Identification and preliminary characterization of a

0 downloads 0 Views 209KB Size Report
Microtubules are composed principally of tubulin and microtubule-associated proteins (MAPs), as determined from studies of animal cells (Dustin, 1984; Olmsted ...
Journal of Cell Science 105,891-901 (1993) Printed in Great Britain © The Company of Biologists Limited 1993

891

Identification and preliminary characterization of a 65 kDa higher-plant microtubule-associated protein Jiang Chang-Jie* and Seiji Sonobe Department of Biology, Faculty of Science, Osaka University, Toyonaka, Osaka 560, Japan *Author for correspondence at present address: Biotechnology Research Team, Regional Environment Division, National Institute for Environmental Studies, Onogawa 16-2, Tsukuba 305, Japan

SUMMARY Microtubules in plant cells, as in animal cells, are dynamic structures. However, our lack of knowledge about the constituents of microtubules in plant cells has prevented us from understanding the mechanisms that control microtubule dynamics. To characterize some of these constituents, a cytoplasmic extract was prepared from evacuolated protoplasts (miniprotoplasts) of tobacco BY-2 cells, and microtubules were assembled in the presence of taxol and disassembled by cold treatment in the presence of Ca 2+ and a high concentration of NaCl. SDS-PAGE analysis of triplecycled microtubule protein revealed the presence of 120 kDa, 110 kDa and a group of 60-65 kDa polypeptides in addition to tubulin. Since these polypeptides had copolymerized with tubulin, through the three cycles of assembly and disassembly, and they bundle microtubules, we tentatively identified the three polypeptides as microtubule-associated proteins (MAPs). To characterize these factors further, triple-cycled

microtubule protein was fractionated by Mono-Q anionexchange chromatography and the microtubulebundling activity of each fraction was examined. Fractions having microtubule-bundling activity contained only the 65 kDa MAP, an indication that the 65 kDa MAP is responsible for the bundling of microtubules. Purified 65 kDa MAP formed cross-bridge structures between adjacent microtubules in vitro. Polyclonal antibodies were raised in mice against the 65 kDa MAP. Immunofluorescence microscopy revealed that the 65 kDa MAP colocalized with microtubules in BY-2 cells throughout the cell cycle. Western blotting analysis of extracts from several species of plants suggested that the 65 kDa MAP and/or related peptides are widely distributed in the plant kingdom.

INTRODUCTION

et al., 1991; Baskin and Cande, 1990; Gunning, 1982; Gunning and Wick, 1985). The events associated with these changes in the arrangement of microtubules and their possible functions have been well documented. However, the molecular mechanisms by which microtubules can assume such diverse orientation and participate in so many different functions in living plant cells remain unknown. Our lack of knowledge of the constituents of microtubules in plant cells has prevented us from understanding these mechanisms. Microtubules are composed principally of tubulin and microtubule-associated proteins (MAPs), as determined from studies of animal cells (Dustin, 1984; Olmsted, 1986). Tubulin, a highly conserved heterodimer of α and β subunits, self-assembles to form the microtubule filaments. Plant tubulin is remarkably similar to animal tubulin in many respects (Fosket, 1989). For example, antibodies against brain tubulin cross-react with plant tubulin; tubulin from plant and animal sources can be copolymerized to form heterogeneous microtubules both in vitro (Yadav and

In plant cells, as in animal cells, microtubules are dynamic structures (Zhang et al., 1990) and are involved in a variety of cellular functions. Four typical discrete arrays of microtubules within cells appear and disappear sequentially during the progression of the cell cycle and play distinct roles in the growth and morphogenesis of plant cells (Seagull and Heath, 1980; Lloyd, 1987; Seagull, 1989; Staiger and Lloyd, 1991). In interphase cells, microtubules lie parallel to and close to the plasma membrane. They are known as cortical microtubules, and appear to control the orientation of cellulose microfibrils in the cell wall (Gunning and Hardham, 1982; Giddings and Staehelin, 1991). In premitotic cells, microtubules form the preprophase band, which is considered to determine the site and plane of formation of the cell plate (Gunning, 1982; Wick, 1991; Lloyd, 1991). In dividing cells, microtubules appear as the spindle apparatus and the phragmoplast, and play key roles in both karyokinesis and cytokinesis (Tiwari et al., 1984; Lambert

Key words: cross-bridge (microtubules), microtubule-associated proteins (MAPs), microtubule bundles, microtubule protein, miniprotoplast, tobacco BY-2 cells

892

C.-J. Jiang and S. Sonobe

Filner, 1983) and in vivo (Zhang et al., 1990; Vantard et al., 1990; Asada et al., 1991); and the amino acid sequences of tubulin deduced from nucleotide sequences of plant tubulin genes are 79-87% identical to those of mammalian tubulin (Silflow et al., 1987). Plant tubulin exhibits differences in its sensitivity to various drugs, as compared to animal tubulin (Morejohn and Fosket, 1984a,b). MAPs affect assembly, stability, morphology and therefore function of microtubules (Olmsted, 1986; Tucker, 1990; Cyr, 1991). MAPs also involve in interaction of microtubules with other cytoskeletal components and cellular organelles (Gelfand and Bershadsky, 1991). The functional diversity of microtubules seems to be ascribed to the diversity of MAPs (Cyr, 1991). In spite of the potentially important roles of MAPs in the mediation of microtubule-dependent cellular functions, almost all our knowledge of MAPs comes from studies of animal cells, and very little information is available about plant MAPs at present. Cyr and Palevitz (1989) identified microtubule-binding proteins from an extract of carrot cultured cells using taxol-stabilized neuronal microtubules as a high-affinity substrate and Maekawa et al. (1990) reported identification of a green algal (Dichotomosiphon tuberosus) MAP (90 kDa) by virtue of its cross-reaction with antibodies against bovine adrenal 190 kDa MAP. More recently, Vantard et al. (1991) reported the taxol-mediated assembly of microtubules in cytosolic extracts of maize cultured cells and the usefulness of such assembled microtubules as the native matrix for isolation of putative plant MAPs. However, no distinct protein(s) have been purified and characterized. It should be pointed out that neuronal microtubules can only be used to identify MAPs that have binding sites common to both neuronal and plant microtubules and, therefore, they cannot be used to identify MAPs that associate specifically with plant microtubules. We previously reported success in the assembly of microtubules in a cytoplasmic extract from evacuolated protoplasts (miniprotoplasts) of tobacco BY-2 cells (Jiang et al., 1992). In this paper, we report (1) the isolation of microtubule protein (tubulin and MAPs) from the cytoplasmic extract by assembly and disassembly; (2) purification and partial characterization of a higher-plant MAP (65 kDa MAP) that can form cross-bridge structures between adjacent microtubules in vitro and (3) the localization of the 65 kDa MAP in BY2 cells using antibodies raised against this protein. MATERIALS AND METHODS Plant materials Tobacco BY-2 cells (Nicotiana tabacum ‘Bright Yellow-2’) were cultured in suspension in modified Linsmaier and Skoog’s medium (LS medium), supplemented with 200 mg l−1 KH2PO4, 3% sucrose and 0.2 mg l−1 2,4-dichlorophenoxyacetic acid (2,4-D), pH 5.8, at 26°C in the dark (Nagata et al., 1981). They were subcultured every 7 days by transfer of 2 ml of culture into 95 ml of fresh medium. Five-day-old cells, i.e. cells in the logarithmic phase of growth, were used. Arabidopsis Fl-3 cells were kindly supplied by Mr T. Yokoi in our laboratory. These cells were derived from A. thaliana strain Fl-3 (Arabidopsis Information Service, Frankfurt, Germany) by Dr Davis (Ohio State Univ., Ohio, USA). They were cultured in

suspension in a newly designed medium; namely, LS medium supplemented with 200 mg l−1 KH2PO4, 3% sucrose, 0.2 mg l−1 2,4D, 10 mg l−1 thiamine-HCl, 1 mg l−1 pyridoxine-HCl, 1 mg l−1 nicotinic acid, and 100 mg l−1 myo-inositol, pH 5.8, at 26°C in the dark. They were subcultured every 7 days by the transfer of 5 ml of culture into 95 ml of fresh medium (T. Yokoi, personal communication). Pea (Pisum sativun L. cv. Alaska), azuki bean (Vigna angularis Ohwi et Ohashi cv. Takarawase) and maize (Zea mays L. cv. Honey Bantam) seeds were sown on moist vermiculite and germinated under white light at 27°C. Seedlings were grown under the same conditions for 5 to 7 days. Barley (Hordeum vulgare L. cv. Fuji-nijou) and rice (Oryza sativa L. cv. Koshihikari) seeds were germinated and the seedlings were grown on moist filter paper in Petri dishes at 27°C for 3 and 5 days, respectively. The seeds of pea, azuki bean and maize were purchased from Watanabe Saishujo Co. (Miyagi, Japan) and barley seeds were supplied by Dr K. Takeda (Okayama Univ., Kurashiki, Japan). Thalli of Dichotomosiphon tuberosus were kindly supplied by Prof. R. Nagai (Osaka Univ., Osaka, Japan). Samples were originally collected in a paddy field in Okinawa, Japan, and they have been cultured unialgally in her laboratory, since that time (Maekawa et al., 1986). Carrots were purchased from a local vegetable market.

Isolation of miniprotoplasts Miniprotoplasts were isolated as described previously (Sonobe, 1990; Jiang et al., 1992). BY-2 cells of 300-350 g fresh weight, were treated with a solution of wall-digesting enzymes (1% Cellulase Onozuka RS (Yakult Pharmaceutical Co., Takarazuka, Hyogo, Japan) and 0.1% Pectolyase Y-23 (Seishin Pharmaceutical Co., Nagareyama, Chiba, Japan) in 0.45 M mannitol, pH 5.5) for isolation of protoplasts. The isolated protoplasts were suspended in 400 ml of a solution of 30% Percoll (Pharmacia LKB Biotechnology AB, Uppsala, Sweden) in 0.6 M mannitol, pH 5.8, and centrifuged at 11,000 g for 60 min. Miniprotoplasts were collected from the layer near the bottom of each centrifuge tube and washed with a cooled (4°C) 0.6 M solution of mannitol by centrifugation at 450 g for 3 min.

Preparation of cytoplasmic extract The cytoplasmic extract was prepared as described previously (Jiang et al., 1992). In brief, washed miniprotoplasts (8 ml in packed volume) were suspended in one third volume of an icecold extraction solution (50 mM Pipes containing 2 mM MgCl2, 5 mM EGTA, 25 µg ml−1 leupeptin, 3 mM dithiothreitol, 1 mM PMSF, 0.5 mM GTP, pH 7.0) and homogenized by passage through the 23-gauge needle of a microsyringe. The homogenate was centrifuged at 248,000 g for 40 min at 2°C. The supernatant was used as cytoplasmic extract.

Isolation of microtubule protein (MTP) by assembly and disassembly To obtain a large amount of MTP and to ensure the successful passage of multiple cycles of assembly and disassembly of microtubules, the method of Collins and Vallee (1987) and Collins (1991) was employed with modifications. The extract (4 ml) was incubated with 20 µM taxol and 0.2% Triton X-100 for 25 min at 30°C and then centrifuged at 50,000 g for 20 min at 25°C. The resultant pellet was suspended in 400 µl of an ice-cold depolymerizing solution (50 mM Pipes containing 1 mM MgCl2, 3 mM CaCl2, 0.3 M NaCl, 10 µg ml −1 leupeptin, 0.5 mM dithiothreitol, 0.5 mM PMSF, pH 7.0) and placed on ice for 40 min to depolymerize the microtubules. Then it was centrifuged at 176,000 g for

A plant microtubule-associated protein 20 min at 2°C. The supernatant was termed the once-cycled MTP (1c-MTP) preparation. The 1c-MTP preparation was supplemented with EGTA, taxol and GTP to final concentrations of 1 mM, 20 µM and 0.5 mM, respectively, and incubated at 30°C for 20 min. During the incubation, assembly of microtubules in the 1c-MTP preparation was confirmed by electron microscopy with negative-staining procedures. The incubated 1c-MTP preparation was then diluted 3-fold with extraction solution to lower the concentration of NaCl to 0.1 M. After further incubation for 5 min, the preparation was centrifuged at 30,000 g for 15 min at 25°C and the resultant pellet was suspended in 200 µl of depolymerizing solution. By repeating above procedures, we obtained preparations designated 2c-MTP and 3c-MTP. After addition of EGTA to a final concentration of 1 mM, the 3c-MTP preparation (200 µl) was used immediately or after storage at −70°C for two weeks or less in the following studies.

Purification of the microtubule-bundling factor The 3c-MTP preparation was diluted 3-fold with an elution buffer (20 mM Pipes, 1 mM MgCl 2, 1 mM EGTA, 10 µg ml−1 leupeptin, 0.5 mM PMSF, 0.5 mM dithiothreitol, pH 7.0) and applied to a fast protein liquid chromatography system (FPLC, Pharmacia LKB Biotechnology, Uppsala, Sweden) equipped with a Mono-Q column that has been pre-equilibrated with 0.1 M NaCl in the elution buffer and was eluted with a linear gradient of NaCl (from 0.1 M to 0.5 M, total 10 ml) in the same buffer. Fractions of 0.5 ml were collected and dialyzed against the elution buffer. All procedures were performed at 4°C. To examine the microtubulebundling activities in the fractions, 10 µl of each fraction was mixed with an appropriate volume of microtubule solution (see below) and incubated for 5 min at 30°C. The mixtures were negatively stained and examined with an electron microscope. The microtubules used here were assembled from pure tubulin that had been recovered from fractions that eluted at 0.4-0.5 M NaCl from the same column and had been desalted and concentrated with a microdialyzer (Molecut II, Millipore). This tubulin was capable of polymerizing into microtubules even after storage for six months at −70°C.

Preparation of antibodies Considering that only a small amount of immunogen (65 kDa protein) was available, we employed an intrasplenic immunization method (Nilsson et al., 1987) with slight modifications. About 10 µg of purified 65 kDa protein that was responsible for microtubule-bundling activity was adsorbed to a 10 mm2 piece of nitrocellulose membrane by repeated direct application and air-drying. The nitrocellulose membrane with adsorbed 65 kDa protein was crushed in 400 µl of PBS with a pestle and mortar, and injected into exposed spleens of two 8-week-old male Balb/c mice. The operation was performed by the procedure described by Nilsson et al. (1987). Mice were given two more intrasplenic injections of 65 kDa protein at 14-day intervals. They were killed 4 days after the last injection and their sera were collected by cardiac puncture with subsequent clotting and centrifugation. For some experiments, antibodies were band-affinity purified by the method of Olmsted (1981) with exceptions that: (1) peroxidase-conjugated secondary antibodies (see below) were used instead of iodinated Protein A; and (2) antibodies were eluted from the appropriate strip of nitrocellulose membrane (2-3 mm wide, 10 cm long) by sandwiching 20 µl of 0.2 M glycine-HCl, pH 2.8, between the strip and a piece of parafilm; the eluate was immediately neutralized by 5-fold dilution with Tris-buffered saline (100 mM Tris-HCl, 150 mM NaCl, pH 7.5) containing 1% BSA instead of NaOH. The purified antibodies were directly used for western blotting analysis.

893

Immunolocalization of the 65 kDa protein in BY-2 cells The method of Katsuta et al. (1990) was employed. In brief, the walls of tobacco BY-2 cells were partially digested for 2.5-3.0 min with an enzyme solution that contained 0.5% cellulase Onozuka RS, 0.05% pectolyase Y-23, 5 mM EGTA, 25 µg ml −1 leupeptin, 0.5 mM PMSF, 0.25 M mannitol, pH 5.5, and the cells were fixed in 3.7% formaldehyde in extraction solution that had been supplemented with 0.05% Nonidet P-40 (NP-40). After two washes with PBS-Tween, the cells were incubated successively with mouse antibodies against the 65 kDa protein (diluted 1:50) and, FITC-conjugated rabbit antibodies (Seikagaku Kogyo Co., Ltd. Tokyo, Japan) against mouse IgG (diluted 1:30) for 1 h each. After each incubation, cells were washed twice with PBS-Tween. For double labeling, the fixed cells were successively incubated with a mixture of antibodies against the 65 kDa protein and rabbit antibodies against mung bean tubulin (kindly provided by Dr Mizuno, Osaka Univ., Osaka, Japan; diluted 1:200), rhodamineconjugated goat antibodies against rabbit IgG (TAGO, Inc., Burlingame, USA; diluted 1:100), and FITC-conjugated rabbit antibodies against mouse IgG for 1 h each. All these procedures were performed at room temperature. As controls, samples were incubated with preimmune serum instead of antibodies against the 65 kDa protein. Stained samples were examined under a microscope equipped with epifluorescence illumination (Olympus BHSRFK) and photographed on Kodak T-MAX 400 film.

Electrophoresis and immunoblotting Electrophoresis was performed by the method of Laemmli (1970) using SDS-polyacrylamide (7.5% (w/v) acrylamide) slab gels. Proteins in gels were transferred electrophoretically to nitrocellulose membrane as described by Towbin et al. (1979). Non-specific binding sites were blocked by incubation with 3% skim milk in PBS for 1 h. Antibodies against the 65 kDa protein were applied as the first antibody (diluted 1:200), and horseradish peroxidaseconjugated antibodies raised in goat against mouse IgG (Bio-Rad Lab., Richmond, CA, USA) were used for the second antibody. All procedures were performed at room temperature. For immunoblotting analysis, BY-2 cells, Fl-3 cells, the root tips (10-15 mm) of pea, azuki bean, maize, rice and barley seedlings, thalli of Dichotomosiphon tuberosus, and phloem parenchyma cells of carrot roots were rinsed in distilled water and crushed in liquid nitrogen with a pestle and mortar. Sample buffer (pH 6.8) containing 250 mM Tris-HCl, 4% SDS, 10% (v/v) 2mercaptoethanol, 20% (v/v) glycerol and an appropriate amount of bromophenol blue was added to each frozen powder, and each mixture was boiled for 3 min. After centrifugation at 1,200 g for 20 min, the supernatants were analyzed.

Electron microscopy An aliquot of the assembled microtubules was centrifuged at 30,000 g for 20 min at 25°C and the resultant pellet was fixed with 2.5% glutaraldehyde and then with 2% osmium tetroxide, dehydrated with an ethanol series and finally embedded in Spurr’s resin. Sections were cut, stained with uranyl acetate and lead citrate, and then observed with an electron microscope (JEM-100s; Jeol, Tokyo). Assembly and bundling of microtubules were also examined by a negative-staining method using 2% uranyl acetate.

Quantitation of protein Concentrations of proteins were determined by the method of Bradford (1976) using the Bio-Rad dye reagent (Bio-Rad Lab.) and bovine serum albumin as the standard.

894

C.-J. Jiang and S. Sonobe

RESULTS Preparation of miniprotoplasts and cytoplasmic extract Typically, 8 ml of miniprotoplasts were obtained from 300350 g fresh weight of cells. The yield of miniprotoplasts was dependent on the condition of the isolated protoplasts. Incomplete digestion of cell walls drastically reduced the yield of miniprotoplasts. To obtain protoplasts in an optimal condition, cells were used on or before the fifth day of culture. The closer cells were to the stationary phase of growth, the less complete was cell wall digestion. We found later that addition of 1-2 mM MgCl2 to the Percoll solution can increase the yield of miniprotoplasts by about 20%. About 4 ml of cytoplasmic extract was obtained from 8 ml of packed miniprotoplasts. The protein concentration of the extract was up to 35-45 mg ml−1. High protein concentration was very important for the polymerization of tubulin into microtubules. In extracts below 25 mg ml−1, minimal assembly of microtubules occurred in the absence of taxol. Similarly, extracts with higher protein concentrations reduced the total yield of proteins. Isolation of MTP by assembly and disassembly procedures Microtubules were assembled in the presence of taxol, and they were disassembled by cold treatment in the presence of 3 mM CaCl 2 and 0.3 M NaCl (Collins and Vallee, 1987; Collins, 1991). It was possible to repeat three cycles of microtubule assembly and disassembly. Fig. 1 shows SDSPAGE of the cytoplasmic extract and the preparations of microtubule protein obtained at each stage of the three cycles of assembly and disassembly. In the trice-cycled microtubule protein (3c-MTP) preparation (lane 4 in Fig. 1), polypeptides with molecular masses of 120 kDa and 110 kDa, and a group of 60-65 kDa (for convenience designated as 65 kDa protein in this paper), were detected in addition to tubulin. These polypeptides had copurified with tubulin through the three cycles of assembly and disassembly; hence they seemed to be MAPs. About 3.6 mg of 1c-MTP, 400 µg of 2c-MTP and 240 µg of 3c-MTP were recovered from 4 ml of extract. Electron microscopy of negatively stained preparations revealed that MTP was assembled into microtubules even in the presence of 0.3 M NaCl (data not shown). No bundles of microtubules were formed under these conditions. However, large bundles of microtubules, visible even to the naked eye as cotton-wad-like material, occurred immediately after the MTP preparation was diluted 3-fold with extraction buffer, which reduced the concentration of NaCl to 0.1 M and presumably led MAP(s) to re-associated to microtubules (data not shown). A pellet of microtubule bundles formed from the 3c-MTP preparation was suspended in 0.3 M NaCl and 10 µM taxol, and the mixture was centrifuged at 55,000 g for 15 min at 25°C. The resultant supernatant (MAP fraction) was retained and the pellet (microtubules) suspended in extraction solution that contained 10 µM taxol. The MAPs and microtubule fraction were subjected to SDS-PAGE (Fig. 2). Almost all MAPs (lane 1) were released from microtubules

Fig. 1. Patterns after SDS-PAGE of preparation of microtubule protein (MTP) at different stages of isolation. Lanes: 1, cytoplasmic extract; 2, 1c-MTP; 3, 2c-MTP; 4, 3c-MTP (see text for definition); M, molecular mass standards (kDa). Gels were stained with Coomassie blue.

Fig. 2. Profiles after SDS-PAGE of MAPs and tubulin. A pellet of microtubule bundles, formed in the preparation designated 3cMTP, was suspended in extraction solution plus 0.3 M NaCl and then centrifuged. Lanes: 1, supernatant (MAPs); 2, pellet (tubulin); M, molecular mass standards (kDa). The tubulin was recovered as microtubules, as shown in Fig. 3.

(lane 2) by 0.3 M NaCl. The microtubules that released MAPs were negatively stained and observed by electron microscopy. As shown in Fig. 3, microtubules were scattered and no bundles were observed. Incubation of the MAP fraction with microtubules at a low NaCl concentration (0.1 M) again gave rise to large bundles of microtubules (data not shown). This observation indicates that the microtubulebundling factor(s) was recovered in the MAP fraction that had been solubilized by 0.3 M NaCl from microtubule bundles formed in the 3c-MTP preparation. Western blotting analysis of the MAP fraction using antibodies against mung bean tubulin identified the lower bands of around 55 kDa as tubulin (data not shown). Some microtubules seem depolymerized by suspending, although 10 µM taxol were added in the extraction solution.

A plant microtubule-associated protein

895

Fig. 3. Electron micrograph of microtubules that had released MAPs. The pellet obtained in the experiment described in the legend to Fig. 2 was suspended in the extraction solution that had been supplemented with 10 µM taxol and a 5 µl sample was applied to a Formvar-carbon-coated microgrid, washed with water, and negatively stained with 2% uranyl acetate. No bundles of microtubules were observed. Bar, 0.25 µm. Fig. 5. Electron micrographs of microtubule bundles reconstituted by the addition of purified 65 kDa protein to microtubules. Samples were negatively stained. Bars: 2 µm (A) and 0.5 µm (B).

found to have microtubule-bundling activity. These were pooled, concentrated and used in the following studies. SDS-PAGE revealed that these fractions contained the 65 kDa protein (Fig. 4, lane 1). The 120 kDa protein was eluted before the 65 kDa protein and the 110 kDa protein was eluted between the 65 kDa protein and tubulin. These proteins had no microtubule-bundling activity.

Fig. 4. Profile after SDS-PAGE of purified microtubule-bundling factor and tubulin. Proteins in the preparation designated 3c-MTP were fractionated by FPLC on a Mono-Q column. The fraction having microtubule-bundling activity contained only the 65 kDa protein (lane 1). Highly purified tubulin was obtained by this procedure (lane 2). M, molecular mass standards (kDa).

Purification of tubulin and the microtubulebundling factor To identify and purify the protein(s) that cross-bridge adjacent microtubules and cause microtubule bundling, the 3cMTP preparation was fractionated by Mono-Q anionexchange chromatography. Highly purified tubulin (Fig. 4, lane 2) was recovered from fractions that eluted at 0.45-0.5 M NaCl. The fractions were pooled, concentrated and desalted. The tubulin was polymerized into microtubules by incubation with 20 µM taxol and the polymerized microtubules used for the examination of microtubule-bundling activity of the other fraction. Fractions that eluted between 0.25-0.3 M NaCl were

Formation of cross-bridge structures between adjacent microtubules by the 65 kDa protein Bundles of microtubules formed by 65 kDa protein (Fig. 5) were investigated by thin-section electron microscopy (Fig. 6). Microtubules were arranged in parallel (Fig. 6B) and cross-bridges were frequently observed between adjacent microtubules in cross-sections (Fig. 6C, arrowheads). The length of the cross-bridges was about 10-12 nm, similar to that of the shorter cross-bridges observed in bundles formed directly in the cytosolic extract (Fig. 2b,d of Jiang et al., 1992). As can be seen in cross-section, cross-linked microtubules were associated laterally to form ‘sheets’ rather than three-dimensional bundles. Characterization of antibodies against the 65 kDa protein Polyclonal antibodies were raised in mice against the 65 kDa protein. Since only a small amount of 65 kDa protein was available as immunogen, we employed an intrasplenic immunization method (Nilsson et al., 1987, with slight modifications; see Materials and Methods). This method seems to be very useful for the preparation of antibodies against molecules that are difficult to isolate in large quantities. Western blotting analysis of cytoplasmic extracts (Fig. 7A, lane 1; B, lane 1), 3c-MTP (Fig. 7A, lane 2) and

896

C.-J. Jiang and S. Sonobe

Fig. 7. (A) Cytoplasmic extract (lane 1) and 3c-MTP (lane 2, 3) were western blotted using antibodies against the 65 kDa protein (lane 1, 2) and tubulin (lane 3), respectively. Antibodies against the 65 kDa protein were highly specific to the 65 kDa protein and showed no cross-reactivity with tubulin or with other polypeptides. (B) Cytoplasmic extract were western blotted using band-affinity-purified antibodies against the 65 kDa protein. Lanes: 1, antibodies not purified; 2-4, band-affinity-purified antibodies from first and second bands, third band and fourth band, counted downwards, respectively (cut at positions indicated by arrows). Every kind of band-affinity-purified antibody recognized all four bands in the cytoplasmic extract, indicating that these polypeptide bands are closely inter-related. Fig. 6. Ultrastructure of microtubule bundles (B,C) like the ones shown in Fig. 5. Microtubules were arranged parallel to each other (B) and cross-bridge structures were observed between adjacent microtubules in cross-sections of the microtubule bundles (arrowheads in C). Microtubules (A) that sedimented in the absence of the 65 kDa protein showed no ordered arrangement. Bars: 0.5 µm (A, B) and 0.1 µm (C).

whole-cell extracts of BY-2 cells (Fig. 10, lane 1) using antibodies against the 65 kDa protein indicated that the antibodies are highly specific to the 65 kDa protein and show no cross-reactivity to tubulin (Fig. 7A, lane 3) or other polypeptides. Absence in crude extracts of cross-reactive polypeptides with molecular masses higher than 65 kDa (Fig. 7A,B lane 1; Fig. 10, lane 1) indicates that the 65 kDa protein is not a product of the degradation of a polypeptide of higher molecular mass. It is unclear at present whether the 120 kDa and 110 kDa polypeptides are derived from polypeptides of higher molecular mass. The antibodies recognized the group of 60-65 kDa polypeptides (designated as 65 kDa protein in this paper), and typically four bands were distinguishable in the cytoplasmic extract. To examine whether or not these polypeptides are inter-related, antibodies were band-affinity puri-

fied from: (1) first and second bands; (2) third band; and (3) fourth band in the cytoplasmic extract, counted downwards, respectively (arrows in Fig. 7B, lane 1) by the method of Olmsted (1981). Western blotting analysis revealed that all of these purified antibodies recognized all the four bands in the cytoplasmic extract (Fig. 7B, lanes 24), indicating that these polypeptide bands are closely interrelated. The antibodies were examined for their ability to inhibit the microtubule-bundling activity of the 65 kDa protein. The antibodies were mixed with the 65 kDa protein and microtubules assembled from BY-2 cell tubulin at the same time or with bundles of microtubules like the ones shown in Fig. 5, and mixtures were incubated for 20 min. The incubated mixtures were negatively stained and observed with an electron microscope. The antibodies inhibited formation of microtubule bundles by the 65 kDa protein but had no effect on pre-formed bundles of microtubules (data not shown). Immunolocalization of the 65 kDa protein in BY-2 cells Tobacco BY-2 cells were subjected to immunofluorescence staining with the antibodies against the 65 kDa protein. Fig.

A plant microtubule-associated protein

897

Fig. 8. Cells immunostained for the 65 kDa protein (A) and cells immunostained for microtubules (B) and stained with pre-immune serum (C). Cortical microtubule-like arrays are seen in (A). The arrow indicates a phragmoplast. No distinct structures are seen in cells stained with pre-immune serum (C). Bar, 20 µm.

Fig. 9. Colocalization of the 65 kDa protein (B, D, F, H) with microtubules (A, C, E, G) throughout the cell cycle. (A, B) Cortical microtubules in interphase cells; (C, D) microtubules in the preprophase band; (E, F) mitotic spindle; (G, H) phragmoplast. Bar, 20 µm.

8a shows that the antibodies stained numerous filamentous structures that lay parallel to and close to cell cortices, suggesting the co-localization of the 65 kDa protein with cortical microtubules. Antibodies also stained other typical arrays of microtubules in plant cells. A phragmoplast structure is indicated by an arrow in Fig. 8A. No distinct staining patterns were seen with pre-immune serum (see Fig. 8C). Double staining with antibodies against the 65 kDa protein and tubulin revealed that the 65 kDa protein is co-

localized with microtubules in BY-2 cells throughout the cell cycle. Fig. 9 shows representative images of the colocalization of the 65 kDa protein (B, D, F, H) with microtubules (A, C, E, G) in the typical arrays of microtubules that appear sequentially during the cell cycle, i.e. the cortical microtubules (Fig. 9A, B) and the microtubules in the preprophase band (Fig. 9C, D), mitotic spindle (Fig. 9E, F) and phragmoplast (Fig. 9G, H). In cells treated with 10−4 M propyzamide for 2.5 h,

898

C.-J. Jiang and S. Sonobe Fig. 10. Western blotting analysis of extracts prepared from various plant tissues and cultured cells. Lanes 1-9, profile after SDS-PAGE of extracts of BY-2 cells, pea, azuki bean, Arabidopsis Fl-3 cells, maize root tips, rice root tips, barley root tips, Dichotosiphon tuberosus thalli and carrot roots, respectively; 1′-9′, western blots of lanes 1-9. In all extracts tested, polypeptide(s) that cross-reacted with antibodies against the 65 kDa protein was present. Mobilities of the polypeptides from rice root tips and Fl-3 cells differed from that of the 65 kDa protein from BY-2 cells. M, molecular mass standards (kDa.

almost all microtubules were disrupted (Akashi et al., 1988). The disruption of microtubules by propyzamide also abolished staining with antibodies against the 65 kDa protein (data not shown). Propyzamide-treated cells were washed twice with fresh medium and cultured in the absence of propyzamide for 15 min. Cells were sampled and double-stained for the 65 kDa protein and tubulin. The distinct patterns of staining with antibodies against the 65 kDa protein reappeared concomitantly with the reorganization of microtubules (data not shown). Distribution of proteins related to the 65 kDa protein in the plant kingdom To determine whether or not the 65 kDa protein or related polypeptides are widely distributed in the plant kingdom, whole-cell extracts were prepared from eight further plant species and subjected to western blotting procedures with antibodies against the 65 kDa protein. As shown in Fig. 10, polypeptide(s) that cross-reacted with the antibodies were present in all plants examined. The size of the polypeptide from pea, azuki bean, barley, maize, carrot and Dichoto mosiphon tuberosus was very close to that of the 65 kDa protein from BY-2 cells. A polypeptide related to the 65 kDa protein from rice was slightly smaller, and that from Fl-3 cultured cells was larger than the 65 kDa protein from BY-2 cells. In order to determine whether or not the 65 kDa protein shares some immunological homology with known neuronal MAPs, western blotting analysis was carried out using several commercially available antibodies against neuronal MAPs: (1) monoclonal antibody against bovine MAP2 (Sigma Chemical Co. Ltd., St. Louis, MO, USA); (2) monoclonal antibody against bovine tau protein (Chemicon International, Inc., Town, CA, USA); (3) rabitt polyclonal antibodies against bovine MAPs (Sigma Chemical Co. Ltd., St. Louis, MO, USA); (4) rabbit polyclonal antibodies against chick embryo brain tau (Sigma Chemical Co. Ltd., St. Louis, MO, USA); (5) rabbit polyclonal antibodies against a sythetic peptide of 24 amino acids coresponding to a tubulin-binding domain common to the bovine adrenal 190 kDa MAP and tau protein (AP-1 and AP-4; Aizawa et al. 1989; Maekawa et al. 1990; kindly provided by Dr H. Murofushi, University of Tokyo, Tokyo, Japan). All these antibodies failed to identify the 65 kDa protein in the 3cMTP (data not shown).

DISCUSSION Microtubules in plant cells are dynamic structures (Zhang et al., 1990). If we are to understand how they assume their functionally diverse arrangements in living plant cells, we must identify all the protein components that are involved in the dynamic processes that determine the organization of microtubules. Studies on animal cells have revealed that microtubules are composed of tubulin and microtubuleassociated proteins (MAPs; Dustin, 1984; Olmsted, 1986). Microtubule protein has been isolated from a variety of animal sources by procedures that involve the temperaturedependent assembly and disassembly of microtubules. In spite of many the efforts at adapting this procedures to the isolation of microtubule protein from higher plants, all previous attempts have encountered difficulties because the yield of assembled microtubules in plant extracts is too low for repeated assembly and disassembly, even when taxol is used (Morejonh and Fosket, 1982; Cyr and Palevitz, 1989; Vantard et al., 1991). Recent success in obtaining good yields of assembled microtubules in extracts prepared from evacuolated protoplasts (Jiang et al., 1992) prompted us to attempt to isolate higher-plant microtubule protein by the assembly and disassembly procedure. Initially, we intended to work without using taxol, which binds to polymerized tubulin (Parness and Horwitz, 1981) and might therefore reduce the yields of some MAPs. However, we soon found that the amounts of microtubules assembled in a small volume of extract (4 ml) in the absence of taxol were too small to allow repeated assembly and disassembly. To overcome this problem, we employed the method of Collins and Vallee (1987) and Collins (1991), a variant of the temperature-dependent assembly and disassembly procedure, in which the assembly of microtubules is enhanced by taxol. Microtubules treated with taxol are usually very resistant to depolymerization but, in the presence of multiple destabilizing factors they can be depolymerized into free tubulin (Collins, 1991). In the present study, a combination of low temperature, Ca2+ and a high concentration of NaCl (0.3 M) was used. Cold treatment and Ca2+ are well known to cause depolymerization of microtubules. Addition of 0.3 M NaCl to the depolymerizing buffer presumably dissociates MAPs from microtubules and thereby increases the rate of disassembly of microtubules. Almost all MAPs identified to date have been found to stabilize microtubules.

A plant microtubule-associated protein In the present study, we were able to repeat the cycles of assembly and disassembly up to three times. In the preparation of triple-cycled microtubule protein (Fig. 1), three polypeptides were detected in addition to the band of tubulin by SDS-PAGE. From our observations that these polypeptides copolymerized with tubulin and were enriched during the three cycles of assembly and disassembly (Fig. 1), and that they bundled microtubules, we concluded that the three polypeptides are higher plant MAPs. The 65 kDa protein appeared as multiple bands that are closely inter-related (Fig. 1 and Fig. 7A,B). This was reminiscent of the size-heterogeneity of tau protein, one of the MAPs in the mammalian brain, which appears on gel as a group of isoforms with molecular mass from 45 to 62 kDa (Cleveland et al., 1977). Heterogeneity of tau arises by alternative splicing of a single gene (Neve et al., 1986; Himmler, 1989; Kanai et al., 1989) and multiple phosphorylation states (Lindwall and Cole, 1984a,b; Kanai et al., 1989). The phosphorylation state of tau can affect its binding to microtubules. Alkaline phophatase treatment of tau resulted in a more rapid and more extensive polymerization of microtubules (Lindwall and Cole, 1984a). The multiple bands of 65 kDa protein might result in a similar way to those of tau protein. Naturally, further studies must be carried out before drawing the final conclusion, which is very important for understanding the mechanism by which the 65 kDa protein is involved in the regulation of microtubule dynamics in living cells. Another possible explanation for the multiple bands of 65 kDa protein could be proteolysis during the isolation of the cytoplasmic extract. But this does not seem very likely, because two kinds of protease inhibitors were added in the extraction solution and antibodies against the 65 kDa protein failed to identify smaller polypeptides (Fig. 7A,B). The identification and purification of the 65 kDa protein that forms cross-bridge structures between adjacent microtubules and causes bundling of microtubules in vitro may provide important clues to the mechanism of the spatial organization of microtubules in situ. Similar cross-bridge structures have been observed in a variety of plant cells, as mentioned in our previous paper (Jiang et al., 1992), and microtubules in plant cells are often found to be bundled together (see the Discussion in the paper by Cyr and Palevitz, 1989). In BY-2 cells, cross-bridge structures have been observed between adjacent microtubules in isolated phragmoplasts (Kakimoto and Shibaoka, 1988). Their lengths were similar to those of the cross-bridges formed by the 65 kDa protein in vitro (Fig. 6C). Our immunocytochemical study revealed that the 65 kDa protein is co-localized with microtubules in situ throughout the cell cycle. These results suggest the possibility that the 65 kDa protein is involved in the architectural organization of microtubules in living cells. The 65 kDa protein formed cross-bridges of similar length to the shorter type of cross-bridges observed in the bundles of microtubules that were formed in the crude extract (Jiang et al., 1992), but they did not form the longer type of cross-bridge. These observations suggest that the two types of cross-bridge are not composed of the same protein. However, the possibility that the 65 kDa protein forms two types of cross-bridge cannot be excluded and,

899

therefore, it should be examined. We note, in this context, that the tau protein has been reported to be able to stretch or contract by more than 300%, depending on its state of phosphorylation (Lichtenberg et al., 1988; Hagestedt et al., 1989). Possible explanations for the presence of two different types of cross-bridge are as follows: (1) each type of cross-bridge is composed of different protein species and the 65 kDa protein is responsible only for the short bridges; (2) the two types of cross-bridge are composed of the 65 kDa protein, the length of which varies depending on modifications of the protein; and (3) the 65 kDa protein is capable of participating in the formation of long bridges via interactions with itself or other proteins. If the antibodies against the 65 kDa protein can inhibit the formation of both types of cross-bridge in the extract, then the latter two explanations would seem to be more plausible. However, the concentration of the antibodies in the antisera was not high enough to allow us to test these possibilities. The addition of a large volume of antiserum to the extract reduced the concentration of tubulin in the extract, with a resultant decrease in the ability of tubulin to polymerize into microtubules. Preparation of a monoclonal antibody, which can be produced at quite high concentrations, against the 65 kDa protein may allow us to test our hypotheses. Cyr and Palevitz (1989) reported isolation of microtubule-binding proteins, from carrot cultured cells, that form cross-bridges between adjacent microtubules and cause bundling of microtubules. Two major polypeptides, of 129 kDa and 76 kDa, were responsible for the microtubule-bundling activities. These proteins seem to differ from the 65 kDa protein in two respects. First, they differ in terms of heat-stability. The 65 kDa protein is very heatsensitive; it is easily denatured and can be completely inactivated by heating at 70°C for 2 min (data not shown). Carrot microtubule-binding proteins, by contrast, retained 20% of their bundling activity after they had been heated at 90°C for 5 min. Second, the 65 kDa protein is an acidic protein, whereas carrot microtubule-binding proteins are basic proteins. The relationship between the 65 kDa protein and carrot microtubule-binding proteins remains to be clarified. Maekawa et al. (1990) identified a green algal MAP (90 kDa) that cross-reacted with antibodies against bovine adrenal 190 kDa MAP. The 90 kDa protein shares several properties with the 65 kDa protein, as follows: (1) partially purified 90 kDa protein has microtubule-bundling activity in vitro; (2) immunocytochemical studies revealed that the antigen recognized by the antibodies against bovine adrenal 190 kDa MAP was localized along microtubules in cells of Dichotomosiphon tuberosus; and (3) the 90 kDa protein bound to DEAE resin and was eluted between 0.18 and 0.39 M NaCl. The 90 kDa protein seems to differ from the 65 kDa protein in that: (1) the 90 kDa protein is heat-stable (90°C, 1 min); and (2) the 90 kDa protein was not recognized by the antibodies against the 65 kDa protein during our western blotting analysis (Fig. 10). Vantard et al. (1991) reported the presence, in cultured maize cells, of an 83 kDa protein that cross-reacted with antibodies against tau proteins from rat brain. Western blotting analysis in the present study failed to identify the 65 kDa protein with antibodies against mammalian neuronal

900

C.-J. Jiang and S. Sonobe

MAPs (data not shown). By contrast, western blotting analysis using antibodies against 65 kDa protein revealed the presence in maize root-tip cells of a polypeptide of 65 kDa that cross-reacted with antibodies against the 65 kDa protein (Fig. 10). These results indicate that the 65 kDa protein is immunologically distinct from neuronal MAPs and from the 83 kDa protein of maize. Finally, we would like to emphasize that the extract from evacuolated protoplasts can serve as excellent experimental material for biochemical studies of proteins from plant sources. Using this material, we succeeded for the first time in achieving the assembly of microtubules in a crude extract of plant cells in the absence of microtubule-stabilizing agents (Jiang et al., 1992) and in the isolation of plant microtubule protein by assembly and disassembly procedures (this study). Furthermore, we were able to purify tubulin and a plant MAP (65 kDa) that formed cross-bridge structures in vitro and co-localized with microtubules in situ. The results described in this paper indicate that the 65 kDa protein is a novel plant MAP, which is widely distributed in the plant kingdom even though its role in the regulation of microtubule arrays in vivo remains to be clarified. We thank Professor H. Shibaoka for his encouragement throughout this work and for his critical reading of the manuscript. We thank Dr S. Ogihara for helpful advice on the operation of the FPLC system. This work was supported in part by Grants-inAid for the Encouragement of Young Scientists (no. 03740346) and for Scientific Research (no. 03223211) from the Ministry of Education, Science and Culture, Japan.

REFERENCES Akashi, T., Izumi, K., Nagano, E., Enomoto, M., Mizuno, K. and Shibaoka, H. (1988). Effects of propyzamide on tobacco cell microtubules in vivo and in vitro. Plant Cell Physiol. 29, 1053-1062. Aizawa, H., Kawasaki, H., Murofushi, H., Kotani, S., Suzuki, K. and Sakai, H. (1989). A common amino acid sequence in 190 kDa microtubule-associated protein and tau for promotion of microtuuble assembly. J. Biol. Chem. 264, 5885-5890. Asada, T., Sonobe, S. and Shibaoka, H. (1991).Microtubule translocation in the cytokinetic apparatus of cultured tobacco cells. Nature 350, 238241. Baskin, T. I. and Cande, W. Z. (1990). The structure and function of the mitotic spindle in flowering plants. Annu. Rev. Plant Physiol. Plant Mol. Biol. 41, 277-315. Bradford, J. M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Anal. Biochem. 72, 248-251. Cleveland, D. W., Hwo, S. Y. and Kirschner, M. W. (1977). Physical and chemical properties of purified tau factor and role of microtubule assembly. J. Mol. Biol. 116, 227-247. Collins, C. A. (1991). Reversible assembly purification of taxol-treated microtubules. Meth. Enzymol. 196, 246-253. Collins, C. A. and Vallee, R. B. (1987). Temperature-dependent reversible assembly of taxol-treated microtubules. J. Cell Biol. 105, 2847-2854. Cyr, R. J. (1991). Microtubule-associated proteins in higher plants. In The Cytoskeletal Basis of Plant Growth and Form (ed. C. W. Lloyd), pp. 5767. Academic Press, London. Cyr, R. J. and Palevitz, B. A. (1989). Microtubule-binding proteins from carrot. Planta 177, 245-260. Dustin, P. (1984). Microtubules, 2nd edn. Springer Verlag, New York. Fosket, D. E. (1989). Cytoskeletal proteins and their genes in higher plants. In The Biochemistry of Plants, vol. 15 (ed. P. K. Stumpf), pp. 392-454. Academic Press, New York.

Gelfand, V. I. and Bershadsky, A. D. (1991). Mirotubule dynamics: mechanism, regulation, and function. Annu. Rev. Cell Biol. 7, 93-116. Giddings, T. H. and Staehelin, L. A. (1991). Microtubule-mediated control of microfibril deposition: a re-examination of the hypothesis. In TheCytoskeletal Basis of Plant Growth and Form (ed. C. W. Lloyd), pp. 85-99. Academic Press, London. Gunning, B. E. S. (1982). The cytokinetic apparatus: its development and spatial regulation. In TheCytoskeleton in Plant Growth and Development (ed. C. W. Lloyd), pp. 229-292. Academic press, London. Gunning, B. E. S. and Hardham, A. R. (1982). Microtubules. Annu. Rev. Plant Physiol. 33, 651-698. Gunning, B. E. S. and Wick, S. M. (1985). Preprophase bands, phragmoplasts, and spatial control of cytokinesis. J. Cell Sci. Suppl. 2 , 157-179. Hagestedt, T., Lichtenberg, B., Wille, H., Mandelkow, E.-M. and Mandelkow, E. (1989). Tau protein becomes long and stiff upon phosphorylation: correlation between paracrystalline structure and degree of phosphorylation. J. Cell Biol. 109, 1643-1651. Himmler, A. (1989). Structure of the bovine tau gene: Alternatively spliced transcripts generate a protein family. Mol. Cell. Biol. 9, 1389-1396. Jiang, C. J., Sonobe, S. and Shibaoka, H. (1992). Assembly of microtubules in a cytoplasmic extract of tobacco BY-2 miniprotoplasts in the absence of microtubule-stabilizing agents. Plant Cell Physiol. 33, 497-501. Kakimoto, T. and Shibaoka, H. (1988). Cytoskeletal ultrastructure of phragmoplast-nuclei complexes isolated from cultured tobacco cells. Protoplasma (Suppl. 2), 95-103. Kanai, Y., Takemura, R. Oshima, T., Mori, H., Ihara, Y., Yanagisawa, M, Masaki, T. and Hirokawa, N. (1989). Expression of multiple tau isoforms and microtubule bundle formation in fibroblasts transfected with a single tau cDNA. J. Cell Biol. 109, 1173-1184. Katsuta, J., Hashiguchi, Y. and Shibaoka, H. (1990). The role of cytoskeleton in positioning of the nucleus in premitotic tobacco BY-2 cells. J. Cell Sci. 95, 413-422. Laemmli, U. K.(1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680-685. Lambert, A. M., Vantard, M., Schmit, A. C. and Stoeckel, H. (1991). Mitosis in plants. In The Cytoskeletal Basis of Plant Growth and Form (ed. C. W. Lloyd), pp. 199-207. Academic Press, London. Lichtenberg, B., Mandelkow, E.-M., Hagestedt, T. and Mandelkow, E. (1988). Structure and elasticity of microtubule-associated protein tau. Nature 334, 359-362. Lindwall, G. and Cole, R. D. (1984a). Phosphorylation affects the ability of tau protein to promote microtubule assembly. J. Biol. Chem. 259, 53015305. Lindwall, G. and Cole, R. D. (1984b). The purification of tau protein and the occurrence of two phosphorylation states of tau in brain. J. Biol. Chem. 259, 12241-12245. Lloyd, C. W. (1987). The plant cytoskeleton: The impact of fluorescence microscopy. Annu. Rev. Plant Physiol. 38, 119-139. Lloyd, C. W. (1991). Cytoskeletal elements of the phragmosome establish the division plane in vasculated higher plant cells. In The Cytoskeletal Basis of Plant Growth and Form (ed. C. W. Lloyd), pp. 245-258. Academic Press, London. Maekawa, T., Ogihara, S., Murofushi, H. and Nagai, R. (1990). Green algal microtubule-associated protein with a molecular weight of 90 kDa which bundles microtubules. Protoplasma 158, 10-18. Maekawa, T., Tsutsui, I. and Nagai, R. (1986). Light-regulated translocation of cytoplasm in green alga Dichotomosiphon. Plant Cell Physiol. 27, 837-851. Morejohn, L. C. and Fosket, D. E. (1982). Higher plant tubulin identified by self-assembly into microtubules in vitro. Nature 297, 426-428. Morejohn, L. C. and Fosket, D. E. (1984a). Taxol-induced rose microtubule polymerization in vitro and its inhibition by colchicine. J. Cell Biol. 99, 141-147. Morejohn, L. C. and Fosket, D. E. (1984b). Inhibition of plant microtubule polymerization in vitro by the phosphoric amide herbicide amiprophosmethyl. Science 224, 874-876. Nagata, T., Okada, K., Takebe, I. and Matsui, C. (1981). Delivery of tobacco mosaic virus RNA into plant protoplasts mediated by reversephase evaporation vesicles (liposomes). Mol. Gen. Genet. 184, 161-165. Neve, R. L., Harris, P., Kosik, K. S., Kurnit, D. M. and Donlon, T. A. (1986). Identification of cDNA clones for the human microtubule-

A plant microtubule-associated protein associated protein, tau, and chromosomal localization of the gene for tau and microtubule-associated protein 2. Mol. Brain Res. 1, 271-280. Nilsson, B. O., Svalander, P. C. and Larsson, A. (1987). Immunization of mice and rabbits by intrasplenic deposition of nanogram quantities of protein attached to sepharose beads or nitrocellulose paper strips. J. Immunol. Meth. 99, 67-75. Olmsted, J. B. (1981). Affinity purification of antibodies from diazotized paper blots of heterogeneous protein samples. J. Biol. Chem. 256, 1195511957. Olmsted, J. B. (1986). Microtubule-associated proteins. Annu. Rev. Cell Biol. 2, 421-457. Parness, J. and Horwitz, S. B. (1981). Taxol binds to polymerized tubulin in vitro J. Cell Biol. 91, 479-487. Seagull, R. W. (1989). The plant cytoskeleton. Crit. Rev. Pl. Sci. 8, 131167. Seagull, R. W. and Heath, I. B. (1980). The organization of cortical microtubule arrays on the radish root hair. Protoplasma 103, 205-229. Silflow, C. D., Oppenheimer, D. G., Kopczak, S. D., Ploense, S. E., Ludwig, S. R., Haas, N. and Shustad, D. P. (1987). Plant tubulin genes: structure and differential expression during development. Dev. Genet. 8, 435-460. Sonobe, S. (1990). Cytochalasin B enhances cytokinetic cleavage in miniprotoplasts isolated from cultured tobacco cells. Protoplasma 155, 239-242. Staiger, C. J. and Lloyd, C. W. (1991). The plant cytoskeleton. Curr. Opin. Cell Biol. 3, 33-42. Tiwari, S. C., Wick, S. M., Williamson, R. E. and Gunning, B. E. S.

901

(1984). Cytoskeleton and integration of cellular function in cells of higher plants. J. Cell Biol. 99, 63s-69s. Towbin, H., Staehelin, T. and Gordon, J. (1979). Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedures and some applications. Proc. Nat. Acad. Sci. USA 76, 4350-4354. Tucker, R. (1990). The role of microtubule-associated proteins in brain morphogenesis: a review. Brain Res. Rev. 15, 101-120. Vantard, M., Levilliers, N., Hill, A. M. and Adoutte, A. (1990). Incorporation of Paramecium axonemal tubulin into higher plant cells reveals functional sites of microtubule assembly. Proc. Nat. Acad. Sci. USA 87, 8825-8829. Vantard, M., Schellenbaum, P., Fellous, A. and Lambert, A. M. (1991). Characterization of maize microtubule-associated proteins, one of which is immunologically related to tau. Biochemistry 30, 9334-9340. Wick, S. M. (1991). The preprophase band. In The Cytoskeletal Basis of Plant Growth and Form (ed. C. W. Lloyd), pp. 85-99. Academic Press, London. Yadav, N. S. and Filner, P. (1983). Tubulin from cultured tobacco cells: isolation and identification based on similarities to brain tubulin. Planta 157, 46-52. Zhang, D., Wadsworth, P. and Heplar, P. K. (1990). Microtubule dynamics in living dividing plant cells: confocal imaging of microinjected fluorescent brain tubulin. Proc. Nat. Acad. Sci. USA 87, 8820-8824. (Received 1 April 1993 - Accepted 6 May 1993)