part of xenopus translin is localized in the centrosomes during mitosis

Pi

AD*5

R

atlsptTpetqhtiTccappcMiwXiidcf

o

m

c

Biochemical and Biophysical Research Communications 276, 515–523 (2000)

doi:10.1006/bbrc.2000.3482, available online at http://www.idealibrary.com on

art of Xenopus Translin Is Localizedn the Centrosomes during Mitosis

nna Castro,* Marion Peter,* Laura Magnaghi-Jaulin,* Suzanne Vigneron,*enis Loyaux,† Thierry Lorca,* and Jean-Claude Labbe*,1

Centre de Recherches de Biochimie Macromoleculaire, CNRS UPR 1086, 1919 Route de Mende, 34293 Montpellier cedex, France; and †Structure de Proteines, Sanofi-synthelabo, 10 Rue des Carrieres, 92504 Rueil Malmaison cedex, France

eceived August 7, 2000

control involves the presence of regulatory elements attbtcsl

aa(ete(tmaYsalpm

tfiniwcstom(orap

During oogenesis, maternal mRNAs are synthesisednd stored in a translationally dormant form due tohe presence of regulatory elements at the 3* untrans-ated regions (3*UTR). In Xenopus oocytes, severaltudies have described the presence of RNA-bindingroteins capable to repress maternal-mRNA transla-ion. The testis-brain RNA-binding protein (TB-RBP/ranslin) is a single-stranded DNA- and RNA-bindingrotein which can bind the 3* UTR regions (Y and Hlements) of stored mRNAs and can suppress in vitroranslation of the mRNAs that contain these se-uences. Here we report the cloning of the Xenopusomologue of the TB-RBP/Translin protein (X-ranslin) as well as its expression, its localisation, andts biochemical association with the protein namedranslin associated factor X (Trax) in Xenopus oo-ytes. The fact that this protein is highly present in theytoplasm from stage VI oocytes until 48 h embryosnd that it has been described as capable to inhibitaternal mRNA translation, indicates that it couldlay an important role in maternal mRNA translationontrol during Xenopus oogenesis and embryogenesis.oreover, we investigated X-translin localisation dur-

ng cell cycle in XTC cells. In interphase, although aeak and diffuse nuclear staining was observed,-translin was mostly present in the cytoplasm where

t exhibited a prominent granular staining. Interest-ngly, part of X-translin underwent a remarkable re-istribution throughout mitosis and associated withentrosomes, which may suggest a new unknown roleor this protein in cell cycle. © 2000 Academic Press

Key Words: translin; TB-RBP; centrosome; embryo;ocyte; Xenopus; Trax.

In many eukaryotic cells utilisation of functionalRNA is regulated at a posttranscriptional level. This

1 To whom correspondence should be addressed. E-mail: [email protected].

515

he 39 untranslated regions (39UTR) as well as theinding of specific cytoplasmic proteins. The associa-ion of these proteins to the 39UTR cis-acting sequencesontrols mRNA translation by either a direct suppres-ion or by modulating the stability and the subcellularocalisation of individual mRNAs (1).

During oogenesis, maternal mRNAs are synthesisednd stored in a translationally dormant form and arectivated either upon re-entry into the meiotic divisionsoocyte maturation) or after fertilisation. These mRNAsncode a number of important products such as thosehat drive the early embryonic cell divisions, establishmbryonic polarity and determine certain cell lineages2–4). In Xenopus oocytes, several studies have describedhe presence of RNA-binding proteins capable to repressaternal-mRNA translation (1). In this regard, Bouvet et

l. have demonstrated that the overexpression of the-box protein 2 (FRGY2) facilitates translational repres-ion of mRNA synthesised within the oocytes (5). As wells a direct repression, mRNA translation is also modu-ated in Xenopus germ cells by the binding of differentroteins that control the transport and localisation ofaternal mRNAs (6, 7).TB-RBP (for testis-brain-RNA-binding protein) is

he mouse homologue of Translin, a human proteinrstly identified as a DNA binding protein that recog-ises consensus sequences at the breakpoint junctions

n chromosomal translocations (8). Subsequently, itas also identified as a mRNA-binding protein, which

an bind the 39 UTR regions (Y and H elements) oftored mRNAs and can suppress in vitro translation ofhe mRNAs that contain these sequences (9). More-ver, TB-RBP serves as an attachment protein for theicrotubule association of testicular and brain mRNAs

10). In neural cell culture, administration of antisenseligonucleotides to the TB-RBP-binding Y element dis-upted RNA-protein interaction and sorting of-CAMKII and ligation mRNAs (11) suggesting an im-ortant role of this protein in mRNA localisation. In

0006-291X/00 $35.00Copyright © 2000 by Academic PressAll rights of reproduction in any form reserved.

male germ cells, TB-RBP/Translin is a nuclear proteinisisp

otsfTmc

X(iwbpvmiectbioto

M

pdwgwlef5l0piu

XwGbo

tKg

ester (Pierce) for immunisation and to immobilised bovine seruma

mts1wpN1(f

p

tcpCrm1wtMcDaA

Sb

ftbXaSt

5tmTbefuw

mbaraBwaw2ecsc

Vol. 276, No. 2, 2000 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

n pachytene spermatocytes. In the post-meiotic roundpermatids it is mainly localised in the cytoplasm andn the intercellular bridges where it may play a role inuppression of mRNA translation and in mRNA trans-ort, respectively (12).During Xenopus oocyte maturation and early embry-

genesis, the tight regulation of maternal mRNAranslation plays a major role to induce specific proteinynthesis and it is mediated by the association of dif-erent proteins to the 39UTR. We hypothesised thatranslin, as an RNA-binding protein capable to inhibitRNA translation and to modulate mRNA localisation

ould play an important role in this process.Here we report the cloning of the cDNA encoding the

enopus homologue of the TB-RBP/Translin proteinX-translin). X-translin is present at stage VI as well asn maturing oocytes. No differences on the protein levelere observed either during oocyte maturation or em-ryogenesis. In meiotic oocytes, X-Translin wasresent in both, the cytoplasm and in the germinalesicle. In somatic XTC cells in culture, X-translin isainly cytoplasmic although a residual nuclear stain-

ng of this protein is also observed in these cells. Inter-stingly, throughout mitosis a part of X-translin asso-iates with centrosomes suggesting a possible role ofhis protein in cell cycle. Finally, we have initiated aiochemical study to characterise proteins specificallynteracting with X-translin in metaphase II-arrestedocytes. Immunoprecipitation assays with anti-X-ranslin antibodies demonstrated the specific bindingf a 35 kDa protein that we identified as Trax.

ATERIAL AND METHODS

cDNA cloning of Xenopus translin. Total mRNAs from meta-hase II-arrested eggs and single-strand cDNAs were prepared asescribed in Faure et al. 1997 (13). A PCR fragment was amplifiedith two degenerate Translin oligonucleotides (cartaytayaarttycay-aycartgg and tartcnccngcdatnacnmsrttnac). Translin PCR productsere subcloned into Blue-Script vector and used as a probe to isolate

onger Translin cDNAs from a lgt10 library of metaphase II arrestedggs (Clontech). A total of 5 3 105 independent phages were trans-erred to duplicate filters and hybridised at 57°C (23 SSC, 0.1% SDS,3 Denhardt’s, 100 mg/ml of salmon sperm DNA), to the [32P]-abelled random-primed PCR insert. Filters were washed (23 SSC,.1% SDS) at 57°C for 2 h. Three rounds of plaque purification wereerformed. Purification of l phage was performed from master platenfection. After subcloning, cDNAs were sequenced and analysed bysing BLAST sequence analysis program.

Immunisation procedures and antibodies. Fusion protein of GST--translin was expressed in E. coli. Purified soluble GST-X-translinas firstly submitted to Factor-X treatment in order to eliminate theST tag (DGST-translin) and subsequently used to immunise rab-its. Immune sera directed against X-translin was affinity purifiedn immobilised GST-X-translin.Antibodies against C-terminus X-translin (a-Cter-trans) were ob-

ained by immunising rabbit with the C-terminal peptide (C)LS-EEPTPAEGK (C for coupling). The peptide was coupled to thyro-lobulin using m-maleimidobenzoyl-N-hydroxysulfosuccinimide

516

lbumin for affinity purification.

Xenopus oocytes and egg extracts. Ovaries were surgically re-oved from Xenopus laevis females and dissociated by collagenase

reatment (1 mg/ml). Oocytes were enucleated as previously de-cribed (14). Maturation in vitro was performed by incubation with0 mg/ml of progesterone. Oocyte at different times of maturationere subjected to homogenisation with oocyte buffer (20 mM Tris,H 7.5, 50 mM NaCl, 50 mM NaF, 10 mM b-gycerophosphate, 5 mMa4P2O7, 1 mM Na3VO4, 1 mM EDTA, 0.5 mM DTT, 0.1 mM PMSF,mM benzamidine). Embryos were obtained as previously described

15, 16) and subjected to homogenisation at different times afterertilisation.

Crude extracts were prepared from unfertilised Xenopus eggs asreviously described by Lorca et al. 1998 (17).

Amino acid sequencing. Peptides were generated by in-gel diges-ion as described by Shevchenko et al. (18). The sample was reduced,arbamidomethylated and digested with trypsin (Promega). Gelieces were extracted successively with 25 mM NH4HCO3 and 50%H3CN in 5% HCOOH. Extracts were pooled and the volume was

educed by evaporation in a Speed Vac concentrator. The peptideixture was desalted by purification on a Gel Loader tip containingml of Poros R2 (Perseptive Biosystems). After washing the peptidesere eluted with 60% MeOH in 5% HCOOH. The mixture was

ransferred into a nanospray needle. Nanoelectrospray MS andS/MS spectra were carried out on a Qtof mass spectrometer (Mi-

romass, Manchester). Sample was sprayed for about 20–30 min.uring the spraying, peptides were selected by the quadrupole lensnd fragmented in the collision chamber (low energy collision gas:rgon).NCBInr data banks search has been realised with a Sanofi-

ynthelabo intranet software developed for combined search motiveselonging to the same protein.

Immunoprecipitation assays. Twenty microlitres of a crude un-ertilised egg extract were diluted with 500 ml of RIPA buffer con-aining 10 mM NaH2PO4, 300 mM NaCl, 50 mM NaF, 80 mM Na-glycerophosphate, 5 mM EDTA, 0.5% Na deoxycholate, 1% Triton-100 and incubated for one hour at 4°C with 2 mg of affinity-purifiedntibodies covalently bound or not to protein A-Sepharose beads.ubsequently, immunoprecipitates were washed three times withhe same RIPA buffer and eluted by 2% SDS.

Gel filtration. An unfertilised egg extract was diluted twice with0 mM Tris-HCl pH 7.5, 100 mM NaCl, 1 mM EGTA, 1 mM dithio-hreitol (DTT) (buffer A) and clarified by centrifugation at 4°C for 20in at 50 000 rpm in the TLA 100.4 rotor of a Beckman centrifuge.he supernatant was filtered through a 0.45 mm Millipore mem-rane and applied (2 ml) to a Hiload 16/60 Superdex 200 columnquilibrated with buffer A. The flow rate was 1 ml/min, and 2 mlractions were collected. Calibration of the column was performedsing catalase (230 kDa) and ovalbumin (43 kDa) as moleculareight markers (Pharmacia).

Biochemical purification of X-translin associated proteins. Tenillilitres of crude unfertilised egg extract were diluted twice with

uffer B (50 mM Tris–HCl pH 7.8, 0.5 mM EDTA and 1 mM DTT)nd centrifuged for 30 min at 35 000 rpm in a TFA 50.38 Kontronotor. The supernatant was applied to a 50 ml DEAE Affi-Gel Blueffinity chromatography gel (Bio-Rad) preequilibrated with buffer150 mM NaCl. After washing with the same buffer, the columnas eluted with buffer B1250 mM NaCl. Ammonium sulphate wasdded in the eluent until 70% of saturation, the resulted precipitateas collected by centrifugation and the pellet was resuspended withml of 50 mM Tris–HCl pH 7.4. This fraction was divided into two

qual portions and precleared by a phase of control immunoglobulinsross-linked to protein A-Sepharose beads. Once centrifuged theupernatants were applied to 100 ml of protein A-Sepharose beadsross-linked to a-Cter-trans antibodies or 100 ml of the same phase

pre-blocked with 1 ml of a solution at 1 mg/ml of the C-terminalXtttwa

C[raos(emt2mw8pgd

(fm

c0fiwibcrflfatwwwec

R

uTtahEoachhgt

Production and characterisation of X-translin poly-citosttlitt1iCnb

nsaagtbBsmgwcsst(Xo

tsgassmbmaptgcomet


-translin peptide. Subsequently, the phases were washed threeimes with RIPA buffer adjusted to 0.2% SDS and 500 mM NaCl,hree times with RIPA buffer adjusted to 0.2% SDS and three moreimes with 50 mM Tris–HCl pH 7.5. Finally immunoprecipitatesere eluted with 150 ml of 2% SDS and assayed by silver stainnalysis and immunoblotting.

DNA gel shift analysis. The Bcl-CL1 oligonucleotide (GCCCTC-TGCCCTCCTTCCGCGGG) (8) was labelled at its 59 end with

g-32P]ATP and T4 polynucleotide kinase. Excess of [g-32P]ATP wasemoved by passage through a G25 column. For the DNA bindingssay, a control protein (10 mg) or recombinant DGST-X-translin (5r 10 mg) was incubated at room temperature with 5 mg of salmonperm DNA for 10 min. For the supershift assays DGST-X-translin10 mg) was incubated for 5 min prior addition of salmon sperm withither control immunoglobulins (2 mg) or a-X-translin antibodies (2g). Subsequently, the [g-32P]labelled Bcl-CL1 (1 ng) was added to

he mix and incubated for 15 additional minutes in a final volume of5 ml containing 20 mM Hepes pH 7.6, 3 mM MgCl2, 40 mM KCl, 2M DTT, and 5% glycerol. The DNA-protein complex was mixedith 80% glycerol, 0.1% bromophenol blue to a final concentration of% and 0.01%, respectively, and resolved by electrophoresis in a 4%olyacrylamide gel (29:1) run at 90 V for 2 h. Following drying of theel, radiolabelled DNA-protein complexes were detected by autora-iography.

Immunofluorescence microscopy. The Xenopus tadpole cell lineXTC) was grown on 80% L-15 medium supplemented with 10%oetal calf serum (FCS), 100 mg/ml penicillin and 100 mg/ml strepto-

ycin at 27°C and 2% CO2.For immunofluorescence, cells were grown on coverslips to 80%

onfluence, fixed and permeabilised with formaldehyde 3.7% and.5% Triton X-100 in IF buffer (5% FCS, 0.05% Triton X-100 in PBS)or 10 min and washed with IF buffer. Subsequently, cells werencubated with 4 mg/ml of a-X-translin antibodies for one hour andashed four times with IF buffer. Anti-rabbit biotin conjugated

mmunoglobulins (1/200, Amersham) were used as secondary anti-odies and incubated for 30 min. After washing, anti-rabbit biotinonjugated antibodies were developed either by streptavidin-texased (colocalisation assays, 1/200 Amersham) or streptavidin-uorescein (1/200 Amersham) for 10 min. Slides were then washedour times with IF buffer, stained, when required (no colocalisationssays), with 1 mg/ml propidium iodide and washed four additionalimes with IF buffer before mounting. When co-localisation assaysere performed, cells were submitted to a subsequent incubationith anti-b-tubulin monoclonal antibody (1:200 Sigma) for 1 h. Afterashing, an anti-mouse fluorescein conjugated antibody (1:200 Am-rsham) was added for 30 min. Cell preparations were analysed byonfocal laser scanning microscopy.

ESULTS

Identification of a Xenopus Translin homologue. Bysing PCR fragments amplified with two degenerateranslin oligonucleotides as a probe we isolated fromhe screening of a ggt10 library of metaphase II-rrested eggs a cDNA sequence encoding the Xenopusomologue of Translin. This X-translin cDNA (DDBJ/MBL/GenBank Accession No. AF169343) contains anpen reading frame (ORF) of 687 base pairs, predicting

protein of 228 amino acid residues. This proteinontains a leucine zipper motif within its C-terminalalf and two short basic regions within its N-terminalalf, as well as a potential transmembrane helix re-ion. The amino acid sequence shares 80% overall iden-ity to human, mouse, chicken and hamster translins.

517

lonal antibodies. Polyclonal antibodies were raisedn rabbit against a recombinant GST-X-translin pro-ein previously submitted to Factor-X treatment inrder to eliminate the GST tag (DGST-translin) andubsequently affinity purified on immobilised GST-X-ranslin (a-X-translin). When tested for immunoblot-ing on unfertilised Xenopus egg extracts (Fig. 1A, topane E), antibodies recognised a single protein migrat-ng at 28 kDa. This band completely disappeared fromhe supernatant and was quantitatively recovered inhe pellet of an a-X-translin immunoprecipitation (Fig.A, top lanes SN and IP, respectively). Likewise, affin-ty purified anti-C-terminus X-translin antibodies (a-ter-trans) specifically immunoprecipitated endoge-ous X-translin from Xenopus egg extracts (Fig. 1A,ottom lanes E, SN, and IP).

Xenopus translin binds single-stranded DNA oligo-ucleotide that contains chromosomal breakpoint con-ensus sequence. It has been reported that humannd mouse translin binds the Bcl-CL1 oligonucleotide,target sequence within the clustered breakpoint re-

ion of the Bcl-2 oncogene in follicular lymphoma pa-ients (9, 19). To test whether X-translin also binds thisreakpoint consensus sequence we synthesised thecl-CL1 oligonucleotide and confirmed by gel shift as-ay, using increasing doses of DGST-translin (5 and 10g), that it strongly binds to the labelled Bcl-CL1 oli-onucleotide (see Fig. 1B, lanes 2 and 3). This bindingas not observed when an equivalent amount of a

ontrol protein was used (10 mg, lane 1). We then usedpecific a-X-translin antibodies to characterise the gelhift band. As shown in Fig. 1 (panel B, lane 4), a-X-ranslin antibodies, but not control immunoglobulinsFig. 1B, lane 5), could supershift the band formed by-translin and Bcl-CL1 confirming the specific bindingf this protein to the breakpoint consensus sequence.

X-translin expression during Xenopus oocyte matura-ion and embryogenesis. It has been previously de-cribed that the amount of TB-RBP in mouse maleerm cells changes according to the state of differenti-tion. The levels of this protein in meiotic pachytenepermatocytes and post-meiotic round spermatids areignificantly higher than that observed for the post-eiotic elongated spermatids (12). This difference has

een proposed to correlate with a role of this protein ineiotic recombination in pachytene spermatocytes,

nd also in storage and translational suppression ofaternal mRNAs in round spermatids, meanwhilehese two processes are not present any more in elon-ated spermatids. In order to investigate whetherhanges in X-translin amount also occurs in Xenopusocytes, we have analysed the levels of this protein ineiotic cells (stage VI and maturing oocytes), and in

mbryo. In contrast to the male mouse germ cell sys-em, X-translin levels in both, diplotene-arrested stage

V(ia2

pcattamirXidopAd

Vm3

plibcisnti2sIsiuse

iebtc(sDtmiw


I oocytes and mature metaphase II-arrested oocytesFig. 1C, lanes VI and MII, respectively) were founddentical to that in early embryo, at least until 48 hfter fertilisation (Fig. 1C, lanes 3 h, 5 h, 6 h, 7 h, 9 h,4 h, 31 h, and 48 h).

Cellular localisation of X-translin in maturing Xeno-us oocytes and XTC cells in culture. Besides thehanges in TB-RBP levels on distinct male germ cells,different nuclear/cytoplasmic distribution of this pro-

ein has also been described (12). In this regard, al-hough Western blot analysis shows an identicalmount of TB-RBP in the nucleus and the cytoplasm ofouse pachytene spermatocytes, nearly all the protein

s present in the cytoplasmic fraction of post-meioticound spermatids. We have also investigated-translin distribution by Western blot analysis on

solated germinal vesicle (GV) and cytoplasm ofiplotene-arrested oocytes (stage VI) and in maturingocytes at different times following the release ofrophase block and until germinal vesicle breakdown.s shown in Fig. 2A, no changes of nuclear/cytoplasmistribution of X-translin was observed either in stage

FIG. 1. (A) Specificity of anti-X-translin and anti-C-terminus X-trmmunoprecipitated with a-X-translin antibodies (a-X-translin) or aach matrix (IP), as well as 10% of the corresponding supernatants (Sy Western blotting and probed with a-X-translin antibodies in aranslin immunoprecipitation. (B) Binding of the recombinant DGontaining specific chromosomal breakpoint sequences. [32P]-labelledlane 1) or with two different concentrations of recombinant DGST-Xhift analysis. a-X-translin antibodies (2 mg) (lane 4) or controlGST-X-translin (10 mg), then, [32P]-labelled Bcl-CL1 (1 ng) was add

ranslin complex supershifted. (C) X-translin expression during Xetaphase II arrested oocyte (MII) or embryos at different times afte

n 20 ml of oocyte buffer (see Methods) and centrifuged 3 min at 13,0ith a-X-translin antibodies. Equal protein amounts corresponding

518

I oocytes (lanes VI, GV and Cy) or throughout oocyteaturation (lanes 1 h, 2 h, 3 h, 4 h, GV, and 1 h, 2 h,h, 4 h Cy).Previous studies have shown that translin was

resent in the cytoplasm of most human somatic cellines in culture, whereas nuclear localisation was lim-ted to hematopoietic cells (8). Likewise, in mouserain, TB-RBP is present either in the nucleus or in theytoplasm depending on the neuron cell type (20). Tonvestigate the subcellular localisation of X-translin inomatic cells, we have performed confocal laser scan-ing microscopy analysis in Xenopus XTC cells in cul-ure. Asynchronous XTC cells were stained with affin-ty purified a-X-translin antibodies in interphase (Fig.B, 1) and in mitosis as monitored by propidium iodidetaining (Fig. 2B, 2, X-translin and DNA respectively).n interphase, although a weak and diffuse nucleartaining was observed, X-translin was mostly presentn the cytoplasm where it exhibited a prominent gran-lar staining (Fig. 2B, 1) similar to that obtained forome mitochondrial proteins (data not shown). Inter-stingly, part of X-translin underwent a remarkable

lin polyclonal antibodies. Twenty ml of unfertilised egg extracts wereterminus X-translin antibodies (a-Cter-trans). Material eluted fromand an equivalent volume of the starting material (E) were analyseder-trans immunoprecipitation and a-Cter-trans antibodies in a-X--X-translin to the Bcl-CL1 single-stranded DNA oligonucleotidel-CL1 (1 ng) was incubated either with a 10 mg of a control proteinnslin (5 mg and 10 mg, lanes 2 and 3, respectively) and tested by gel

unoglobulins (2 mg) (lane 5) were incubated with recombinantto the mixture and tested for detection of [32P]-labelled Bcl-CL1/X-pus oocyte maturation and embryogenesis. Stage VI oocyte (VI),rtilisation (3 h, 5 h, 6 h, 7 h, 9 h, 24 h, and 48 h) were homogenisedrpm. The supernatant was run on SDS–PAGE and immunoblottedne oocyte was loaded per lane.

ans-C-N),-Ct

STBc

-traimmed

enor fe00to o

rcsia

bceds

d((pcmi


edistribution throughout mitosis and associated withentrosomes (Fig. 2B, 2 and 3 X-Translin). This centro-omal localisation was confirmed by using a doublemmunofluorescent labelling with anti X-translin andnti b-tubulin antibodies (Fig. 2B, 3, X-translin and

FIG. 2. (A) Distribution of X-translin in stage VI and maturing Xeeveloped in the germinal vesicle (VG) or the cytoplasm (Cy) of stage V1 h, 2 h, 3 h, 4 h, respectively). Equivalent protein amounts correspB) Subcellular localisation of X-translin in XTC cells. Single a-Xro-metaphase cell (2, X-translin) as visualised by propidium iodideence (3, X-translin, 3, b-tubulin, respectively) in a metaphase cellagnification for all micrographs). The localisation of X-translin in th

n all the immunostainings.

519

-tubulin). Moreover, immunostaining with the mono-lonal antibody a-6Histidine in transfected XTC cellsxpressing 6Histidine-tagged X-translin have alsoemonstrated its centrosomal pattern (data nothown). This is the first report describing the localisa-

pus oocytes. Western blot analysis with a-X-translin antibodies wereVI) or maturing oocytes at 1, 2, 3, and 4 h after progesterone additioning to one germinal vesicle or one cytoplasm were loaded per lane.

anslin immunofluorescence in interphase cell (1, X-translin) andining (2, DNA) and double a-X-translin/a-b-tubulin immunofluores-e bar in the lower right micrograph corresponds to 50 mm (same

ifferent stages of the cell cycle was identical for all the cells observed

noI (ond-trsta. The d

tmp

pdomswtTIwpeWpapX

mXpXpuftTu7rmSatteoaX

cffsot


ion of TB-RBP/Translin at the centrosomes duringitosis and could suggest a new important role of this

rotein in cell cycle.

X-translin is present as a complexed protein in Xeno-us egg extracts. Native form of Translin has beenescribed as a ring-shaped structure with an assemblyf eight subunits of 27 kDa monomer whose molecularass is approximately 220 kDa (21). Moreover, several

tudies have shown that TB-RBP/Translin can interactith different proteins such as Trax, g-Actin, and the

ransitional endoplasmic reticulum ATPase (22–25).o investigate if X-translin is present in metaphaseI-arrested oocytes in association with other proteins,e have developed a gel filtration analysis. Figure 3,anel A, shows the elution profile of the unfertilisedgg extracts submitted to gel filtration and probed byestern Blot with a-X-translin antibodies. Two major

eaks with approximate molecular weight of 43 kDand 230 kDa were observed. This pattern indicates theresence of at least two different populations of-translin, which could reflect either the dimeric or

FIG. 3. (A) Side-by-side analysis by Western blotting with ahromatography by gel filtration on a Superdex 200 column of an unfeor catalase (230 kDa) and ovalbumin (43 kDa). (B) Analysis of a Xenorom a-Cter-trans antibodies was analysed by SDS–PAGE and silveraturated with the antigen peptide encoding the C-terminal sequencbtained from p35. The four-peptide sequences obtained from p35 prohe database (right). Numbers denote the positions of the peptides i

520

ultimeric forms of this protein or the association of-translin with other proteins. To characterise therotein/s that could specifically interact with-translin we developed a subsequent biochemical ap-roach by using a a-Cter-trans immunoaffinity col-mn. The starting material was an extract preparedrom unfertilised Xenopus eggs. A low speed superna-ant was applied onto a DEAE Affi-Gel Blue column.he major part of X-translin was retained on the col-mn, eluted with 250 mM NaCl and precipitated at0% of ammonium sulphate saturation. The pellet wasesuspended and clarified with a matrix of control im-unoglobulins cross-linked to protein A-Sepharose.ubsequently, it was submitted to X-translin immuno-dsorption on the a-Cter-trans antibodies cross-linkedo protein A-Sepharose or the same column blocked byhe corresponding antigen peptide. Proteins wereluted from the affinity matrix by 2% SDS. One tenthf the eluent was submitted to SDS–polyacrylamide gelnd silver staining. As shown in Fig. 3B, besides p28/-translin (confirmed by Western blot) another protein

translin antibodies of consecutive fractions (30 ml loaded), afterised egg extract. The vertical arrows indicate the positions of elutionX-translin complex by SDS–PAGE. The purified preparation eluted

ined (lane 1). Lane 2, same experiment, but the antibodies were firstf X-translin. (C) Amino acid sequences of the proteolytic fragmentslysis (left) are compared to the human Trax sequence obtained from

he protein.

-X-rtilpusstae oteon t

of 35 kDa was specifically immunoprecipitated anddws

Ttaiuesf(LTadwsso

D

tgoswosdbsmmst

pnmUlsfmpmeqttT

trans-acting factor for paternal mRNA translation con-tem

fisdIibAXtqXeTpsItastmstpsswds

XptTpctctspaiia

wdosr3


isappeared when the antibody was first saturatedith the antigen peptide (compare lanes 1 and 2, re-

pectively).

Identification of p35-X-translin interacting protein asrax. To identify p35 protein, the remaining eluent of

he a-Cter-trans column was loaded onto SDS–PAGE,nd Coomassie-blue stained. The gel band correspond-ng to p35 was in-gel digested, treated as describednder Materials and Methods and submitted to nano-lectrospray MS and MS/MS spectra on a Qtof masspectrometer. From the recorded MS/MS mass spectraour partial sequences were found: XXF[IL][IL]HRMW 5 898.5), XnAY[IL]GNTG (MW 5 2132.9), Xn[I-][IL]GVAD[IL]TGE[IL]MR (MW 5 2318.1), and XX-PG[IL]QEYVEA[IL]T[FM]K (MW 5 1812.9) ([IL]: themino acid is either I or L, X: Unknown AA). NCBInrata banks search revealed a high degree of homologyith human, rat and chicken Trax, a previous de-

cribed Translin-associated protein which has beenuggested to be involved in the active nuclear transportf Translin (22, 23, 25) (Fig. 3C).

ISCUSSION

The translational regulation of maternal mRNAs ishe primary mechanism by which stage-specific pro-rams of protein-synthesis are executed during devel-pment (26). In Xenopus oocytes, maternal mRNAs aretored as messenger ribonucleoprotein particles inhich the proteic components act as direct repressorsf translation (1, 5, 27). Maternal mRNAs are alsoelectively and spatially distributed in Xenopus oocytesuring oogenesis and early development. This distri-ution results in targeted translation of mRNAs topecific regions of the cell (28). These mechanisms ofRNA-translation control implies the presence ofRNA binding proteins that are capable to directly

uppress mRNA translation, to anchor mRNAs to cy-oskeletal structures or both of them.

Translin has been described both, as a DNA-bindingrotein (8) and as an RNA-binding protein that recog-ises conserved sequences in the 39 UTRs of storedRNAs. In vitro, TB-RBP/Translin binds to the 39TRs at the Y and H elements and suppresses trans-

ation of testicular mRNAs (9). Moreover, there areome evidences that translin serves as a linker proteinor the association between microtubule and specificRNAs, indicating that it acts as a multifunctional

rotein (10). It has been proposed that in the nucleus ofouse male germ cells, TB-RBP/Translin would pref-

rentially bind to meiotic chromosomal breakpoint se-uences, while in the cytoplasm could bind specificesticular mRNAs to microtubules and inhibit paternalranslation of mRNAs by facilitating their storage (12).he fact that TB-RBP/Translin could be an essential

521

rol in mouse male germ cells tempted us to hypoth-sise a possible role of this protein on the regulation ofaternal mRNA in Xenopus oocytes.To begin to characterise Xenopus Translin we have

rst cloned the Xenopus homologue of this protein. Itsequence shares 80% overall identity to that previousescribed for human, mouse, chicken and hamster.nterestingly, no homologue sequences are present innvertebrates, which could indicate that Translin haseen acquired along evolution in the vertebrate group.s reported by Aoki et al. (8), we have confirmed that-translin binds single-stranded DNA oligonucleotide

hat contains chromosomal breakpoint consensus se-uence. We subsequently proceed to the analysis of-translin levels in Xenopus stage VI oocytes and earlymbryos. It has been previously described (12) thatB-RBP/Translin protein levels is high in mouseachytene spermatocytes and round spermatids andignificantly drop in late stage elongated spermatids.n keeping with its suggested role in mRNA transla-ion control, this decrease may reflect the necessity toctivate protein synthesis in order to undergo the mas-ive structural changes induced as the round sperma-id transforms into shaped spermatozoon (29). Unlikeale spermatogenesis, maternal mRNAs translation is

pecifically inhibited not only during oogenesis but alsohroughout early development where stage-specificrograms of protein synthesis are executed. Our re-ults show constant TB-RBP/Translin levels at alltages from diplotene-arrested oocytes to 48 h embryos,hich suggest that this protein is at work not onlyuring gametogenesis (spermatogenesis and oogene-is) but also during early development.We also analysed the localisation pattern of

-translin in stage VI oocytes. Our results show that arominent level of X-translin is in the cytoplasm al-hough it is also detectable in the germinal vesicle.his localisation pattern agrees with what has beenreviously described in mouse germ cells. In theseells, the highest amount of nuclear Translin (50% ofhe total protein) was observed in pachytene spermato-ytes where it could play a role in meiotic recombina-ion. This protein moves to the cytoplasm in roundpermatids where it would function in mRNA trans-ort and translation. Similarly, stage VI oocytes, whichre arrested at diplotene where meiotic recombinations already finished, present a mainly cytoplasmic local-sation of X-translin where it probably may also playn important role in mRNA transport and translation.One of the more interesting results of the presentork is the presence of X-translin in the centrosomesuring mitosis, which could suggest an important rolef this protein in cell cycle control. In this regard,everal important proteins for cell cycle control areedistributed to the centrosome during mitosis (30–3). As well as a role in mitotic cell cycle control, the

centrosomal localisation of this protein during cell di-vtidtlm

sedtoATodca2wRcpRoamspb

leatopmitistdTmttpr

A

m

was supported by grants from the “Association pour la Recherche surllFb

R

1

1

1

1

1

1

1

1

1

1

2

2

2

2

2

2


ision could also allow a correct redistribution of cer-ain mRNAs between the two daughter cells. In keep-ng with this hypothesis, Morales et al. haveemonstrated the presence of TB-RBP/Translin pro-ein at intercellular bridges of male germ cells. Thisocalisation was interpreted as a mean to distribute

RNAs between male germ cells (12).Finally, the results of the gel-filtration analysis

howed the presence of a large X-translin-containingdifice of 230 kDa. To deeper study this complex weeveloped a biochemical approach by using a a-Cter-rans immunoaffinity column which allowed us dem-nstrating the specific binding of a protein of 35 kDa.minoacid sequencing of p35 reveals this protein asranslin associated factor X or Trax. Trax is a memberf the TB-RBP/Translin family and it has already beenescribed associated to TB-RBP/Translin in both, nu-lear and cytoplasmic complexes from neural sources,s well as in cytosolic extracts from mouse testis (22–4). Although the presence of this complex seems to beidely distributed in the different tissues where TB-BP/Translin is expressed the role of this binding is

ompletely unknown. In this regard, some studies hy-othesise a possible role of Trax in transporting TB-BP/Translin to the nucleus. This idea has been basedn the fact that, unlike TB-RBP/Translin, Trax amino-cid sequence contains a bipartite nuclear targetingotif in its N-terminal region (25). However, our re-

ults show that this association is also present in meta-hase II-arrested egg extracts where nuclear mem-rane has disappeared.In summary, we have cloned the Xenopus homo-

ogue of TB-RBP/Translin and we have studied itsxpression, its localisation as well as the biochemicalssociation with its partners. The fact that this pro-ein is highly present in the cytoplasm from stage VIocytes until 48 h embryo, as well as that it has beenreviously described as capable to inhibit paternalRNA translation, indicate that it could also play an

mportant role in maternal mRNA translation con-rol during oogenesis and embryogenesis. Moreover,ts localisation in the centrosomes during mitosis inomatic XTC cells suggests a new unknown role ofhis protein in cell cycle control. Finally, we haveemonstrated the association of TB-RBP/Translin torax. This complex has also been described in mouseale germ cells and brain which suggests an impor-

ant role of Trax binding to TB-RBP/Translin. Fur-her studies are needed to define whether these pro-osed functions of TB-RBP/Translin pertain to itsole in Xenopus oocytes and somatic cells.

CKNOWLEDGMENTS

We thank Alain Devault and Marcel Doree for critical reading theanuscript and Nicole Lautredou for technical assistance. This work

522

e Cancer” and the “Ligue Nationale contre le Cancer-Equipe label-isee LA LIGUE.” A.C. is a postdoctoral fellow supported by “Humanrontier Science Program,” L.M.J. is a postdoctoral fellow supportedy the “Ligue Nationale contre le Cancer.”

EFERENCES

1. Sommerville, J. (1992) Bioessays 14, 337–339.2. Hake, L. E., and Richter, J. D. (1997) Biochim. Biophys. Acta

133, M31–M38.3. St. Johnston, D., and Nusslein-Volhard, C. (1992) Cell 68, 210–

219.4. Wickens, M., Kimble, J., and Strickland, S. (1996) Translational

control of developmental decisions, Cold Spring Harbor Labora-tory Press, Cold Spring Harbor, NY.

5. Bouvet, P., and Wolffe, A. P. (1994) Cell 77, 931–941.6. Havin, L., Git, A., Elisha, Z., Oberman, F., Yaniv, K., Schwartz,

S. P., Standart, N., and Yisraeli, J. K. (1998) Genes Dev. 12,1593–1598.

7. Deshler, J. O., Highett, M. I., and Schnapp, B. J. (1997) Science276, 1128–1131.

8. Aoki, K., Suzuki, K., Sugano, T., Tasaka, T., Nakahara, K., Kuge,O., Omori, A., and Kasai, M. (1995) Nat. Genet. 10, 167–174.

9. Wu, X. Q., Gu, W., Meng, X., and Hecht, N. B. (1997) Proc. Natl.Acad. Sci. USA 94, 5640–5645.

0. Han, J. R. Y., G. K., and Hecht, N. B. (1995) Proc. Natl. Acad.Sci. USA 92, 9550–9554.

1. Severt, W. L., Biber, T. U. L., Wu, X. Q., Hecht, N. B., DeLorenzo,R. J., and Jakoi, E. R. (1999) J. Cell Sci. 112, 3691–3702.

2. Morales, C. R., Wu, X. Q., and Hecht, N. B. (1998) Dev. Biol. 201,113–123.

3. Faure, S., Vigneron, S., Doree, M., and Morin, N. (1997) EMBOJ. 16, 5550–5561.

4. Dabauvalle, M. C., Doree, M., Bravo, R., and Karsenti, E. (1988)Cell 52, 525–533.

5. Sagata, N., Watanabe, N., Vande Woude, G. F., and Ikawa, Y.(1989) Nature 342, 512–518.

6. Furuno, N., Nishizawa, M., Okazaki, K., Tanaka, H., Iwashita,J., Nakajo, N., Ogawa, Y., and Sagata, N. (1994) EMBO J. 13,2399–2410.

7. Lorca, T., Castro, A., Martinez, A. M., Vigneron, S., Morin, N.,Sigrist, S., Lehner, C., Doree, M., and Labbe, J. C. (1998) EMBOJ. 17, 3565–3575.

8. Shevchenko, A., Chernushevich, I., Ens, W., Standing, K. G.,Thomson, B., Wilm, M., and Mann, M. (1997) Rapid Commun.Mass Spectrom. 11, 1015–1024.

9. Kasai, M., Matsuzaki, T., Katayanagi, K., Omori, A., Maziarz,R. T., Strominger, J. L., Aoki, K., and Suzuki, K. (1997) J. Biol.Chem. 272, 11402–11407.

0. Wu, X. Q., Petrusz, P., and Hecht, N. B. (1999) Brain. Res. 819,174–178.

1. Aoki, K., Suzuki, K., Ishida, R., and Kasai, M. (1999) FEBS Lett.443, 363–366.

2. Wu, X. Q., Lefrancois, S., Morales, C. R., and Hecht, N. B. (1999)Biochemistry 38, 11261–11270.

3. Taira, E., Finkenstadt, P. M., and Baraban, J. M. (1998) J. Neu-rochem. 71, 471–477.

4. Muramatsu, T., Ohmae, A., and Anzai, K. (1998) Biochem. Bio-phys. Res. Commun. 247, 7–11.

5. Aoki, K., Ishida, R., and Kasai, M. (1997) FEBS Lett. 401, 109–112.

26. Stebbins-Boaz, B., and Richter, J. D. (1994) Mol. Cell. Biol. 14,

22

23

31. Freed, E., Lacey, K. R., Huie, P., Lyapina, S. A., Deshaies,

3

3


5870–5880.7. Richter, J. D., and Smith, L. D. (1984) Nature 309, 378–380.8. Litman, P., Barg, J., Rindozoonski, L., and Ginzburg, I. (1993)

Neuron 10, 627–638.9. Hecht, N. B. (1998) BioEssays 20, 555–561.0. Debec, A., and Montmory, C. (1992) Biol. Cell. 75, 121–126.

523

R. J., Stearns, T., and Jackson, P. K. (1999) Genes Dev. 13,2242–2257.

2. Fry, A. M., Arnaud, L., and Nigg, E. A. (1999) J. Biol. Chem. 274,16304–16310.

3. Bailly, E., Doree, M., Nurse, P., and Bornens, M. (1989) EMBOJ. 8, 3985–3995.

part of xenopus translin is localized in the centrosomes during mitosis

Documents