinsare

distributedin

thecytoplasm

(Cyto),nucleoplasm

(Np),orlamina(Ln)in

disassem

bled(D

)orassem

bled(A

)form

s.Microinjected

S28A

aggregatesin

thecytoplasm

(AC).S28D

andND6are

notassem

bledcompletely

inthenuclearlamina(Ln*).In

vitro

assem

bly

waschecked

after

dialysisoflamin

againstMESbuffer

(pH

6.5).After

dialysis,thesampleswere

centrifuged

at10,000

gfor20min.Thelaminswerecollectedin

theassem

bledform

(A)in

thepellets

orin

thedisassem

bledform

(D)in

thesupernatants.n.d.,notdone.


C7B8D failed to recognize the GV lamina, whereas bothO3E6 and the polyclonal antibody did recognize it (Fig.4D, left rectangle). This result indicates that GFLB3 isspecifically dephosphorylated by l-PP.

Finally, in vitro phosphorylation with [g-32P]ATP wascarried out using homogenates of Sf21 insect cells co-expressing cdc2 kinase and cyclin A. Autoradiography didnot reveal any significant difference in incorporation of32P into GFLB3 for the actin control extract versus thecdc2/cyclinA extracts (Fig. 4E, upper panel). However, inWestern blots, specific phosphorylation due to cdc2/cyclinA was detectable for C7B8D, but not for O3E6 (Fig.4E, lower and middle panels, respectively). These resultsindicate that the goldfish GV lamina consists of a mixtureof Ser-28-phosphorylated and Ser-28-dephosphorylatedforms of lamin B3. Therefore, we conclude that thesemAbs recognize N-terminal structures that are induced bySer-28 phosphorylation (C7B8D) or dephosphorylation(O3E6) within the SPTR(I/L) motif of GFLB3.

Identification of phosphorylated GFLB3 in the GV

lamina

We next separated the lamina proteins by (2-D) IEF/SDS PAGE and analyzed the separated proteins byWestern blotting (Fig. 5). The goldfish GV laminacontained two distinct spots with similar apparentmolecular masses (67 kDa), and slightly different neutralisoelectrofocusing values (pI 5.9 and 6.1), as shown bysilver staining (Fig. 5A). Polyclonal GFLB3 ascites fluidimmunoreacted with both spots with similar intensity(Fig. 5B). The mAb O3E6 detected the major spot at pI6.1 (Spot b) but not the minor spot at pI 5.9 (Spot a)(Fig. 5C), whereas mAb C7B8D detected the minor spotbut not the major spot (Fig. 5D).

To confirm that Spot a was phosphorylated, weincubated isolated GV lamina with l-PP, and thensubjected it to 2-D PAGE. Staining with Coomassiebrilliant blue revealed that Spot a disappeared upontreatment with l-PP, whereas Spot b was still present(Fig. 6, left panel) and no new spot was evident. Theseresults suggest that the pI of the proteins in Spot achanged from 5.9 to 6.1 upon l-PP treatment. Further-more, when the lamin in Spot a was eluted and thendephosphorylated with l-PP, it was no longer recog-nized by mAb C7B8D (Fig. 6, right panel). These resultsindicate that the GV lamina contained a mixture ofSer-28-phosphorylated (pI 5.9) and Ser-28-depho-sphorylated (pI 6.1) lamin B3.

Oocyte-specific phosphorylation of GFLB3

Immunocytological analyses have shown that Xenopus

GV lamin B3 (XLB3) is expressed in oocytes and somespecial cells, such as neurons, muscle cells, and testicular

ARTICLE IN PRESS

Fig. 5. Two-dimensional PAGE and Western blotting analysis of goldfish GV lamina proteins. Nuclear lamina proteins extracted

from bulk-isolated GVs (2� 103) were separated by 2-D PAGE using isoelectrofocusing from pH 4–7 and SDS PAGE. (A) Two

silver-stained spots with apparent molecular masses of �67 kDa were observed: a minor spot with pI ¼ 5.9 (Spot a) and a major

spot with pI ¼ 6.1 (Spot b). (B-D) Western blotting was carried out using polyclonal anti-GFLB3 ascites (B), mAb O3E6 (C), and

mAb C7B8D (D). Note that O3E6 recognized only Spot b (pI 6.1), and C7B8D recognized only Spot a (pI 5.9). The data for all

antibodies were obtained using the same electroblotted membrane by reprobing (see Materials and methods). Molecular masses

(Mr) are in kDa.


Sertoli cells. Here, we investigated the tissue-specificexpression of B-type lamins in goldfish using Westernblotting (Fig. 7). The mAb L-200 recognizes all somatic-type lamins in which the coil 2 region contains a Y(Q/E)(E/Q)LL motif, which is present at amino acidresidues 358–362 of goldfish lamin B2 (Yamaguchi andNagahama, 2001). Since the corresponding region ofGFLB3 (residues 365–369) has the sequence HQNLL,mAb L-200 recognized E. coli-produced somatic laminB2, but not GFLB3. The anti-GFLB3mAb O3E6 didnot recognize somatic lamin B2, as expected (Fig. 7,right two lanes).

Oocyte GVs also contain somatic lamins (lamins B1 andB2), but an A-type lamin has not yet been identified inisolated goldfish GVs. GFLB3 had a slightly highermobility on SDS-PAGE than did the somatic lamins B1and B2. The somatic lamins were found in all of the goldfishtissues examined (brain, gill, heart, intestine, head kidney,

spleen, testis, and oocyte GVs) (upper panel). GFLB3 wasabundant in the oocyte GVs, and it was also detected in allsomatic tissues, although it was at significantly lower levelsthan the somatic lamins in the latter tissues (middle panel).Therefore, GFLB3 is not exclusive to oocytes, but it is mostabundant in the oocytes of adult goldfish. When tissueswere examined for expression of the phosphorylated formof lamin B3 using the anti-phosphoSer mAb C7B8D,positive signals were detected only in the oocyte GVs (lowerpanel). Thus, we conclude that phosphorylation of Ser-28 inGFLB3 is oocyte-specific in adult goldfish.

Phosphorylation of GFLB3 in vivo following

microinjection into Xenopus oocytes

Biotinylated GFLB3 was microinjected into Xenopus

oocytes, and its presence was monitored usinglaser-confocal microscopy after the oocytes were fixed

ARTICLE IN PRESS

Fig. 6. Dephosphorylation of GFLB3 with l-PP. (Left panel)Disappearance of Spot a (pI 5.9) after l-PP treatment. Native

proteins from GV lamina (1� 104) were treated with (+) or

without (�) l-PP or mixed (Mix). The lamin proteins were

separated by 2D-PAGE, and the position of each spot was

visualized using high-sensitivity staining with Coomassie

brilliant blue. Treatment with l-PP appeared to eliminate

Spot a or to shift it to pI 6.1 (Spot b), whereas it had no effect

on Spot b. Ac indicates the position of actin. (Right panel)

Dephosphorylation of eluted GFLB3 with l-PP. Purified

GFLB3 was eluted from Spots a and b (50 and 100 ng,

respectively), incubated with l-PP (50U), and analyzed by

Western blotting. Upper lane, polyclonal anti-GFLB3 ascites

fluid; lower lane, mAb C7B8D. l-PP treatment of Spot a

protein dephosphorylated it so that it could not be recognized

by C7B8D.

Fig. 7. Distribution of goldfish lamin proteins in adult tissue.

Western blot analysis using anti-goldfish lamin mAbs. Proteins

from tissue (about 1mg each of brain, gill, heart, intestine,

head kidney, spleen, and testis) and from isolated GVs (20)

were separated by SDS-PAGE (7.5%). Purified E. coli-

produced goldfish lamin B2 and B3 (full-length) were used as

controls. Lamins were detected by immunoblotting and

chemiluminescence using mAb L-200 (specific for somatic

goldfish lamins B1 and B2 (upper panel), mAb O3E6 (specific

for dephosphorylated GFLB3) (middle panel), and mAb

C7B8D (specific for phosphorylated GFLB3) (lower panel).

The mobility of each lamin in SDS-PAGE is indicated on the

right side of the panels. Lamin B3 has a higher mobility than

lamin B1 or B2.

Fig. 8. Nuclear transport and lamina assembly of GFLB3

accompanies phosphorylation in Xenopus oocytes. (Upper

panels) Confocal images of the biotinylated GFLB3

(bGFLB3) after microinjection into immature, fully grown,

Xenopus oocytes at various time points. No signal was detected

before microinjection (A, 0 h). GFLB3 accumulated in the GV

nucleoplasm (B, 8 h) and then assembled in the GV lamina

(C, 16 h). (Lower panels) Immunodetection of lamins from

isolated GVs at each time point. Total protein and lamina

protein fractions from 10 isolated Xenopus GVs were prepared

at each time point and then separated by SDS-PAGE (7.5%).

Lamins were detected by immunoblotting and chemilumines-

cence using mAbs L-200 (upper), O3E6 (middle), and C7B8D

(lower). Initial, GFLB3 before microinjection (a volume

equivalent to half the volume of the microinjected bGFLB3

was electrophoresed). L-200 immunoreacted with XLB3 but

not with GFLB3. The data for all antibodies were obtained

using the same electroblotted membrane by reprobing (see

Materials and methods).


and stained with fluorescence-labeled streptavidin. Assum-ing that phosphorylation of biotinylated GFLB3 occurs invivo, anti-phosphoSer antibody should be able to detect thephosphorylated protein in the extracts of isolated XenopusGVs after nuclear transport. Because biotinylated GFLB3and XLB3 are indistinguishable by SDS-PAGE owing totheir similar mobilities, mAb L-200, which recognizesXLB3 but not GFLB3, was used to verify the amount ofisolated GV protein at each time point (Fig. 8, blotting).

Prior to microinjection, no signal was detected in theGVs (Fig. 8A), and the purified GFLB3 was in thedephosphorylated form and was therefore recognized byO3E6, but not by C7B8D (blotting, initial). Eight hourspost-injection, almost all of the GFLB3 had accumulatedin the nucleoplasm, and a portion of the GFLB3 wasassembled in the nuclear lamina (Fig. 8B). In addition,some of the GFLB3 in the GVs and in the nuclear laminawas phosphorylated, as judged by Western blotting withmAbs C7B8D and O3E6 (blotting, 8 h). At 16h post-injection, almost all of the GFLB3 was assembled in thenuclear lamina (Fig. 8C). The amount of GFLB3 thatwas associated with the nuclear lamina at 16h was higherthan that at 8 h, although no significant change in theratio of phosphorylated/dephosphorylated protein was


observed in the total GV and lamina protein fractions(blotting, 16h). These results indicate that cross-species,in vivo phosphorylation of GFLB3 had already occurredin the interphase (late G2 stage) Xenopus oocytes,possibly inside the GVs. Thus, equilibration betweenthe phosphorylated and dephosphorylated forms ofGFLB3 appears to occur continuously in the GV lamina,at least until oocyte maturation begins.

Discussion

Lamin assembly/disassembly is dependent on the

N-terminal head domain structure

In this study, we produced monoclonal, site- andphosphorylation state-specific anti-GFLB3 antibodiesto investigate stage-specific, in vivo phosphorylation oflamin B3 in goldfish oocytes. Western blotting analysisof recombinant mutant proteins using mAbs clearlyshowed that the N-terminal structure of lamin B3(which includes Ser-28) can be either dephosphorylated(assembled) or phosphorylated (disassembled) (Table 1).Phosphorylation of Ser-28 was easily mimicked bysubstitution of Asp or Glu, as shown by the fact thatthe mutant lamins S28D and S28E behaved like Ser-28-phosphorylated GFLB3 in microinjection studies.

In contrast, various mutations of amino acid residuesnear Ser-28 had distinct effects. The T30A mutantassembled poorly in the GV lamina, like S28E, whereasS25A and T33A assembled in the GV lamina similarlyto the wild-type protein (Fig. 3). Thr-30 is conservedamong vertebrate lamins (see Fig. 1) but has not beenshown to be phosphorylated in all organisms. Therefore,the effects of Ala substitutions on lamin polymerizationappear to be dependent on their proximity to Ser-28.Thus, the conformation of the polypeptide region nearSer-28 appears to be important for lamin assembly/disassembly. Thr-33 is phosphorylated by protein kinaseA (PKA), which inhibits head-to-tail binding of theDrosophila lamin Dmo (Stuurman, 1997). In our study,an Ala substitution at Thr-33 exhibited normal behaviorin the GV and GVBD stages, suggesting that PKA is notinvolved in lamin assembly/disassembly in oocytes.

The pseudo-phosphorylated GFLB3 mutants S28Dand S28E lost the ability to assemble in vitro and invivo. In contrast, the C-terminal CaaX box mutants(SVVM and MVVM), as well as the recombinant wild-type (no substitution) behaved as dephosphorylatedforms in vitro, although they too were disassembled invivo. Thus, Cys modification (isoprenylation) is neces-sary for lamin assembly as well as formation of thecorrect N-terminal structure in vivo. These resultsindicate that the assembly of GFLB3 is regulated co-operatively by the N-terminal head domain structureand by C-terminal Cys modification in vivo.

Nuclear (GV) import of GFLB3

The S28A mutant was expected to behave like the wild-type protein (in terms of being able to assemble) in vivo,despite the fact that its N-terminal region adopts anunusual structure in vitro. The defective nuclear transportof S28A indicates that the N-terminal structure isimportant for lamin transport into the nucleus. PKC sites,which are located immediately N-terminal to the minimalNLS, regulate nuclear transport of lamins B1 and B2(Hennekes et al., 1993). Since the GV-type lamin B3 doesnot contain the corresponding Ser/Thr residues (Yama-guchi et al., 2001), we investigated the roles of other Ser/Thr residues near the NLS (Ser-396, -398, -404, and -406,and Thr-408) in this study. Substitution of an Ala residueat any of these positions did not affect normal transport orassembly in the nuclear lamina (Fig. 3). In contrast,deletion of the N-terminal 20 amino acids was sufficient toblock nuclear import of the ND21 mutant (Table 1).Therefore, a novel finding of this study is that theN-terminal head structure functions as another domainregulating nuclear transport of lamin B3 in oocytes.

How the N-terminal head structure affects nucleartransport remains unclear. In in vitro reconstitutionexperiments, coiled-coil lamin dimers associate long-itudinally to form polar head-to-tail polymers, indicat-ing that the head and tail domains are important in theearly process of lamin assembly (reviewed in Stuurmanet al., 1998). Furthermore, head-less chicken lamin B2dimers lose their propensity to associate head-to-tail(Heitlinger et al., 1992). Considering occurrence oflamin dimerization in the cytoplasm prior to transportinto the nucleus (Loewinger and McKeon, 1988), thesestudies of the lamin polymerization process lead us toconclude that the N-terminal structure of lamin B3 maybe necessary to prevent aggregation of oligomeric laminstructures in the oocyte cytoplasm in vivo.

An as-yet-unidentified cellular factor has been re-ported to interact directly with IF head domains tocontrol cytoplasmic IF organization. This factorblocked the formation of vimentin filaments in thecortex of the Xenopus animal hemisphere (Dent et al.,1992). Such a factor controlling lamin organization maybe associated with the N-terminal head of lamin B3 inthe cytoplasm. Firm proof of the existence of a proteinbinding to the lamin N-terminal head awaits its isolationand biochemical characterization.

The role of site-specific phosphorylation in the oocyte

nucleus

Another novel finding of this study is that a fractionof GFLB3 in the GV lamina reacts with the anti-phosphoSer-28mAb. Ser-28 is one of the highlyconserved phosphoacceptor sites within the flankingregions of the a-helical rod domain. All of the lamins


that have been identified to date (with the exception ofthe Caenorhabditis elegans lamin) carry this site, which ispredicted to undergo phosphorylation by the cdc2kinase (Fig. 1). Although M-phase-specific phosphor-ylation of this site within the N-terminus of theDrosophila lamin has been confirmed by mass spectrom-etry, phosphorylation does not occur in interphase cells(Schneider et al., 1999). Our results are the first evidencethat the cdc2 consensus site in lamin is alreadyphosphorylated prior to nuclear lamina disassembly,which suggests that a kinase unrelated to p34cdc2 plays arole in lamin construction in interphase nuclei.

Very little is known about the role of phosphorylationin IF dynamics during interphase. Photobleachingexperiments have shown that IF subunits can beincorporated along the entire surface of pre-existing IFs(Okabe et al., 1993; Vikstrom et al., 1992). Cytoskeletalor IF integrity is perturbed concomitant with hyperpho-sphorylation, i.e., upon exposure of the cells to potentphosphatase inhibitors (Eriksson et al., 1992). Theseresults suggest that an equilibrium between kinase andphosphatase activities is crucial for the structural main-tenance of IFs, which itself requires a steady-stateequilibrium between soluble and assembled IF subunits.

Our results indicate that lamin phosphorylation ingoldfish and Xenopus oocyte nuclei occurs beforeassembly of the nuclear lamina. The addition of aphosphoric acid to the consensus Ser (Ser-28) decreasedthe pI by 0.2 units (Fig. 5). Furthermore, treatment withl-PP eliminated Spot a (pI 5.9), which appeared to shiftto the position of Spot b (pI 6.1) (Fig. 6). Lamin subunitexchange occurs frequently, thereby continuously re-arranging and remodeling the oocyte nuclear laminastructure via mechanisms dependent on the phosphory-lation/dephosphorylation state of Ser-28, the putativeM-phase p34cdc2 targeting site, following CaaX-depen-dent membrane targeting.

The tissue expression pattern of GFLB3 differed fromthat previously described for XLB3 in immunocytologicalstudies of Xenopus. In Xenopus, the somatic laminsappeared in the mid-blastula stage during embryogenesisand completely replaced XLB3 by the tadpole stage(Benavente et al., 1985; Stick and Hausen, 1985).Although GFLB3 was detected in all tissues, itsphosphorylation was oocyte-specific. RT-PCR analysisconfirmed that GFLB3-specific transcripts were present inall tissues identified in Fig. 7, as well as in liver/pancreas,muscles, and eyes (data not shown). Thus, phosphoryla-tion appears to play a special role in the oocyte nuclearlamina. The enormous GV nuclear structure may besupported by lamin subunit exchange that occurs morefrequently in the GV than in the somatic nucleus.

Although all lamins (B1, B2, and B3) expressed ingoldfish oocytes contain the same motifs (p34cdc2 site,NLS, and CaaX box), lamins B1 and B2 are restricted tothe GV nucleoplasm and are not found in the lamina,

whereas GFLB3 is found in the lamina (Yamaguchi andNagahama, 2001). The spatial distribution of lamins inthe oocyte suggests that they play distinct roles in thedifferent regions. Site-specific phosphorylation of oocytelamins appears to regulate the spatial distribution ofeach lamin to meet various demands of the nucleus, suchas mechanical preservation of the nuclear membrane,transcription, RNA processing, DNA repair, andchromosome condensation (as described in the Intro-duction). Thus, phosphorylation of the SPTR(I/L) motifin the somatic lamins (B1 and B2) is also an importanttopic for oocyte research.

The GV-stage lamin kinases and phosphatases

Immunodetection of phosphorylated lamin B3 ingoldfish GV lamina suggests that lamin kinases arelocalized inside or adjacent to the lamina/matrix. Afterheterogeneous microinjection of GFLB3 into Xenopus

oocytes, as described in this study, GFLB3 was found inits phosphorylated form in the GV. Although ourattempts to prepare nuclear extracts with kinase activityfrom bulk-isolated goldfish GVs have not yet beensuccessful, a lamin kinase activity that was very similar tothat of the cdc2 kinase was detected in isolated Xenopus

GV lamina/matrix in vitro (Yamaguchi, unpublisheddata). Since the anti-phosphoSer antibody used in thepresent study does not recognize XLB3, we were not ableto determine whether phosphorylation of the consensusSer in the N-terminal region of XLB3 occurs in Xenopus

GV lamina in vivo. Indeed, our in vitro phosphorylationexperiments suggested that the p34cdc2 kinase might bethe lamin kinase (Fig. 4E). On the other hand, the laminkinase is not a cdk/cyclin complex, as shown byobservations that no cyclins are present before hor-mone-initiated maturation of goldfish oocytes (Yama-shita et al., 1995) and that the kinase activity of theimmunoprecipitated cdc2/cyclin A complex increasesabruptly just after GVBD (Katsu et al., 1995).

The identity of the interphase lamin phosphatase isalso unclear. In this study, l-PP was the onlyphosphatase to dephosphorylate GFLB3 (see the section‘‘Results’’). l-PP is a member of a large family of metal-containing phosphodiesterases. This family includes Ser/Thr pyrophosphatases and also contains eukaryoticPP1, PP2A, and PP2B. The phosphodiesterase motif isconserved in a number of other enzymes involved in thehydrolysis of phosphate esters (Rusnak et al., 1996).Therefore, other enzymes that contain this motif arecandidates for the interphase lamin phosphatase. Takentogether, these results lead to the conclusion that novellamin kinases and phosphatases are involved in phos-phorylation/dephosphorylation of the N-terminal con-served Ser (Ser-28) of lamin B3 prior to oocytematuration.


Acknowledgements

This study was carried out under the NIBB CooperativeResearch Program (3-108). A laser confocal microscope(Zeiss LSM510) from the Laboratory of Electron Micro-scopy, Okazaki National Research Center, was used. Thisresearch was also aided in part by Grants-in-Aid forScientific Research on Priority Areas from the Ministry ofEducation, Science, Culture, and Sports, Japan(07283104), CREST, JST (Japan Science and TechnologyCorporation), and Nissan Science Foundation, NarishigeZoological Science Award for A. Yamaguchi.

References

Aebi, U., Cohn, J., Buhle, L., Gerace, L., 1986. The nuclear

lamina is a meshwork of intermediate-type filaments.

Nature 323, 560–564.

Baskin, Y., 1995. Mapping the cell’s nucleus. Science 268,

1564–1565.

Belmont, A.S., Zhai, Y., Thilenius, A., 1993. Lamin distribu-

tion and association with peripheral chromatin revealed

by optical sectioning and electron microscopy topography.

J. Cell Biol. 123, 1671–1685.

Benavente, R., Krohne, G., Franke, W.W., 1985. Cell type-

specific expression of nuclear lamina proteins during

development of Xenopus laevis. Cell 41, 177–190.

Berezney, R., Wei, X., 1998. The new paradigm: integrating

genomic function and nuclear structure. J. Cell. Biochem.

Suppl. 30/31, 238–242.

Berezney, R., Mortillaro, M.J., Ma, H., Wei, X., Samaraban-

du, J., 1995. The nuclear matrix: a structural milieu for

genomic function. Int. Rev. Cytol. 162A, 1–65.

Bridger, J.M., Kill, I.R., O’ Farrell, M., Hutchison, C.J., 1993.

Internal lamin structures within G1 nuclei of human dermal

fibroblasts. J. Cell Sci. 104, 297–306.

Collas, P., 1999. Sequential PKC- and Cdc2-mediated

phosphorylation events elicit zebrafish nuclear envelope

disassembly. J. Cell Sci. 112, 977–987.

Collas, P., Thompson, L., Fields, A.P., Poccia, D.L.,

Courvalin, J.-C., 1997. Protein kinase C-mediated inter-

phase lamin B phosphorylation and solublization. J. Biol.

Chem. 272, 21274–21280.

Cox, L.S., Laskey, R., 1991. DNA replication occurs at

discrete sites in pseudonuclei assembled from purified DNA

in vitro. Cell 66, 271–275.

Cyert, M.S., Scherson, T., Kirschner, M.W., 1988. Mono-

clonal antibodies specific for thiophosphorylated proteins

recognize Xenopus MPF. Dev. Biol. 129, 209–216.

Dent, J.A., Cary, R.B., Bechant, J.B., Domingo, R., Klymkows-

ky, M.W., 1992. Host cell factors controlling vimentin

organization in the Xenopus oocyte. J. Cell Biol. 119, 855–866.

Dessev, G., Iovcheva, C., Tasheva, B., Goldman, R., 1988.

Protein kinase activity associated with the nuclear lamina.

Proc. Natl. Acad. Sci. USA 85, 2994–2998.

Dessev, G., Iovcheva-Dessev, C., Bischoff, J.R., Beach, D.,

Goldman, R., 1991. A complex containing p34cdc2 and

cyclin B phosphorylates the nuclear lamin and disassembles

nuclei of clam oocytes in vitro. J. Cell Biol. 112, 523–533.

Eggert, M., Radomski, N., Tripier, D., Traub, D., Jost, E.,

1991. Identification of phosphorylation sites on murine

nuclear lamin C by RP-HPLC and microsequencing. FEBS

Lett. 292, 205–209.

Eriksson, J.E., Brautigan, D.L., Vallee, R., Olmsted, J., Fujiki,

H., Goldman, R.D., 1992. Cytoskeletal integrity in inter-

phase cells requires protein phosphatase activity. Proc.

Natl. Acad. Sci. USA 89, 11093–11097.

Fields, A.P., Pettit, G.R., May, W.S., 1988. Phosphorylation

of lamin B at the nuclear membrane by activated protein

kinase C. J. Biol. Chem. 263, 8253–8260.

Fuchs, E., Weber, K., 1994. Intermediate filament proteins:

structure, dynamics, function and disease. Annu. Rev.

Biochem. 63, 345–382.

Gard, D., Kropf, D., 1993. Confocal immunofluorescence

microscopy of microtubules in oocytes, eggs, and em-

bryos of algae and amphibians. Methods Cell Biol. 37,

147–169.

Gerace, L., Burke, B., 1988. Functional organization of the

nuclear envelope. Annu. Rev. Cell Biol. 4, 335–374.

Heald, R., McKeon, F., 1990. Mutations of phosphorylation

sites in lamin A that prevent nuclear lamina disassembly in

mitosis. Cell 61, 579–589.Heitlinger, E., Peter, M., Haner, M., Lustig, A., Aebi, U.,

Nigg, E.A., 1991. Expression of chicken lamin B2 in

Escherichia coli: characterization of its structure, assembly,

and molecular interaction. J. Cell Biol. 113, 485–495.

Heitlinger, E., Peter, M., Lustig, A., Villiger, W., Nigg, E.A.,

Aebi, U., 1992. The role of the head and tail domain in

lamin structure and assembly: analysis of bacterially

expressed chicken lamin A and truncated B2 lamins.

J. Struct. Biol. 108, 74–89.

Hennekes, H., Peter, M., Weber, K., Nigg, E.A., 1993.

Phosphorylation on protein kinase C sites inhibits nuclear

import of lamin B2. J. Cell Biol. 120, 1293–1304.

Hocevar, B.A., Burns, D.J., Fields, A.P., 1993. Identification

of protein kinase C (PKC) phosphorylation sites on human

lamin B. J. Biol. Chem. 268, 7545–7552.

Hofemeister, H., Kuhn, C., Franke, W.W., Weber, K., Stick,

R., 2002. Conservation of the gene structure and mem-

brane-targeting signals of germ cell-specific lamin LIII in

amphibian and fish. Eur. J. Cell Biol. 81, 51–60.

Hoffman, M., 1993. The cell’s nucleus shapes up. Science 259,

1257–1259.

Hozak, P., Sasseville, A.M.-J., Raymod, Y., Cook, P.R.,

1995. Lamin proteins form an internal nucleoskeleton

as well as a peripheral lamina in human cells. J. Cell Sci.

108, 635–644.

Inagaki, M., 1994. Visualization and function of vimentin

phosphorylation by cdc2 kinase during mitosis. J. Cell Biol.

269, 31097–31106.

Inagaki, M., Matsuoka, Y., Tsujimura, K., Ando, S., Tokui,

T., Takahashi, T., Inagaki, N., 1996. Dynamic property of

intermediate filaments: regulation by phosphorylation.

Bioessays 18, 481–487.

Kagawa, H., Young, G., Nagahama, Y., 1984. In vitro

estradiol-17b and testosterone production by ovarian

follicles of the goldfish, Carassius auratus. Gen. Comp.

Endocrinol. 54, 139–143.

Katsu, Y., Yamashita, M., Hirai, T., Tokumoto, T., Kajiura,

H., Nagahama, Y., 1995. Molecular cloning and


immunological analysis of goldfish cyclin A during oocyte

maturation. Dev. Biol. 170, 616–625.

Kill, I.R., Hutchison, C.J., 1995. S-phase phosphorylation of

lamin B2. FEBS Lett. 377, 26–30.

Krohne, G., Dabauvalle, M.-C., Franke, W., 1981. Cell type-

specific differences in protein composition of nuclear pore

complex-lamina structures in oocytes and erythrocytes of

Xenopus laevis. J. Mol. Biol. 151, 121–141.

Loewinger, L., McKeon, F., 1988. Mutations in the nuclear

lamin proteins resulting in their aberrant assembly in the

cytoplasm. EMBO J. 7, 2301–2309.

Ma, H., Samarabandu, J., Devdhar, R.S., Acharya, R., Cheng,

P.-C., Meng, C., Berezney, R., 1998. Spatial and temporal

dynamics of DNA replication sites in mammalian cells.

J. Cell Biol. 143, 1415–1425.

McKeon, F.D., Kirschner, M.W., Caput, D., 1986. Homol-

ogies in both primary and secondary structure between

nuclear envelope and intermediate filament proteins.

Nature 319, 463–468.

Mills, A.D., Blow, J.J., White, J.G., Amos, W.B., Wilcock, D.,

Laskey, R.A., 1989. Replication occurs at discrete foci spaced

throughout nuclei replicating in vitro. J. Cell Sci. 94, 471–477.

Misteli, T., Spector, D.L., 1998. The cellular organization of

gene expression. Curr. Opin. Cell Biol. 10, 323–331.

Moir, D.M., Montag-Lowry, M., Goldman, R.D., 1994.

Dynamic properties of nuclear lamins: lamin B is associated

with sites of DNA replication. J. Cell Biol. 125, 1201–1212.

Molloy, S., Little, M., 1992. p34cdc2 kinase-mediated

release of lamins from nuclear ghosts is inhibited by

cAMP-dependent protein kinase. Exp. Cell Res. 201,

494–499.

Nickerson, J.A., Blencowe, B.J., Penman, S., 1995. The

architectural organization of nuclear metabolism. Int.

Rev. Cytol. 162, 67–123.Nigg, E.A., 1992. Assembly disassembly of the nuclear lamina.

Curr. Opin. Cell Biol. 4, 105–109.

Okabe, S., Miyasaka, H., Hirokawa, N., 1993. Dynamics of

neuronal intermediate filaments. J. Cell Biol. 121, 375–386.

Ottaviano, Y., Gerace, L., 1985. Phosphorylation of the

nuclear lamins during interphase and mitosis. J. Biol.

Chem. 260, 624–632.

Parry, D.A.D., Steinert, P.M., 1995. Intermediate Filament

Structure. R.G. Landes Co., Austin, TX, pp. 49–122.

Peter, M., Nakagawa, J., Doree, M., Labbe, J.C., Nigg, E.A.,

1990. In vitro disassembly of the nuclear lamina and M

phase-specific phosphorylation of lamins by cdc2 kinase.

Cell 61, 591–602.

Peter, M., Heitlinger, E., Haner, M., Aebi, U., Nigg, E.A.,

1991. Disassembly of in vitro-formed lamin head-to-tail

polymers by cdc2 kinase. EMBO J. 10, 1535–1544.

Rusnak, F., Yu, L., Mertz, P., 1996. Metalloenzymes and

signal transduction: the protein serine/threonine phospha-

tases, a novel class of binuclear metal-containing enzymes.

J. Biol. Inorg. Chem. 1, 388–396.

Schneider, U., Mini, T., Jeno, P., Fisher, P.A., Stuurman, N.,

1999. Phosphorylation of the major Drosophila lamin in

vivo: Site identification during both M-phase (meiosis)

and interphase by electrospray ionization tandem mass

spectrometry. Biochemistry 38, 4620–4632.

Sekimata, M., Tsujimura, K., Tanaka, J., Takeuchi, Y.,

Inagaki, N., Inagaki, M., 1996. Detection of protein kinase

activity specifically activated at metaphase-anaphase transi-

tion. J. Cell Biol. 133, 141–149.

Skalli, O., Goldman, R.D., 1991. Recent insight into assembly,

dynamics, and function of intermediate filament networks.

Cell Motil. Cytoskeleton 19, 67–79.Spector, D.L., 1993. Nuclear organization of pre-mRNA

processing. Curr. Opin. Cell Biol. 5, 442–447.

Stick, R., 1988. cDNA cloning of the developmentally regulated

lamin Llll of Xenopus laevis. EMBO J. 7, 3189–3197.

Stick, R., Hausen, P., 1985. Changes in the nuclear lamina

composition during early development of Xenopus laevis.

Cell 41, 191–200.

Stuurman, N., 1997. Identification of a conserved phosphor-

ylation site modulating nuclear lamin polymerization.

FEBS Lett. 401, 171–174.

Stuurman, N., Heins, S., Aebi, U., 1998. Nuclear lamins: their

structure, assembly, and interactions. J. Struct. Biol. 122, 42–66.

Tawfic, S., Ahmed, K., 1994. Association of casein kinase 2

with nuclear matrix. J. Biol. Chem. 269, 7489–7493.

Van Driel, R., Wansink, D.G., van Steensel, B., Grande,

M.A., Schul, W., de Jong, L., 1995. Nuclear domains and

nuclear matrix. Int. Rev. Cytol. 162A, 151–189.

Vikstrom, K.L., Lim, S.-S., Goldman, R.D., Borisy, G.G.,

1992. Steady state of intermediate filament networks. J. Cell

Biol. 118, 121–129.

Ward, G.E., Kirschner, M.W., 1990. Identification of cell

cycle-regulated phosphorylation on nuclear lamin C. Cell

61, 561–577.

Xing, Y.G., Lawrence, J.B., 1991. Preservation of specific

RNA distribution within the chromatin-depleted nuclear

substructure demonstrated by in situ hybridization coupled

with biochemical fractionation. J. Cell Biol. 112, 1055–1063.

Yamaguchi, A., Nagahama, Y., 2001. Somatic lamins in

germinal vesicles of goldfish (Carassius auratus) vitellogenic

oocytes. Cell Struct. Funct. 26, 693–703.

Yamaguchi, A., Yamashita, M., Yoshikuni, M., Nagahama,

Y., 2001. Identification and cDNA cloning of germinal

vesicle lamin B3 in goldfish (Carassius auratus) oocytes.

Eur. J. Biochem. 268, 932–939.

Yamashita, M., Fukada, S., Yoshikuni, M., Bulet, P., Hirai,

T., Yamaguchi, A., Lou, Y.-H., Zhao, Z., Nagahama, Y.,

1992. Purification and characterization of maturation-

promoting factor in fish. Dev. Biol. 149, 8–15.

Yamashita, M., Kajiura, H., Tanaka, T., Onoe, S., Nagahama,

Y., 1995. Molecular mechanisms of the activation of

maturation-promoting factor during goldfish oocyte ma-

turation. Dev. Biol. 168, 62–75.

Documents

Phosphorylation of the p34cdc2 target site on goldfish germinal vesicle lamin B3 before oocyte maturation