Multiple isoforms of the high molecular weight microtubule associated protein XMAP215 are expressed during development in Xenopus.

We have cloned and sequenced cDNAs encoding two isoforms of XMAP215, a high molecular weight microtubule-associated protein identified in Xenopus eggs. XMAP215 is approximately 80% identical in amino acid sequence to the product of ch-TOG, a cDNA that is over expressed in certain human tumors [Charrasse et al., 1995: Eur J Biochem 234:406-413]. Northern and Western blots demonstrated that XMAP215 is expressed throughout development, from oogenesis to tadpole. We identified two XMAP215 transcripts differing only in the presence of a 108-bp sequence encoding a 36 amino acid insert. RT-PCR revealed that the transcripts encoding these two isoforms are expressed at distinct times during development: a transcript containing the insert (encoding XMAP215(M)) is expressed during oogenesis and is present through gastrulation. The second transcript (encoding XMAP215(Z)) lacks the 108-bp insert and is expressed from gastrulation onward. In situ hybridization demonstrated that XMAP215 transcripts are localized to the ectoderm of early embryos and in the developing nervous system during later development. These results suggest that XMAP215 plays important roles in at least two phases of development: (1) regulating the assembly of MTs during the rapid cell divisions after fertilization, and (2) regulating MT assembly during the development of the nervous system.

XMAP230 is required for normal spindle assembly in vivo and in vitro.

XMAP230 is a high molecular mass microtubule-associated protein isolated from Xenopus oocytes and eggs, and has been recently shown to be a homolog of mammalian MAP4. Confocal immunofluorescence microscopy revealed that XMAP230 is associated with microtubules throughout the cell cycle of early Xenopus embryos. During interphase XMAP230 is associated with the radial arrays of microtubules and midbodies remaining from the previous division. During mitosis, XMAP230 is associated with both astral microtubules and microtubules of the central spindle. Microinjection of affinity-purified anti-XMAP230 antibody into blastomeres severely disrupted the assembly of mitotic spindles during the rapid cleavage cycles of early development. Both monopolar half spindles and bipolar spindles were assembled from XMAP230-depleted extracts in vitro. However, spindles assembled in XMAP230-depleted extracts exhibited a reduction in spindle width, reduced microtubule density, chromosome loss, and reduced acetylation of spindle MTs. Similar defects were observed in the spindles assembled in XMAP230-depleted extracts that had been cycled through interphase. Depletion of XMAP230 had no effect on the pole-to-pole length of spindles, and depletion of XMAP230 from both interphase and M-phase extracts had no effect on the rate of microtubule elongation. From these results, we conclude that XMAP230 plays an important role in normal spindle assembly, primarily by acting to stabilize spindle microtubules, and that the observed defects in spindle assembly may result from enhanced microtubule dynamics in XMAP230-depleted extracts.

Microtubules in Xenopus oocytes are oriented with their minus-ends towards the cortex.

Despite lacking centrosomes, stage VI Xenopus oocytes contain extensive networks of cytoplasmic microtubules (MTs). To gain additional insight into the factors regulating MT organization during oogenesis, we have used electron microscopy and "hook decoration" to examine the distribution and orientation of MTs in Xenopus oocytes. A limited survey of two "undecorated" stage VI oocytes revealed 218 MTs in images covering approximately 2,500 microm(2), and indicated that the MT number density of the animal cytoplasm was greater than that of the vegetal cytoplasm. Examination of five "decorated" stage VI oocytes (three animal and five vegetal hemispheres) revealed 653 MTs. Of these, 76% could be scored as having exclusively counterclockwise (CCW) or clockwise (CW) hooks. In the animal hemispheres, 93% of the scored MTs exhibited CCW hooks when viewed from the direction of the cortex, indicating that most MTs were oriented with their minus-ends out. MT orientation appeared relatively uniform throughout the animal cytoplasm: more than 90% of the scored MTs in the cortical (90%), subcortical (96%), or perinuclear (98%) cytoplasm were oriented with their minus-ends out. In the vegetal hemispheres, approximately 80% of the scored MTs exhibited CCW hooks, and thus were oriented with their minus-ends out; 96% of the scored MTs in stage III oocytes were oriented minus-end out. These observations support a model in which the cortex plays a significant role in MT nucleation and organization in Xenopus oocytes, and have significant implications for the MT-dependent transport and localization of cytoplasmic organelles and RNAs during oogenesis.

Phosphorylation by CDK1 regulates XMAP215 function in vitro.

XMAP215, a microtubule-associated protein isolated from Xenopus eggs, promotes microtubule assembly dynamics in an end-specific manner: addition of XMAP215 to purified porcine tubulin increases both elongation and shortening rates at microtubule plus ends, with minimal effects at minus ends. Previous results indicated that XMAP215 is phosphorylated during M phase, suggesting that its activity may be regulated by cell cycle phosphorylation. To test this hypothesis, we used video-enhanced DIC microscopy to examine the effects of XMAP215 phosphorylated by CDK1 on the assembly of purified tubulin. XMAP215 incubated with ATP at 30 degrees C for 10- 20 min in the absence of CDK1 exhibited a 4.1-fold increase in plus end elongation rate compared to purified tubulin. Elongation was promoted to a lesser degree (2.4-fold) by phosphorylated XMAP215. In contrast, XMAP215 phosphorylation did not alter the approximately 3-fold increase in shortening rate. XMAP215 binding to taxol microtubules was also not changed by phosphorylation. To further investigate mechanisms responsible for the faster microtubule shortening rate observed with XMAP215, we made microtubules with segments assembled prior to XMAP215 addition (proximal segments) and segments assembled in the presence of XMAP215 (distal segments). In 9 of 10 microtubules, the distal segment shortened faster (distal = 60.7 microm/min; proximal = 37.5 microm/min), suggesting that MTs assembled in the presence of XMAP215 have an altered lattice that results in subsequent faster shortening.

Confocal microscopy and 3-D reconstruction of the cytoskeleton of Xenopus oocytes.

Xenopus oocytes contain a complex cytoskeleton composed of three filament systems: (1) microtubules, composed of tubulin and at least three different microtubule-associated proteins (XMAPs); (2) microfilaments composed of actin and associated proteins; and (3) intermediate filaments, composed of keratins. For the past several years, we have used confocal immunofluorescence microscopy to characterize the organization of the oocyte cytoskeleton throughout the course of oogenesis. Together with computer-assisted reconstruction of the oocyte in three dimensions, confocal microscopy gives an unprecedented view of the assembly and reorganization of the cytoskeleton during oocyte growth and differentiation. Results of these studies, combined with the effects of cytoskeletal inhibitors, suggest that organization of the cytoskeleton in Xenopus oocytes is dependent upon a hierarchy of interactions between microtubules, microfilaments, and keratin filaments. This article presents a gallery of confocal images and 3-D reconstructions depicting the assembly and organization of the oocyte cytoskeleton during stages 0-VI of oogenesis, a discussion of the mechanisms that might regulate cytoskeletal organization during oogenesis, and speculates on the potential roles of the oocyte cytoskeleton during oogenesis and axis formation.

XMAP230 is required for the organization of cortical microtubules and patterning of the dorsoventral axis in fertilized Xenopus eggs.

The dorsoventral axis of Xenopus embryos is specified by a rotation of the egg cortex relative to the underlying yolky cytoplasm. This cortical rotation, which occurs during the first cell cycle after fertilization, is dependent upon an array of parallel microtubules in the subcortical cytoplasm. We have used confocal immunofluorescent microscopy and microinjection of affinity-purified anti-XMAP230 antibody to address the role of XMAP230, one of three high-molecular-weight microtubule-associated proteins (MAPs) in Xenopus eggs, in the assembly and organization of the cortical microtubule array and specification of the dorsoventral axis. Confocal immunofluorescence microscopy revealed that XMAP230 was associated with cortical microtubules shortly after their appearance in the subcortical cytoplasm. XMAP230 staining became more prominent as microtubules were aligned and bundled during the cortical rotation. Loss of XMAP230 appeared to precede disassembly of cortical microtubules at the end of the first cell cycle. Deeper within the cytoplasm, XMAP230 was associated with microtubules early in the assembly of the sperm aster. However, later in the first cell cycle, XMAP230 was associated with microtubules (MTs) of the first mitotic spindle, spindle asters, and the cortical MTs, but not with microtubule remnants of the sperm aster. Microinjection of anti-XMAP230 antibody locally disrupted the assembly and organization of microtubules in the cortex of activated or fertilized eggs and resulted in defects in the dorsoventral patterning of embryos. These results indicate that the assembly and/or organization of cortical microtubules in fertilized Xenopus eggs and subsequent specification of the dorsoventral axis are dependent upon XMAP230.

Tau interacts with src-family non-receptor tyrosine kinases.

Tau and other microtubule-associated proteins promote the assembly and stabilization of neuronal microtubules. While each microtubule-associated protein has distinct properties, their in vivo roles remain largely unknown. Tau is important in neurite outgrowth and axonal development. Recently, we showed that the amino-terminal region of tau, which is not involved in microtubule interactions, is important in NGF induced neurite outgrowth in PC12 cells. Here we report that a proline rich sequence in the amino terminus of tau interacts with the SH3 domains of fyn and src non-receptor tyrosine kinases. Tau and fyn were co-immunoprecipitated from human neuroblastoma cells and co-localization of tau and fyn was visualized in co-transfected NIH3T3 cells. Co-transfection of tau and fyn also resulted in an alteration in NIH3T3 cell morphology, consistent with an in vivo interaction. Fyn-dependent tyrosine phosphorylation of tau occurred in transfected cells and tyrosine phosphorylated tau was identified in human neuroblastoma cells as well. Our data suggest that tau is involved in signal transduction pathways. An interaction between tau and fyn may serve as a mechanism by which extracellular signals influence the spatial distribution of microtubules. The tyrosine phosphorylation of tau by fyn may also have a role in neuropathogenesis, as fyn is upregulated in Alzheimer's disease.

XMAP230 is required for the assembly and organization of acetylated microtubules and spindles in Xenopus oocytes and eggs.

We used affinity-purified polyclonal antibodies to characterize the distribution and function of XMAP230, a heat-stable microtubule-associated protein isolated from Xenopus eggs, during oogenesis. Immunoblots revealed that XMAP230 was present throughout oogenesis and early development, but was most abundant in late stage oocytes, eggs, and early embryos. Immunofluorescence microscopy revealed that XMAP230 was associated with microtubules in oogonia, post-mitotic stage 0 oocytes, early stage I oocytes, and during stage IV-VI of oogenesis. However, staining of microtubules by anti-XMAP230 was not detectable during late stage I through stage III. In stage VI oocytes, anti-XMAP230 stained a large subset of microtubules that were also stained with monoclonal antibodies specific for acetylated (&agr ;)-tubulin. During oocyte maturation, XMAP230 was associated with the transient microtubule array that serves as the precursor of the first meiotic spindle, as well as both first and second meiotic spindles. The extensive array of cytoplasmic microtubules present throughout maturation was not detectably stained by anti-XMAP230. Microinjection of anti-XMAP230 locally disrupted the organization and acetylation of microtubules in stage VI oocytes, and reduced the re-acetylation of microtubules during recovery from cold-induced microtubule disassembly. Subsequent maturation of oocytes injected with anti-XMAP230 resulted in defects in the assembly of the transient microtubules array and first meiotic spindle. These observations suggest that XMAP230 is required for the stabilization and organization of cytoplasmic and spindle microtubules in Xenopus oocytes and eggs.

The TOGp protein is a new human microtubule-associated protein homologous to the Xenopus XMAP215.

We have recently identified a 6,449 bp cDNA, termed colonic, hepatic tumor over-expressed gene (ch-TOG), that is highly expressed in human tumors and brain. Its single open reading frame encodes a putative 218,000 Da polypeptide, TOGp. Antibodies generated against a bacterially expressed TOGp fragment specifically recognize a 218, 000 Da polypeptide in two human cell lines and in brain. Immunofluorescence microscopy using affinity-purified TOGp antibodies revealed that the distribution of TOGp was dependent upon the cell cycle. During interphase, TOGp was found concentrated in the perinuclear cytoplasm, where it co-localized with ER markers. In contrast anti-TOGp antibodies stained centrosomes and spindles in mitotic cells. TOGp co-sedimented with taxol-stabilized microtubules in vitro. Moreover, a TOGp enriched fraction promotes microtubule assembly both in solution and from nucleation centers. Finally, sequence comparison and immunologic cross-reaction suggest that TOGp is homologous to XMAP215, a previously described microtubule associated protein (MAP) from Xenopus eggs. These results suggest that TOGp is a MAP and that TOGp/XMAP215 may be necessary for microtubules rearrangements and spindle assembly in rapidly dividing cells.

The organization and animal-vegetal asymmetry of cytokeratin filaments in stage VI Xenopus oocytes is dependent upon F-actin and microtubules.

Confocal immunofluorescence microscopy with anti-cytokeratin antibodies revealed a continuous and polarized network of cytokeratin (CK) filaments in the cortex of stage VI Xenopus oocytes. In the animal cortex, CK filaments formed a dense meshwork that both was thicker and exhibited a finer mesh than the network of CK filaments previously observed in the vegetal cortex (Klymkowsky et al., 1987). CK filaments first appeared in association with germinal vesicle (GV) and mitochondrial mass (MM) of oocytes in early mid stage I, indicating that CK filaments are the last of the three cytoskeletal networks to be assembled. By late stage I, CK filaments formed complex networks surrounding the GV, surrounding and penetrating the MM, and linking these networks to a meshwork of CK filaments in the oocyte cortex. During stage III-early IV, CK filaments formed a highly interconnected, apparently unpolarized, radial array linking the perinuclear and cortical CK filament networks. Polarization of the CK filament network was observed during mid stage IV-stage V, as first the animal, then the vegetal CK filament networks adopted the organization characteristic of stage VI oocytes. Treatment of stage VI oocytes with cytochalasin B disrupted the organization of both cortical and cytoplasmic CK filaments, releasing CK filaments from the oocyte cortex and inducing formation of numerous cytoplasmic CK filament aggregates. CB also disrupted the organization of cytoplasmic microtubules (MTs) in stage VI oocytes. Disassembly of oocyte MTs with nocodazole resulted in loss of the characteristic A-V polarity of the cortical CK filament network. In contrast, disruption of cytoplasmic CK filaments by microinjection of anti-CK antibodies had no apparent effect on cytoplasmic or MT organization. We propose a model in which the organization and polarization of the cortical network of CK filaments in stage VI Xenopus oocytes are dependent upon a hierarchy of interactions with actin filaments and microtubules.

Microinjection of anti-alpha-tubulin antibody (DM1A) inhibits progesterone-induced meiotic maturation and deranges the microtubule array in follicle-enclosed oocytes of the frog, Rana pipiens.

Microinjection of anti-alpha-tubulin (Dm1A) inhibited progesterone-induced meiotic maturation in large follicle-enclosed oocytes of the frog, Rana pipiens. DM1A (46 nl; 10 mg/ml) injection significantly increased the ED50 value for progesterone as determined by germinal vesicle dissolution (GVD) bioassay. By contrast, low doses of microinjected DM1A (46 nl; 2.5 mg/ml), anti-actin (clone KJ43A), anti-cytokeratin (C-11), anti-intermediate filament antibody (IFA), generic IgG (46 nl; 20 mg/ml) or sodium azide (46 nl; 1 mg/ml), an antibody preservative, were without inhibitory effect in this bioassay. Microinjected, affinity-purified DM1A (46 nl; 7.5 mg/ml) was also inhibitory, but preabsorption with pure tubulin prior to injection significantly reduced the inhibitory effect. DM1A injection had no effect on centrifugation-induced germinal vesicle migration (GVM). Previous work indicated that drugs (e.g. demecolcine and nocodazole), which destabilise microtubules, enhance both centrifugation-induced GVM and progesterone-induced GVD in Rana oocytes. Taking these results together, it is suggested that DM1A injection may have differential effects on microtubules in this cell. Thus, while the majority of microtubules were apparently depolymerised by DM1A (46 nl; 10 mg/ml) injection, a small subpopulation appeared to be stabilised as bundles. Confocal immunofluorescence microscopy of follicle-enclosed oocytes after DM1A injection revealed a major loss of microtubules throughout the cell; however, apparent sparse bundles of microtubules arranged in an approximately 600 microns shell were associated with the injectate region 24 h post-injection. By contrast, control follicle-enclosed oocytes topically labelled with DM1A post-fixation had extensive microtubule arrays similar to those previously reported in Xenopus oocytes. Intracellular recording after DM1A injection and progesterone treatment yielded an intermediate membrane potential (Vm = -31.8 mV) compared with control (immature) DM1A-injected cells (Vm = -44.7 mV) or potassium balanced salt solution (KBS)-injected cells matured with progesterone (Vm = -13.9 mV). These results suggest that DM1A injection does not completely inhibit electrophysiological changes initiated by progesterone. Working hypotheses are proposed that suggest a role for microtubules in the action of progesterone which normally lifts the prophase I block in the Rana follicle-enclosed oocyte.

Xenopus nonmuscle myosin heavy chain isoforms have different subcellular localizations and enzymatic activities.

There are two isoforms of the vertebrate nonmuscle myosin heavy chain, MHC-A and MHC-B, that are encoded by two separate genes. We compared the enzymatic activities as well as the subcellular localizations of these isoforms in Xenopus cells. MHC-A and MHC-B were purified from cells by immunoprecipitation with isoform-specific peptide antibodies followed by elution with their cognate peptides. Using an in vitro motility assay, we found that the velocity of movement of actin filaments by MHC-A was 3.3-fold faster than that by MHC-B. Likewise, the Vmax of the actin-activated Mg(2+)-ATPase activity of MHC-A was 2.6-fold greater than that of MHC-B. Immunofluorescence microscopy demonstrated distinct localizations for MHC-A and MHC-B. In interphase cells, MHC-B was present in the cell cortex and diffusely arranged in the cytoplasm. In highly polarized, rapidly migrating interphase cells, the lamellipodium was dramatically enriched for MHC-B suggesting a possible involvement of MHC-B based contractions in leading edge extension and/or retraction. In contrast, MHC-A was absent from the cell periphery and was arranged in a fibrillar staining pattern in the cytoplasm. The two myosin heavy chain isoforms also had distinct localizations throughout mitosis. During prophase, the MHC-B redistributed to the nuclear membrane, and then resumed its interphase localization by metaphase. MHC-A, while diffuse within the cytoplasm at all stages of mitosis, also localized to the mitotic spindle in two different cultured cell lines as well as in Xenopus blastomeres. During telophase both isoforms colocalized to the contractile ring. The different subcellular localizations of MHC-A and MHC-B, together with the data demonstrating that these myosins have markedly different enzymatic activities, strongly suggests that they have different functions.

Microtubule organization, acetylation, and nucleation in Xenopus laevis oocytes: II. A developmental transition in microtubule organization during early diplotene.

Confocal immunofluorescence microscopy of ovaries from juvenile frogs revealed changes in the organization, acetylation, and nucleation, of microtubules (MTs), and redistribution of gamma-tubulin (gamma-TB), during early oogenesis in Xenopus laevis. Interphase oogonia contained sparse, radially organized, MT arrays and prominent centrosomes, Acetylated MTs were not commonly found in oogonia. In contrast, small (approximately 12-25 microns), postmitotic (stage 0) oocytes contained dense, highly polarized, MT networks that exhibited little or no evidence of radical organization. Examination of stage 0 oocytes stained with antibodies to gamma-TB, in conjunction with assays of MT nucleation activity, revealed that stage 0 oocytes do contain active centrosomes. In addition, stage 0 oocytes contained numerous acetylated MTs, suggesting that arrest in meiotic prophase is accompanied by MT stabilization. Early stage I oocytes (diameters from approximately 35-50 microns) exhibited a rounded morphology and contained a dispersed, apparently disordered, MT array with a substantial population of acetylated MTs. Examination of stage I oocytes stained with gamma-TB antibodies revealed that this centrosomal protein was present in multiple cytoplasmic foci which did not function as MTOCs following cold-induced MT disassembly. The results presented indicate that the maternal centrosome is inactivated during early stage I, roughly coincident with the onset of the diplotene stage of meiotic prophase and prior to assembly of the mitochondrial mass. Our observations place constraints on the role of MTs and the maternal centrosome during specification of the animal-vegetal axis of Xenopus oocytes and raise questions regarding the mechanisms by which MT assembly and organization are regulated during oocyte differentiation.

F-actin is required for spindle anchoring and rotation in Xenopus oocytes: a re-examination of the effects of cytochalasin B on oocyte maturation.

We used confocal immunofluorescence microscopy to examine spindle migration, morphology and orientation during the maturation of Xenopus oocytes, in the presence or absence of cytochalasin B (CB), an inhibitor of actin assembly. Treatment with CB during maturation (10-50 micrograms/ml beginning 0-3 h prior to addition of progesterone) disrupted the normal organisation of the novel MTOC and transient microtubule array (MTOC-TMA complex) that serves as the immediate precursor of the first meiotic spindle, suggesting that F-actin plays an important role in the assembly or maintenance of this complex. However, CB treatment did not block translocation of the MTOC-TMA complex to the oocyte cortex, suggesting that MTOC-TMA translocation is not dependent on an actin-based mechanism. Bipolar spindles were observed in CB-treated oocytes fixed during both M1 and M2. However, rotation of the M1 and M2 spindles into an orientation orthogonal to the oocyte surface was inhibited by CB. Rhodamine-phalloidin revealed a concentration of F-actin at the site of M1 spindle attachment, further suggesting that cortical actin is required for anchoring and rotation of the meiotic spindles. Finally, the incidence of M1 monasters was significantly increased in CB-treated oocytes, suggesting that interactions between the nascent M1 spindle and cortex are dependent on F-actin.

XMAP from Xenopus eggs promotes rapid plus end assembly of microtubules and rapid microtubule polymer turnover.

We have used video-enhanced DIC microscopy to examine the effects of XMAP, a Mr 215,000 microtubule-associated protein from Xenopus eggs (Gard, D.L., and M. W. Kirschner. 1987. J. Cell Biol. 105:2203-2215), on the dynamic instability of microtubules nucleated from axoneme fragments in vitro. Our results indicate that XMAP substantially alters the parameters of microtubule assembly at plus ends. Specifically, addition of 0.2 microM XMAP resulted in (a) 7-10-fold increase in elongation velocity, (b) approximately threefold increase in shortening velocity, and (c) near elimination of rescue (the switch from rapid shortening to elongation). Thus, addition of XMAP resulted in the assembly of longer, but more dynamic, microtubules from the plus ends of axonemes which upon catastrophe disassembled back to the axoneme nucleation site. In agreement with previous observations (Gard, D.L., and M. W. Kirschner. 1987. J. Cell Biol. 105:2203-2215), the effects of XMAP on the minus end were much less dramatic, with only a 1.5-3-fold increase in elongation velocity. These results indicate that XMAP, unlike brain MAPs, promotes both polymer assembly and turnover, and suggests that the interaction of XMAP with tubulin and the function of XMAP in vivo may differ from previously characterized MAPs.

Confocal microscopy of F-actin distribution in Xenopus oocytes.

We have used rhodamine-conjugated phalloidin and confocal microscopy to examine the organisation of filamentous actin (F-actin) during oogenesis in Xenopus laevis. F-actin was restricted to a thin shell in the cortex of oogonia and post-mitotic oocytes less than 35 microns in diameter. In oocytes with diameters of 35-50 microns, F-actin was observed in three cellular domains: in the cortex, in the germinal vesicle (GV) and in a network of cytoplasmic cables. Initially, actin cables were sparsely distributed in the cytoplasm, with no evidence of discrete organising centres. In larger stage I oocytes, a dense network of actin cables extended throughout the cytoplasm, linking the GV and mitochondrial mass to the cortical actin shell. Apart from the F-actin associated with the mitochondrial mass, no evidence of a polarised distribution of F-actin was apparent in stage I oocytes. F-actin was observed also in the cortex and the GV of stage VI oocytes, and a network of cytoplasmic cables surrounded the GV. Cytoplasmic actin cables extended from the GV to the animal cortex, and formed a three-dimensional network surrounding clusters of yolk platelets in the vegetal cytoplasm. Finally, disruption of F-actin in stage VI oocytes with cytochalasin resulted in distortion and apparent rotation of the GV in the animal hemisphere, suggesting that actin plays a role in maintaining the polarised organisation of amphibian oocytes.

Gamma-tubulin is asymmetrically distributed in the cortex of Xenopus oocytes.

Stage VI Xenopus oocytes contain an extensive network of cytoplasmic microtubules (MTs), with no evidence of a functional centrosome. Recently, Stearns et al. (1991) demonstrated that Xenopus eggs contain a substantial pool of the centrosomal protein gamma-tubulin (gamma-Tb). For this report, I have used confocal immunofluorescence microscopy to examine the distribution of gamma-Tb during the later stages of oogenesis in Xenopus laevis. gamma-Tb was apparent surrounding the germinal vesicle (GV) of stage VI oocytes, consistent with previous results suggesting that the GV serves as an microtubule organizing center in later oogenesis. Surprisingly, gamma-Tb was also concentrated in the cortex of stage VI oocytes, and the distribution of cortical gamma-Tb was polarized along the animal-vegetal (A-V) axis. In the vegetal cortex, gamma-Tb was observed in brightly stained foci, often organized into short linear arrays. In the animal hemisphere, gamma-Tb was more evenly distributed as small cortical foci. Dual immunofluorescence microscopy revealed that gamma-Tb in the vegetal hemisphere was associated with MTs in the cortical cytoplasm. The distribution of gamma-Tb was not significantly affected by either cold or nocodazole, but was partially disrupted by cytochalasin B. gamma-Tb thus may serve as a link between the oocyte MT network and cortical actin. Finally, polarization of the distribution of cortical gamma-Tb temporally coincides with formation of the A-V axis and polarization of the oocyte MT cytoskeleton during stage IV of oogenesis. These observations raise a number of questions regarding the organization and orientation of MTs during Xenopus oogenesis and the role of gamma-Tb in the polarization of the oocyte cytoskeleton during A-V axis formation.