Development 108, 289-298 (1990)Printed in Great Britain © The Company of Biologists Limited 1990

289

A two-step model for the localization of maternal mRNA in Xenopus

oocytes: Involvement of microtubules and microfilaments in the

translocation and anchoring of Vg1 mRNA

JOEL K. YISRAELI, SERGEI SOKOL, D. A. MELTON

Department of Biochemistry and Molecular Biology, Harvard University, 7 Divinity Avenue, Cambridge, MA 02138, USA

Summary

In an effort to understand how polarity is established inXenopus oocytes, we have analyzed the process oflocalization of the maternal mRNA, Vgl. In fully grownoocytes, Vgl mRNA is tightly localized at the vegetalcortex. Biochemical fractionation shows that the mRNAis preferentially associated with a detergent-insolublesubcellular fraction. The use of cytoskeletal inhibitorssuggests that (1) microtubules are involved in the trans-location of the message to the vegetal hemisphere and (2)

microfilaments are important for the anchoring of themessage at the cortex. Furthermore, immunohisto-chemistry reveals that a cytoplasmic microtubule arrayexists during translocation. These results suggest a rolefor the cytoskeleton in localizing information in theoocyte.

Key words: localization, maternal mRNA, Vgl, polarity,microtubules, microfilaments, Xenopus oocytes.

Introduction

The animal-vegetal (A-V) axis of Xenopus oocytesbecomes the primary axis of the embryo after fertiliz-ation. It is along the A-V axis that the only knowndevelopmental differences exist in an unfertilized egg,with ectoderm being formed from the animal hemi-sphere and endoderm from the vegetal hemisphere(Nakamura et al. 1970). The dorsal-ventral axis, estab-lished by cortical rotation after fertilization, formsorthogonal to the A-V axis (Vincent and Gerhart,1987). The asymmetric orientation of the mitochondrialcloud opposite the chromatid attachment site in primor-dial germ cells is thought to presage the A—V polarityobserved in oocytes (Al-Mukhtar and Webb, 1971;Heasman et al. 1984). This polarity is interpreted inoocytes so that organelles and maternally encodedfactors, such as mRNA and protein, necessary for thedevelopment of particular regions of the early embryowill have attained their proper spatial organization bythe end of oogenesis. How the A-V polarity is gener-ated and interpreted is not understood.

Intracellular localization of specific mRNAs has beenreported in a number of different cell types. Cytoplas-mic actin mRNA is localized to the lamellipodia offibroblasts in culture, where active actin protein syn-thesis is occurring (Lawrence and Singer, 1986). RNAencoding MAP2, a dendrite-specific microtubule-associated protein, is enriched in dendrites in thedeveloping brain (Garner et al. 1988). In Drosophila,

both zygotic and maternal mRNAs have been identifiedwhich are localized to particular regions of the syncytialblastoderm or egg. Transcripts of a number of pair-rulegenes, such asfushi tarazu, hairy, and evenskipped, arelocalized apically above the nucleus in embryos beforecellularization (Akam, 1987). Two localized, maternalmRNAs in Drosophila, bicoid and nanos, are localizedat the anterior and posterior ends of the egg, respect-ively, and the spatial organization of their proteinproducts is thought to be essential in setting up the basicbody plan of the fly (Niisslein-Volhard et al. 1987;Driever and Ntisslein-Volhard, 1988; MacDonald andStruhl, 1988; R. Lehmann, personal communication).

In Xenopus, several localized, maternal mRNAshave been isolated (Rebagliati et al. 1985). The bestcharacterized of these is Vgl, a vegetally localizedmessage whose protein product is a member of theTGF/3 family (Melton, 1987; Weeks and Melton, 1987).Several TGF/J-like molecules have been implicated inthe process of mesodermal induction (Smith et al. 1988;Rosa et al. 1988), and our recent unpublished resultssuggest that Vgl, like heterologous TGF/3-1 (Kimelmanand Kirschner, 1987), enhances the induction of meso-derm by FGF. Distributed homogeneously in youngoocytes, Vgl mRNA is translocated in early stage IVoocytes to a tight shell along the vegetal cortex (Melton,1987), a region approximately 5/xm in depth consistingof the plasma membrane and associated internal ma-terial (Franke et al. 1976). Following hormonal matu-ration of an oocyte to an egg, Vgl mRNA is released

290 J. K. Yisraeli, S. Sokol and D. A. Melton

from the tight cortical shell and forms a broader band inthe vegetal hemisphere which remains in vegetal cellsthroughout early embryogenesis (Weeks and Melton,1987). Translation of Vgl protein product beginstoward the end of oogenesis and continues until gastru-lation (Dale et al. 1989; Tannahill and Melton, 1989).Throughout early development, Vgl protein is largelyconfined to the vegetal hemisphere of embryos, whereit is glycosylated and associated with membranes. VglmRNA and protein are thus both located in the regionimportant for the induction of mesoderm during theperiod when this induction is occurring (Nieuwkoop,1969; Smith, 1989).

Regardless of its role in development, understandinghow Vgl mRNA is properly localized should provideinformation on the process of localization of mRNA inparticular, and on the interpretation of polarity withincells in general. Injection of exogenous Vgl mRNA hasshown that ds-acting sequences in the Vgl mRNA areimportant for its localization (Yisraeli and Melton,1988). The cellular machinery involved in the localiz-ation process has not yet been elucidated. In an effort todissect the process into its component parts, we haveexamined the involvement of the oocyte cytoskeleton inattaining the proper distribution of Vgl message. Ourdata suggest a two-step model for the localization ofVgl mRNA in which translocation of Vgl message tothe vegetal cortex is dependent on cytoplasmic micro-tubules and anchoring is achieved with the involvementof cortical microfUaments.

Materials and methods

Oocyte culture and drug treatmentAlbino frogs were anesthetized with 0.5 g benzocaine (Sigma)in 11 of water, and late stage III oocytes (0.5-0.6 mm as sizedby an ocular micrometer) were manually isolated in IXmodified Barth's saline-Hepes (MBSH; Gurdon, 1968) sup-plemented with 0.1 gl"1 penicillin/streptomycin. Oocyteswere then transferred in a sterile fashion to Terasaki plates(Vangard International), one to two oocytes per well, andcultured for five days at 20°C in a humidified chamber in 0.5XLeibowitz medium supplemented with lmM-L-glutamine,l^gmP1 insulin, 15mM-Hepes (pH7.8), 50 units ml"1 nysta-tin, lOOugmP1 gentamycin, lOOjigmP1 penicillin,100^g ml streptomycin, and 10% frog serum containingvitellogenin (Wallace et al. 1980).

Frog serum containing vitellogenin was obtained from frogsinjected three weeks earlier with 0.4 ml/100 g body weight of alOmgmP1 solution of estradiol-17 /S (Sigma) suspended inpropylene glycol (Wallace et al. 1980). Blood was collected byheart puncture, allowed to coaggulate for several hours atroom temperature, and then spun in an analytical table topcentrifuge (~2000revsmin~1) to isolate serum. On average,3-4 ml of green serum was obtained per frog, which wasimmediately aliquoted and frozen until use.

Cytoskeletal inhibitors were added when the oocytes wereseeded and whenever the medium was changed (usually thefirst and third days after seeding) at the following concen-trations: 25/igmP1 cytochalasin B, 1/^gmP1 nocodazole.lmgmP 1 colchicine, 1/igmP1 tubulozole-C, and l/<gmltubulozole-T (all were suspended in DMSO except for colchi-cine, which was dissolved in ethanol). All concentrations of

inhibitors were chosen based on previous reports in whichthese drugs were effective (Coleman et al. 1981; Kimelman etal. 1987; Geuens et al. 1985), with the exception of nocod-azole, which we found to be equally effective over a 10-foldrange of concentrations (10 to ljtgmP1).

Stage VI oocytes were isolated in the same way as thesmaller oocytes but were cultured overnight in IX MBSH in96-well plates using the same concentrations of drugs as withmiddle stage oocytes. Maturation of stage VI oocytes wasachieved in vitro by treating with 1 ^M-progesterone (inethanol) for approximately 8-10 h (Holwill et al. 1987),monitoring germinal vesicle breakdown in cocultured pig-mented oocytes. All oocyte stages mentioned in this paper areaccording to Dumont (1972).

Oocyte extractions and RNA analysisUsing the basic protocol of Jeffery and Meier (1983), wevaried the detergent, salt concentration and temperature ofoocyte extractions in order to optimize conditions for thespecific partitioning of Vgl mRNA into the insoluble fractionand fibronectin mRNA into the soluble fraction. AlthoughBrij, NP40, and Triton X-100 all solubilized fibronectinmRNA well, 0.5% Triton gave the best recovery of VglmRNA in the insoluble pellet fraction. Salt concentrationsbelow 0.3M-KC1 do not solubilize yolk, while increasingconcentrations yield higher percentages of non-localized fi-bronectin mRNA in the insoluble fraction. Temperaturesfrom 4°C to room temperature (23 °C) had little effect on thefractionation, and hence room temperature was chosen forthe extractions so as not to disrupt microtubules. The opti-mized procedure is as follows. Ten oocytes were homogenizedin 0.5 ml of buffer containing 0.5 % Triton X-100,10 mM-Pipes(pH6.8), 0.3M-KC1, lOmM-magnesium acetate, 0.5 ITIM-EGTA, 10 fig of yeast tRNA and 20mM-vanadyl ribonucleo-side complexes. After incubation at room temperature for5-10 min, the lysate was spun at 13 000 revs min"1 for 5min.To prepare RNA from the insoluble fraction, the pellet wasresuspended in 0.4ml of buffer A (50mM-Tris-HCl (pH7.5),O.lM-NaCl, 10mM-EDTA, 0.5% SDS, 10/tg of yeast tRNAand 400/ig per ml of proteinase K) and digested for lh at45 °C. The RNA was then extracted twice with phenol/chloroform (1:1) and precipitated with 0.1 volumes of 3 M-sodium acetate and 2.5 volumes of cold ethanol. The RNAfrom the soluble fraction of the original lysate was preparedby diluting the supernatant 1:4 with buffer A and processingas described above. The relative amounts of poly (A)+ RNAin the fractions was determined by incorporation of labelednucleotide in a reverse transcription reaction primed witholigo (dT).

RNA samples prepared from pellets, supernatants andunfractionated lysates as a control were electrophoresed indenaturing formaldehyde agarose gels and transfered toGeneScreen as described (Rebagliati et al. 1985). In all cases,the same number of oocyte equivalents of RNA was loadedonto each lane of the gel. Northern blot analysis was per-formed with 32P-labeled anti-sense Vgl (Weeks and Melton,1987) and fibronectin probes (Krieg and Melton, 1985).Quantification of the relative amount of signal in the pelletand soluble fractions was performed by scanning the auto-radiographic film with a densitometer.

In situ hybridizationOocytes were fixed, sectioned and hybridized as previouslydescribed (Melton, 1987). Animal-vegetal orientation of thealbino oocyte sections was determined by examining serialsections for the position of the germinal vesicle and thedistribution of yolk platelets. A detailed protocol for in situ

The localization of Vgl mRNA 291

hybridization is being published elsewhere (O'Keefe et al. inpress). The probe used for all the in situ hybridizationspresented here was a 32P-antisense T7 transcript to the entirecoding region of Vgl (see Weeks and Melton, 1987). All insitu hybridizations shown here are dark-field photographs; thesilver grains appear white under these optical conditions.

Whole-mount immunohistochemistryOocytes were defolliculated in 0.5% collagenase type IV(Sigma) in IX PBS at room temperature, with gentle rocking.Oocytes were fixed and stained as described by Dent et al.(1989), using a horseradish peroxidase-conjugated goat anti-mouse second antibody (Cappel) to visualize the monoclonalanti-/3tubulin antibody (a gift from M. Klymkowsky). Controloocytes, incubated with either the second antibody alone or inconjunction with a monoclonal antibody for a muscle-specificmarker not present in oocytes (12101, a gift from C. Kintner),showed no appreciable staining.

Results

Association of Vgl mRNA with a detergent-insolublefraction of oocyte extracts at specific times duringoogenesisVgl mRNA is initially distributed homogeneouslythroughout the cytoplasm of small oocytes and under-goes localization to the vegetal cortex as the oocytegrows from 0.6 to 0.8 mm in diameter (early stage IV)(Fig. 1A; Melton, 1987; Yisraeli and Melton, 1988).Because injected Vgl message, which undergoes asimilar localization in cultured oocytes, and endogen-ous message appear to maintain steady-state RNAlevels throughout localization, the accumulation of VglmRNA at the vegetal cortex seems to be a translocationprocess rather than localized degradation of the mess-age. Translocation of other macromolecules and organ-elles, in different cell types of many organisms, isknown to involve cytoskeletal elements (see below). Inorder to determine whether Vgl mRNA might beassociated with a particular subcellular compartment ofoocytes or eggs, detergent extracts of oocytes wereanalyzed for Vgl mRNA.

The RNA from detergent-soluble and insoluble (pel-let) fractions was isolated and analyzed by Northernblot analysis (Fig. IB). Oocytes in which Vgl mRNAlocalization has occurred (stage VI, lanes 4-6) contain alarge majority of their Vgl mRNA (approximately80%) in the insoluble fraction (lane 4). In eggs (lanes4-7), however, where in situ hybridizations show thatVgl mRNA is released from its tight cortical shell, VglmRNA is found almost exclusively in the solublefraction (lane 8). Other messages, such as fibronectin,which are uniformly distributed, are found almostcompletely in the soluble fraction of extracts from bothof these stages (approximately 90-95 %, lanes 5 and 8).In young oocytes (stage II, lanes 1-3), both Vgl andfibronectin mRNA are found in the detergent solubleand insoluble fractions in about equal proportions. Thepresence of fibronectin RNA in detergent insolublepellets of young oocytes was repeatedly observed, butcannot be explained by the same mechanism that servesto localize Vgl RNA since fibronectin RNA is uni-

st. II early st. VI unfert.st. IV egg

oocytes

B st. II st. VI eggsP S T P S T P S T

fb

vgl

1 2 3 4 5 6 7 8 9

Fig. 1. Specific association of Vgl mRNA with thedetergent-insoluble fraction of extracts. (A) Schematicrepresentation of the distribution of Vgl mRNA in oocytesand eggs. Vgl mRNA (dark shading in picture) is initiallydistributed homogeneously in early stage oocytes (stage II),undergoes a process of localization in middle stage oocytes(early stage IV), and culminates in a tight cortical shell bythe end of oogenesis (stage VI; Melton, 1987). Unfertilizedeggs have Vgl mRNA distributed in a broad band in thevegetal hemisphere along the cortex (egg; Weeks andMelton, 1987). (B) Northern blot analysis of RNAdetergent-extracted from oocytes and unfertilized eggs.Oocytes or eggs were homogenized and fractionated into asoluble fraction and an insoluble pellet by centrifugation, asdescribed in Materials and methods. RNA was thenpurified from the pellet (P, lanes 1,4, and 7), the solublefraction (S, lanes 2, 5, and 8) or total, unfractionatedextract (T, lanes 3, 6, 9) and analyzed by Northern blothybridization with both a fibronectin (fb) and Vgl probe.Two oocyte-equivalents of RNA was run in each lane.Lanes 1-3, from stage II oocytes; lanes 4-6, from stage VIoocytes; and lanes 7-9, from unfertilized eggs.

formly distributed in the cytoplasm. The insolublefraction, which is generally considered to contain mostof the cytoskeletal elements of the cell (Schliwa et al.1981; Jeffery and Meier, 1983), contains only 2-5 % ofthe total poly (A)+ RNA in stage VI oocytes (seeExperimental Procedures). Inasmuch as the same num-ber of oocyte-equivalents was loaded in each lane onthe gel in Fig. IB, we can calculate that, per fig of totalpoly (A)+ RNA, Vgl mRNA is enriched approxi-mately 100-fold in the detergent-insoluble fraction ofonly those oocytes that contain localized Vgl mRNA.In contrast, fibronectin mRNA shows no detectableenrichment in the detergent-insoluble fraction at thisstage. These results are similar to those reported byPondel and King (1988), although the data presentedhere suggest a much higher enrichment of Vgl mRNAin the insoluble fraction.

292 J. K. Yisraeli, S. Sokol and D. A. Melton

Effect of cytoskeletal inhibitors on the localization ofVgl mRNA in late stage oocytesThe biochemical studies described above and those ofPondel and King (1988) suggest that cytoskeletal el-ements may be involved in the localization of VglmRNA. The specific effects of a number of cytoskeletalinhibitors are well established. Cytochalasin B binds tothe barbed end of actin filaments and disrupts theirnormal organization (Cooper, 1987). Nocodazole andcolchicine bind specifically to tubulin and disrupt micro-tubules (Wilson and Bryan, 1974). To explore furtherhow Vgl mRNA interacts with the cytoskeleton, latestage (stage V/VI) oocytes were treated overnight withvarious inhibitors, fixed the next morning, and analyzedby in situ hybridization (Fig. 2). Oocytes incubated insaline (Fig. 2A), nocodazole (Fig. 2C), or colchicine(data not shown) contain Vgl mRNA distributed in atight cortical shell in the vegetal hemisphere, just asseen in stage V/VI oocytes in vivo (see Fig. 1A).Cytochalasin B treatment, however, causes a release ofVgl message from this tight distribution, resulting in abroad band of Vgl mRNA along the cortex (Fig. 2D).The degree to which the message disperses is somewhatvariable, but generally seems to be limited to thevegetal hemisphere. In late stage oocytes, actin micro-filaments are found predominantly along the cortex(Franke et al. 1976). Our results suggest that cytochal-asin B causes a release of Vgl mRNA from its normallytight cortical distribution by disrupting cortical micro-filaments and thereby allowing passive diffusion of themessage.

Vgl mRNA in unfertilized eggs is distributed in abroad band in the vegetal hemisphere (Weeks andMelton, 1987), similar to that seen after cytochalasin Btreatment. To determine when this release occursduring development, stage VI oocytes were matured invitro by culturing in the presence of progesterone,which releases the oocyte from its block at first meioticprophase and, among many other changes, inducesgerminal vesicle breakdown (Fig. 2B). Clearly, the

distribution of Vgl mRNA in progesterone-maturedoocytes is indistinguishable from that seen either incytochalasin-B-treated oocytes (Fig. 2D) or in unferti-lized eggs (Weeks and Melton, 1987). Thus, the releaseof Vgl from its tight cortical position is directly connec-ted to hormonal maturation of oocytes. The cytochal-asin B induced release of Vgl mRNA is not a result ofmaturation, however, as can be seen from the presenceof the germinal vesicle in those sections (Fig. 2D). It isinteresting to note that the few pigment granulespresent in the vegetal cortex undergo the same sort ofinward migration upon maturation as does Vgl mRNA(Hausen et al. 1985), suggesting that the cytoskeletalreorganization known to occur during maturation(Wylie et al. 1985; Dent and Klymkowsky, 1988) isresponsible for both phenomena.

Vgl mRNA is associated with the detergent-insol-uble fraction of late stage oocytes (Fig. IB). To testwhether cytoskeletal inhibitors can affect this associ-ation, RNA was isolated from the soluble and insolublefractions of detergent extracts of drug-treated stage VIoocytes and analyzed as above by Northern blot analy-sis (Fig. 3). As expected from the in situ hybridizations,neither nocodazole (lane 4) nor colchicine (data notshown) had any effect on the presence of Vgl mRNA inthe insoluble fraction. In accordance with its effect onthe distribution of Vgl message, cytochalasin B treat-ment caused a release of most of the Vgl mRNA intothe soluble fraction (lane 8). Densitometer tracings ofNorthern blots show that an average of 65 % of VglRNA is found in the detergent insoluble fraction inuntreated stage VI oocytes, whereas only 28% of VglRNA is found in the pellet after cytochalasin B treat-ment. Although the intracellular distribution of VglmRNA is similar in both unfertilized eggs and cytochal-asin B-treated oocytes, not all of the Vgl message isreleased into the soluble fraction after cytochalasin Btreatment, as appears to be the case after maturation(see Fig. IB). This remnant of Vgl mRNA in thedetergent-insoluble fraction may imply that other cyto-

Fig. 2. Vgl mRNA distribution in drug-treated,late-stage oocytes. In situ hybridization using aVgl probe reveals the localization of themessage in oocytes incubated overnight insaline (A), progesterone (B), nocodazole (C),or cytochalasin B (D). The scale bar indicates500/tm. The germinal vesicle (gv) is indicated inthose sections where it is present.

controlP S T

nocodP S T

cytoBP S T

fb

vg1

The localization of Vgl mRNA 293

-serum +serum

no drug

1 2 3 4 5 6 7 8 9

Fig. 3. Specific release of Vgl mRNA into the solublefraction of detergent extracts by treating with cytochalasinB. Stage VI oocytes cultured overnight in saline with nodrug treatment (control; lanes 1-3), with nocodazole(nocod; lanes 4-6), or with cytochalasin B (cytoB; lanes7-9) were homogenized and their RNA analyzed as inFig. 1. RNA purified from the insoluble pellet was run inlanes 1, 4, and 7, from the soluble fraction in lanes 2, 5,and 8, and from total extracts in lanes 3, 6, and 9.

skeletal elements not sensitive to cytochalasin B areassociated with the message. In fact, Pondel and King(1988) have shown that the insoluble fraction fromdetergent extracts of late stage oocytes is enriched incytokeratins and vimentin. Nevertheless, we can con-clude that at least one component of the structureholding the Vgl mRNA at the cortex is sensitive tocytochalasin B.

Effects of cytoskeletal inhibitors on the distribution ofVgl mRNA in middle stage oocytesExamining the effects of cytoskeletal inhibitors on latestage oocytes provided information about the mechan-ism for anchoring Vgl message at the cortex. In orderto analyze the process of translocation of the message, itwas necessary to use an in vitro system for culturingmiddle stage oocytes which are just beginning toundergo localization. Immature oocytes grown in vitroin the presence of serum from estradiol-injected frogsincrease in diameter as a result of the uptake ofvitellogenin, a yolk protein precursor, from the serum(Wallace et al. 1980). In addition, the changes thatnormally occur during oogenesis in vivo occur incultured oocytes as well, including the migration of thegerminal vesicle to the animal hemisphere and thelocalization of Vgl message to the vegetal cortex(Yisraeli and Melton, 1988).

Late stage III oocytes were cultured in the presenceof various cytoskeletal inhibitors for 5 days and thenanalyzed by in situ hybridization (Fig. 4). Oocytesgrown in the presence of the vitellogenin-containingserum but without any inhibitors demonstrate a markedlocalization of the Vgl message at the vegetal cortex(Fig. 4B), as compared with either uncultured oocytes(data not shown) or those cultured in medium withoutserum (Fig. 4A). Notably, cytochalasin B treatedoocytes are also capable of translocating Vgl mRNA(Fig. 4C), although in these oocytes, the Vgl messageaccumulates in the vegetal cytoplasm and not in a tight

+cytochalasin B

+nocodazole

Fig. 4. The effects of cytoskeletal inhibitors on thetranslocation of Vgl mRNA in middle stage oocytes. Latestage III oocytes (0.5-0.6 mm in diameter) were cultured invitro for 5 days and then assayed for Vgl mRNAdistribution by in situ hybridization. Oocytes were growneither in medium without serum (A) or in medium withserum, either with no drug (B) or in conjunction withcytochalasin B (C) or nocodazole (D). The scale barrepresents 200 pm.

subcortical shell. A possible explanation for this atypi-cal distribution is that the Vgl mRNA cannot beanchored at the cortex when microfilaments are dis-rupted (see Fig. 2D).

The translocation of Vgl message can be disrupted bydrugs that interfere with microtubule polymerization;both nocodazole (Fig. 4D) and colchicine (data notshown) completely prevent translocation of the VglmRNA. Despite the dramatic and specific effects of themicrotubule inhibitors on Vgl mRNA localization, noeffect on the viability of the oocytes was detected withany of the drugs, as assayed by comparing the profile ofproteins synthesized in vitro on the last day of theculture (data not shown). In addition, the amounts ofVgl mRNA, as assayed on Northern blots, are constantand unaffected by any of the treatments (data notshown). Thus, cytoskeletal elements sensitive to nocod-azole and colchicine, but not those sensitive to cytochal-asin B, are evidently associated with the translocationof Vgl mRNA in middle stage oocytes. In other words,these data implicate microtubules as being involved inthe movement of Vgl message to the vegetal hemi-sphere.

Further evidence for the specific involvement ofmicrotubules in the movement of Vgl mRNA comesfrom the use of a relatively new tubulin-binding drug

294 J. K. Yisraeli, S. Sokol and D. A. Melton

tub-C tub-T Table 1. Growth of oocytes cultured under variousconditions

Fig. 5. The specific effect of depolymerizing microtubuleson the translocation of Vgl mRNA in middle stage oocytes.Late stage III oocytes were cultured in vitro in mediumcontaining serum and either the potent microtubuledepolymerizing agent tubulozole-C (tub-C) or its inactivetrans isomer tubulozole-T (tub-T) and assayed as in Fig. 4.The scale bar represents 200 [an.

and its trans isomer. Tubulozole-C is a photoisomer ofcolchicine produced by UV irradiation, which revers-ibly depolymerizes microtubules; tubulozole-T, its transisomer, does not inhibit microtubule polymerizationeven at concentrations well above those of tubulozole-C(Geuens et al. 1985). Oocytes cultured in tubulozole-Tgrow normally and show normal localization of VglmRNA. Those grown in tubulozole-C, however, showno movement of the message and appear identical tooocytes treated with other microtubule inhibitors(Fig. 5). These data suggest that the inhibition oftranslocation is a consequence of depolymerizingmicrotubules, not some side effect of the drugs.

Oocytes grow in diameter, both in vivo and in vitro,by receptor-mediated endocytosis of vitellogenin fromthe serum (Wallace and Jared, 1976). In our earlierwork, we noted a strong correlation between the size ofan oocyte, grown either in vivo or in vitro, and thedegree of localization of Vgl mRNA (Yisraeli andMelton, 1988). It is formally possible, therefore, thatdrugs that prevent the growth of oocytes might preventthe localization of Vgl mRNA as well. Both micro-tubule and microfilament inhibitors severely reducedthe growth of stage III oocytes in culture (Table 1),perhaps by interfering with normal endocytosis (e. g.see Dustin, 1978). Nevertheless, it is clear from the datapresented in Table 1 that it is possible to uncouplegrowth and Vgl localization. Cytochalasin B-treatedoocytes, despite their poor growth, translocate VglmRNA. Also, oocytes cultured in the presence of fetalcalf serum without any vitellogenin do not grow at allbut show normal localization of Vgl message (Table 1and data not shown); this result further suggests thatvitellogenin uptake is not important for translocation.In light of those results, it is interesting that oocytescultured in medium alone without any serum are unableto localize Vgl mRNA. How serum stimulates localiz-ation of the message remains a mystery. It is clear,however, that the lack of localization in oocytes cul-tured in the presence of microtubule inhibitors is not aresult of simply preventing growth of the oocyte.

Growth medium

Saline aloneMedium aloneMedium+frog serum (containing vitellogenin)Medium+frog serum+cytochalasin BMedium+frog serum+colchicineMedium+frog serum+nocodazoleMedium+frog serum+tubulozole-CMedium+frog serum+tubulozole-TMedium+fetal calf serum

% increasein volume

(")

0(12)0(18)

54(20)10 (20)5(23)

11(22)11(23)36(12)5(13)

Late stage III oocytes (0.5-0.6 mm in diameter) were culturedfor 5 days in saline (IX MBSH), medium (supplemented Leibowitzmedium), or medium with serum, with the indicated drug, asdescribed in Materials and Methods. The increase in volume wascalculated for each group of oocytes by cubing the ratio of thediameter of the oocytes at the end of the incubation period to thediameter of the oocytes at the beginning of the incubation period.The data shown represent the average from three independentexperiments.

Visualization of the microtubule array in oocytesAlthough the drug studies provide strong evidence forthe involvement of cytoskeletal elements in both thetranslocation and anchoring of Vgl mRNA, it is import-ant to know that the appropriate cytoskeletal structuresare present at the right time in the proper place. Asmentioned above, actin filaments are highly enriched inthe cortical region of oocytes, where they appear to bearranged in 'whirls' (Franke etal. 1976). Coleman etal.(1981), using electron microscopy, have demonstratedthat cytochalasin B, but not colchicine, completelydisrupts cortical microfilaments in late stage oocytes.Thus, in stage VI oocytes, not only does cytochalasin Brelease Vgl mRNA from its tight localization at thecortex but also disrupts the array of cortical microfila-ments.

The presence of a radial tubulin array in oocytes wasfirst reported by Palecek et al. (1985) using immunohis-tochemistry on oocyte sections. The location and ar-rangement of tubulin in oocytes during oogenesis hasalso been visualized using the whole-mount proceduredeveloped by Klymkowsky and colleagues (Dent andKlymkowsky, 1988). We have employed the lattertechnique to monitor the appearance of the array andits sensitivity to cytoskeletal inhibitors. As seen inFig. 6A, a radial tubulin array, emanating from thegerminal vesicle and extending throughout the cyto-plasm to the periphery of the oocyte, is first evident instage II oocytes, using an anti-/3 tubulin antibody; thesestructures are thus present before translocation of theVgl message begins. This array is fairly symmetric inthe early and middle stage oocytes, but as the germinalvesicle moves up to the animal hemisphere, the stainingbegins to disappear from the vegetal region (Fig. 6C).(By this point in oogenesis, Vgl mRNA is alreadycompletely localized). In stage V/VI oocytes, the radialarray persists, emanating from the germinal vesicle, but

AStage VI

Bo c

cyto B nocod

The localization of Vgl mRNA 295

translocation anchoring

©-still

1 .1T

nocodazolecolchicine

microtubuledependent

earlySt. IV

tcytochalasin B

microfilamentdependent

middlest IV

Fig. 7. A two-step model for the localization of VglmRNA in oocytes. The diagram represents the simplestinterpretation of the data presented in the paper. VglmRNA distribution (indicated by shading) becomesprogressively restricted to the vegetal cortex during normaloogenesis (indicated by the horizontal arrows). Microtubuleinhibitors, such as nocodazole or colchicine, block the firstpart of this process, which we have termed translocation,and result in the prevention of any noticeable migration ofVgl mRNA. Microfilament inhibitors, such as cytochalasinB, prevent the second step of the localization, which wehave called anchoring, and result in the release, in latestage oocytes, of the Vgl mRNA from its tight cortical shellinto a broad band along the cortex. (In middle stageoocytes when the translocation process is occurring,cytochalasin B prevents anchoring at the cortex as well andresults in the ectopic accumulation of Vgl mRNA in thecytoplasm above the cortex; see Fig. 4C).

Fig. 6. Whole-mount immunohistochemistry of /3-tubulin inoocytes. Oocytes were defolliculated, fixed and stainedusing an anti-/3-tubulin antibody according to the procedureof Dent et al. (1989). (A) Untreated oocytes of theindicated stages were stained. The staining in stages II andIII is radially symmetric. The stage VI oocyte is shownviewed from the animal hemisphere. (B) Stage VI oocyteswere incubated in either cytochalasin B (cyto B) ornocodazole (nocod) overnight in saline and then processedas above. Both are shown as viewed from the animalhemisphere. The punctate staining of the nocodazole-treated oocyte, evidence for the disruption of themicrotubule array, is detectable only in the animalhemisphere. The scale bar for A and B represents 500 fan.(C) A lateral view of an untreated stage VI oocyte atslightly higher magnification showing the tubulin arrayemanating from around the gv evident only in the animalhemisphere. The scale bar represents 500/OTI.

is detectable in only the animal hemisphere (Fig. 6A).The microtubule nature of the array is evident from itssensitivity to depolymerizing drugs such as nocodazole;both stage VI and stage IV oocytes treated with thesedrugs have completely disrupted arrays and show onlypunctate staining (Fig. 6B and data not shown). Noeffects on the tubulin array were detected when oocyteswere incubated in the presence of cytochalasin B(Fig. 6B). Thus, a cytoplasmic tubulin array that issensitive to microtubule inhibitors is present through-out the oocyte during the period of translocation of VglmRNA.

Discussion

The two-step model diagrammed in Fig. 7 integratesthe data presented here. We propose that Vgl mRNA isassociated with microtubules in middle stage oocytesand that this association is necessary for the translo-cation of Vgl mRNA to the vegetal hemisphere. Theanchoring of Vgl message at the cortex requires theinvolvement of cortical actin microfilaments. Once VglmRNA is anchored at the vegetal cortex, its distri-bution is not susceptible to disruption by nocodazole orcolchicine. In fact, oocytes that have only partiallylocalized Vgl mRNA do not release the localized RNAwhen the oocytes are cultured with microtubule inhibi-tors, even though further localization is prevented (datanot shown). Previously, we noted that the anchoringprocess seemed to be saturated when excess, exogenousVgl mRNA was injected into oocytes, even though thetranslocation process continued to function (Yisraeliand Melton, 1988). This observation is consistent withthe model presented here.

The use of inhibitors that affect specific cytoskeletalelements does not allow us to conclude that Vgl mRNAor mRNPs interact directly with microtubules duringtranslocation or microfilaments during anchoring. It iscertainly possible that microtubules and microfilamentsact indirectly through contact with other cytoplasmic'components. Nonetheless, these experiments do dem-onstrate a requirement for intact microtubules andfilaments to move and anchor Vgl mRNA.

An association between mRNAs and cytoskeletal

296 /. K. Yisraeli, S. Sokol and D. A. Melton

elements has been noted in a number of different celltypes (Lenk et al. 1977; Cervera et al. 1981; Jeffery,1984). In general, these reports have found that thebulk of mRNA (or, in some cases, polyribosomes) isnot washed away from the cytoskeleton in detergentextracts. The data presented here, however, as well asthose of Pondel and King (1988), suggest that VglmRNA is a member of a very small group of mRNAs(less than 5% of the total poly (A)+ RNA) that isassociated with the subcellular fraction of stage VIoocytes and is insoluble in detergent. Furthermore, ourresults demonstrate that Vgl mRNA can be releasedfrom this association by incubation with cytochalasin B,implying a role for microfilaments in this process. Thediscrepancy between the findings with oocytes and withother cells may reflect the specialized nature of oocytes,which need to spatially organize developmentally im-portant molecules before cleavage begins.

The involvement of microtubules in intracellulartransport has become an area of intensive study. Longknown for their role in ciliary movement (Warner,1979), microtubule-based motors are also responsiblefor both anterograde and retrograde transport in axons(Okabe and Hirokawa, 1989), chromosome segregationduring mitosis (Mitchison, 1989), pigment granule ag-gregation in chromatophores (McNiven and Porter,1984), and general organelle transport in such disparateorganisms as amoeba (Koonce and Schliwa, 1986) andsea urchins (Pryer et al. 1986). In Drosophila oocytesintact microtubules are required for cytoplasmicstreaming (Gutzeit, 1986).

The data presented here provide evidence for micro-tubule-mediated RNA transport. How RNA in general,and Vgl in particular, might move along a microtubuleis not understood. Making the assumption that associ-ation of Vgl mRNA with microtubules is not ratelimiting, we can roughly estimate that the rate ofmovement of Vgl mRNA in late stage III oocytes is onthe order of 100 nm per day. Clearly, the rates reportedfor fast axonal transport of proteins and organellesalong microtubules are several orders of magnitudefaster than that estimated for Vgl mRNA (50-400 mmper day; Lasek et al. 1984). On the other hand, slowaxonal transport rates, which are two to three orders ofmagnitude slower than fast transport, are similar to thehypothesized rate for Vgl mRNA movement, althoughthe molecular basis for this kind of transport is stillunclear (Lasek et al. 1984; Vale, 1987). It is interestingto note that an RNA transport system similar to the onewe observe in oocytes may exist in neurons. Usingcultured hippocampal neurons pulse-labeled in vitrowith [3H]uridine, Davis et al. (1987) observed RNAmigration specifically into dendrites, which can beidentified by the presence of the microtubule-associatedprotein, MAP2. Detergent extraction does not removethe RNA from the dendrites, implying an associationwith the cytoskeleton, and the rate of RNA transport isapproximately the same order of magnitude,400-500^m per day, as seen with Vgl. Recently, in situhybridization has revealed that this transport of RNAmay not be random; MAP2, but not tubulin, mRNA is

found specifically localized in dendrites in the develop-ing brain (Garner et al. 1988).

Movement along microtubule tracks has been sugges-ted as a possible model for the cortical rotation thatoccurs in fertilized eggs before first cleavage (Elinsonand Rowning, 1988). The microtubule array in fertilizedeggs, however, is quite different from that seen inoocytes. In eggs, the microtubule array is much denser,and the density of the array is probably important forproviding the necessary force. In addition, the array ineggs forms transiently in an oriented fashion runningalong the cortex. It is important to note that none of theevidence presented here proves that the microtubules inoocytes are providing the motive force for the translo-cation process. It is formally possible, for instance, thatthe microtubules might function as 'highways', facilitat-ing movement as a result of some other force, perhapseven passive diffusion along the tubule.

The animal-vegetal axis of oocytes is determinedwell before the localization of Vgl mRNA, perhaps asearly as the primordial germ cell (Al-Mukhtar andWebb, 1971; Heasman et al. 1984). Now that some ofthe structural elements involved in localizing RNA havebeen elucidated, the question of how the polarity of theoocyte is interpreted becomes better defined. Severallines of evidence seem to implicate additional, specificfactors in the process of spatially organizing the oocyte.Three mRNAs that are localized to the animal hemi-sphere in late stage oocytes were isolated by the samedifferential screening procedure that yielded Vgl(Rebagliati et al. 1985). These mRNAs are initiallydistributed homogeneously throughout young oocytesand appear to undergo localization within more or lessthe same narrow period of oogenesis as does Vglmessage (O'Keefe et al. 1989). The movement ofdifferent RNAs in opposite directions at the same time,coupled with the ubiquitous nature of the microtubulearray in stage III/IV oocytes, suggests that the direc-tionality of the translocation process is achievedthrough the use of additional factors. In addition,although all of the information necessary for thespecific localization of Vgl mRNA resides in the nakedRNA itself (Yisraeli and Melton, 1988), it is hard toimagine how microtubules could recognize these ex-acting sequences without the intervention of factorsspecific for the RNA. The search for these factorsshould be greatly aided by the identification of ex-acting sequences necessary for localization.

We would like to thank M. Klymkowsky for the anti fi-tubulin antibody and a preprint of the whole-mount immuno-histochemistry procedure. This work was supported by agrant from the National Institutes of Health (D.A.M.) and anNRSA fellowship (J.K.Y.).

References

AKAM, M. (1987). The molecular basis for metameric pattern in theDrosophila embryo. Development 101, 1-22.

AL-MUKHTAR, K. K. AND WEBB, A. C. (1971). An ultrastructuralstudy of primordial germ cells, oogonia, and early oocytes inXenopus laevis. J. Embryo!, exp. Morph. 26, 195-217.

The localization of Vgl mRNA 297

ALLEN, R. D., WEISS, D. C , HAYDEN, J. H., BROWN, D. T.,

FUJIWAKE, H. AND SIMPSON, M. (1985). Gliding movement of andbidirectional transport along single native microtubules fromsquid axoplasm: evidence for an active role of microtubules incytoplasmic transport. J. Cell Biol, 100, 1736-1752.

CERVERA, M., DREYFUSS, G. AND PENMAN, S. (1981). Messenger

RNA is translated when associated with the cytoskeletalframework in normal and VSV-infected HeLa cells. Cell 23,113-120.

C O L E M A N , A . , MORSER, J . , L A N E , C , B E S L E Y , J . , W Y L I E , C . A N DVALLE, G. (1981). Fate of secretory proteins trapped in oocytesof Xenopus laevis by disruption of the cytoskeleton or byimbalanced subunit synthesis. J. Cell Biol. 91, 770-780.

COOPER, J. A. (1987). Effects of cytochalasin and phalloidin onactin. J. Cell Biol. 105, 1473-1478.

DALE, L., MATTHEWS, G., TABE, L. AND COLEMAN, A. (1989). A

developmental expression of the protein product of Vgl, alocalized maternal mRNA in the frog Xenopus laevis. Embo J. 8,1057-1065.

DAVIS, L., BANKER, G. A. AND STEWARD, O. (1987). Selective

dendritic transport of RNA in hippocampal neurons in culture.Nature, Lond. 330, 477-479.

DENT, J. A. AND KLYMKOWSKY, M. W. (1988). Wholemountanalyses of cytoskeletal reorganization and function duringoogenesis and early embryogenesis in Xenopus. In The CellBiology of Fertilization (Shatten, H. and Shatten, G. eds)(Orlando: Academic Press).

DENT, J. A., POLSON, A. G. AND KLYMKOWSKY, M. W. (1989). A

whole mount immunocytochemical analysis of the expression ofthe intermediate filament protein vimentin in Xenopus.Development 105, 61-74.

DRIEVER, W. AND NUSSLEIN-VOLLARD, C. (1988). The bicoidprotein determines position in the Drosophila embryo in aconcentration-dependent manner. Cell 54, 95-104.

DUMONT, J. N. (1972). Oogenesis in Xenopus laevis (Daudin) I.Stages of oocyte development in laboratory maintained animals.J. Morph. 136, 153-180.

DUSTIN, P. (1978). Microtubules. (Berlin: Springer-Verlag).EUNSON, R. P. AND ROWNINC, B. (1988). A transient array of

parallel microtubules in frog eggs: potential tracks for acytoplasmic rotation that specifies the dorso-ventral axis. DeviBiol. 128, 185-197.

FRANKE, W. W., RATHKE, P. C , SEIB, E., TRENDELENBURO, M. F.,

OSBORN, M. AND WEBER, K. (1976). Distribution and mode ofarrangement of microfilamentous structures and actin in thecortex of the amphibian oocyte. Cytobiologie 14, 111-130.

GARNER, C. C , TUCKER, R. P. AND MATUS, A. (1988). Selective

localization of messenger RNA for cytoskeletal protein MAP2 indendrites. Nature, Lond. 336, 674-677.

GEUENS, G. M. A., NUYDENS, R. M., WILLEBRADS, R. E., VAN DE

VEIVE, DRAOONETTI, C. H., MAREEL, M. M. K. AND

DEBRABANDER, M. J. (1985). Effects of tubulozole on themicrotubule system of cells in culture and in vivo. Cancer Res.45, 733-742.

GURDON, J. (1968). Changes in somatic cell nucleii inserted intogrowing and maturing amphibian oocytes. J. Embryol. exp.Morph. 20, 401-414.

GUTZEIT, H. (1986). The role of microtubules in the differentiationof ovarian follicles during vitellogenesis in Drosophila. Roux'sArchiv. devl Biol. 195, 173-181.

HAUSEN, P., WANG, Y. H., DREYER, C. AND STICK, R. (1985).

Distribution of nuclear proteins during maturation of theXenopus oocyte. J. Embryol. exp. Morph. 89, 17-34.

HEASMAN, J., QUARMBY, J. AND WYUE, C. C. (1984). The

mitochondrial cloud of Xenopus oocytes: the source of germinalgranule material. Devi Biol. 105, 458-489.

HOLWILL, S., HEASMAN, J., CRAWLEY, C. R. AND WYLIE, C. C.

(1987). Axis and germ line deficiencies caused by u. v. irradiationof Xenopus oocytes cultured in vitro. Development 100, 735-743.

JEFFERY, W. R. (1984). Spatial distribution of messenger RNA inthe cytoskeletal framework of ascidian eggs. Devi Biol. 103,482-492.

JEFFERY, W. R. AND MEIER, S. (1983). A yellow crescent

cytoskeletal domain in ascidian eggs and its role in earlydevelopment. Devi Biol. 96, 125-143.

KIMELMAN, D. AND KJRSCHNER, M. (1987). Synergistic induction ofmesoderm by FGF and TGF/3 and the identification of an mRNAcoding for FGF in the early Xenopus embryo. Cell 51, 369-377.

KIMELMAN, D., KIRSCHNEB, M. AND SCHERSON, T. (1987). The

events of the midblastula transition in Xenopus are regulated bychanges in the cell cycle. Cell 48, 399-407.

KOONCE, M. P. AND SCHUWA, M. (1986). Reactivationof organellemovements along the cytoskeletal framework of a giantfreshwater amoeba. J. Cell Biol. 103, 605-612.

KRIEG, P. AND MELTON, D. A. (1985). Developmental regulation ofa gastrula-specific gene injected into fertilized Xenopus eggs.EMBO J. 4, 3463-3471.

LASEK, R. J., GARNER, J. A. AND BRADY, S. T. (1984). Axonal

transport of cytoplasmic matrix. J. Cell Biol. 99, 212s-221s.LAWRENCE, J. B. AND SINGER, R. H. (1986). Intracellular

localization of messenger RNAs for cytoskeletal proteins. Cell45, 407-415.

LENK, R., RANSOM, L., KAUFMANN, Y. AND PENMAN, S. (1977). A

cytoskeletal structure with associated polyribosomes obtainedfrom HeLa cells. Cell 10, 67-78.

MACDONALD, P. M. AND STRUHL, G. (1988). cis-acting sequencesresponsible for localizing bicoid mRNA at the anterior pole ofDrosophila embryos. Nature, Lond. 336, 595-598.

McNrvEN, M. A. AND PORTER, K. R. (1984). Chromatophores-models for studying cytomatrix translocations. J. Cell Biol. 99,152-158.

MELTON, D. A. (1987). Translocation of a localized maternalmRNA to the vegetal pole of Xenopus oocytes. Nature, Lond.328, 80-82.

MrrcMSON, T. J. (1989). Mitosis: basic concepts. Current Opinionin Cell Biology 1, 67-74.

NAKAMURA, O., TAKASAKI, H. AND MIZOHATA, T. (1970).

Differentiation during cleavage in Xenopus laevis. I. Acquisitionof self-differentiation capacity of the dorsal marginal zone. Proc.Japan Acad. 46, 694-699.

NIEUWKOOP, P. D. (1969). The formation of mesoderm inurodelean amphibians. I. Induction by the endoderm. WilhelmRoux' Arch. EntwMech. Org. 162, 341-373.

NOSSLEIN-VOLHARD, C , FROHNHOFER, H. G. AND LEHMANN, R.(1987). Determination of anteroposterior polarity in Drosophila.Science 238, 1675-1681.

O'KEEFE, H., KINTNER, C , YISRAEU, J. AND MELTON, D. A.(1989). The use of in situ hybridization to study the localizationof maternal mRNAs during Xenopus oogenesis. In In SituHybridization and the Study of Development and Differentiation.Society for Experimental Biology, vol. 43, (Harris, N. andWilkenson, D. eds) (Cambridge: Cambridge University Press),(in press).

OKABE, S. AND HIROKAWA, N. (1989). Ajtonal transport. CurrentOpinion in Cell Biology 1, 91-97.

PALECEK, J., HABROVA, V., NEDVIDEK, J. AND ROMANOVSKY, A.(1985). Dynamics of tubulin structures in Xenopus laevisoogenesis. J. Embryol. exp. Morph. 87, 75-86.

PONDEL, M. AND KING, M. L. (1988). Localized maternal mRNArelated to transforming growth factor B mRNA is concentrated ina cytokeratin-enriched fraction from Xenopus oocytes. Proc.natn. Acad. Sci. U.S.A. 85, 7612-7616.

PRYER, N. K., WADSWORTH, P. AND SALMON, E. D. (1986).Polarized microtubule gliding and particle saltations produced bysoluble factors from sea urchin eggs and embryos. Cell Motil. 6,537-548.

REBAGLIATI, M. R., WEEKS, D. L., HARVEY, R. P. AND MELTON, D.A. (1985). Identification and cloning of localized maternal RNAsfrom Xenopus eggs. Cell 42, 769-777.

ROSA, F., ROBERTS, A. B., DANIELPOUR, C , DART, L. L., SPORN,M. B. AND DAWID, I. B. (1988). Mesoderm induction inamphibians: the role of TGF-/J2-like factors. Science 239,783-785.

SCHUWA, M., VAN BLERKOM, J. AND PORTER, K. R. (1981).Stabilization of the cytoplasmic gTound substance in detergent-opened cells and a structural and biochemical analysis of itscomposition. Proc. natn. Acad. Sci. U.S.A. 78, 4329-4333.

298 J. K. Yisraeli, S. Sokol and D. A. Melton

SMITH, J. C. (1989). Mesoderm induction and mesoderm-inducingfactors in early amphibian development. Development 105,665-677.

SMITH, J. C , YAQOOB, M. AND SYMES, K. (1988). Purification,

partial charcterization and biological effects of the XTCmesoderm-inducing factor. Development 103, 591-600.

TANNAHILL, D. AND MELTON, D. A. (1989). Localized synthesis ofthe Vgl protein during early Xenopus development.Development 106, 775-785.

VALE, R. D. (1987). Intracellular transport using microtubule-basedmotors. Ann. Rev. Cell Biol. 3, 347-378.

VINCENT, J.-P. AND GERHART, J. C. (1987). Subcortical rotation inXenopus eggs: An early step in embryonic axis specification.Devi Biol. 113, 484-500.

WALLACE, R. A. AND JARED, D. W. (1976). Protein incorporationby isolated amphibian oocytes. V. specificity for vitellogeninincorporation. J. Cell Biol. 69, 345-351.

WALLACE, R. A., MISULOVIN, Z. AND WILEY, H. S. (1980). Growth

of anuran oocytes in serum-supplemented medium. Reprod.Nutr. Develop. 20, 699-708.

WARNER, F. D. (1979). Cilia and flagella: microtubule sliding and

regulated motion. In Microtubules (K. Roberts and J. S. Hyams,eds) (London: Academic Press, Inc.).

WEEKS, D. L. AND MELTON, D. A. (1987). A maternal mRNAlocalized to the vegetal hemisphere in Xenopus eggs codes for agrowth factor related to TGF-j5. Cell 51, 861-867.

WILSON, L. AND BRYAN, J. (1974). Biochemical andpharmacological properties of microtubules. Adv. Cell Mol. Biol.3, 22-72.

WYLIE, C. C , BROWN, C , GODSAVE, S. F., QUARMBY, J. AND

HEASMAN, J. (1985). The cytoskeleton of Xenopus oocytes and itsrole in development. J. Embryol. exp. Morph. 89, 1-15.

YAHARA, I., HARADA, F., SEKITA, S., YOSHIHIRA, K. AND NATORI, S.

(1982). Correlation between effects of 24 different cytochalasinson cellular structures and cellular events and those on actin invitro. J. Cell Biol. 92, 69-78.

YISRAEU, J. K. AND MELTON, D. A. (1988). The maternal mRNAVgl is correctly localized following injection into Xenopusoocytes. Nature, Lond. 336, 592-595.

{Accepted 20 November 1989)