Centrosomal Components Immunologically Related to Tektins from Ciliary and ﬂagellar Microtubules

Journal of Cell Science 107, 2095-2105 (1994) 2095 Printed in Great Britain © The Company of Biologists Limited 1994

Centrosomal components immunologically related to tektins from ciliary and ﬂagellar microtubules

Walter Steffen*, Elizabeth A. Fajer† and Richard W. Linck University of Minnesota, Department of Cell Biology and Neuroanatomy, Minneapolis, MN 55455, USA *Present address: University of Vienna, Institute of Biochemistry and Molecular Cell Biology, Dr. Bohrgasse 9, A-1030 Vienna, Austria †Present address: Florida State University, Institute of Molecular Biophysics, Tallahassee, FL 32306-3015, USA

SUMMARY

Centrosomes are critical for the nucleation and organiz- same protein. Four independently derived monoclonal ation of the microtubule cytoskeleton during both inter- anti-tektins were found to stain centrosomes of S. solidis- phase and cell division. Using antibodies raised against sea sima oocytes and CHO and HeLa cells, by immunofluores- urchin sperm flagellar microtubule proteins, we charac- cence microscopy. In particular, the centrosome staining of terize here the presence and behavior of certain compo- one monoclonal antibody specific for tektin B (tekB3) was nents associated with centrosomes of the surf clam Spisula cell-cycle-dependent for CHO cells, i.e. staining was solidissima and cultured mammalian cells. A Sarkosyl observed only from early prometaphase until late detergent-resistant fraction of axonemal microtubules was anaphase. By immuno-electron microscopy tekB3 specifi- isolated from sea urchin sperm flagella and used to produce cally labeled material surrounding the centrosome, monoclonal antibodies, 16 of which were specific- or cross- whereas a polyclonal anti-tektin B recognized centrioles as specific for the major polypeptides associated with this well as the centrosomal material throughout the cell cycle. microtubule fraction: tektins A, B and C, acetylated α- Finally, by immunoblot analysis tekB3 stained polypep- tubulin, and 77 and 83 kDa polypeptides. By 2-D isoelec- tides of 48-50 kDa in isolated spindles and centrosomes tric focussing/SDS polyacrylamide gel electrophoresis the from CHO cells. tektins separate into several polypeptide spots. Identical spots were recognized by monoclonal and polyclonal antibodies against a given tektin, indicating that the different Key words: centriole, intermediate filament, mitosis, Spisula polypeptide spots are isoforms or modified versions of the solidissima, monoclonal antibody, immuno-electron microscopy

INTRODUCTION Kimble and Kuriyama, 1992). The PCM can be dissociated from the centrioles in vivo by treatment of cells with the micro- Centrosomes are responsible for nucleating and organizing the tubule drug colchicine (Sellitto and Kuriyama, 1988). From microtubule cytoskeleton of interphase and mitotic cells. Cen- these observations questions arise concerning the structural trosomes are complex structures, which in animal cells are and functional nature of the PCM: what is the framework with usually, but not always, composed of two centrioles and a sur- which the pericentriolar components are associated, and what rounding, amorphous, pericentriolar material (PCM). During organizes the PCM around the centrioles under normal con- the last two decades great effort has gone into studying the ditions? function of centrosomes as microtubule organizing centers, and Recently, a number of components have been found to be it has become clear that the capability for microtubule nucle- associated with the centrosome/PCM. A ~51 kDa polypeptide ation in interphase and mitotic cells is associated with the PCM is associated with the PCM of mitotic spindles in sea urchins rather than with the centrioles (Berns and Richardson, 1977; and appears to be involved in nucleating microtubule assembly Brenner et al., 1977; Gould and Borisy, 1977; Toriyama et al., (Toriyama et al., 1988). In addition, several components have 1988). been identified as being part of the PCM, by immunological Our structural knowledge about centrosomes is predomi- methods (for a complete list, see Kimble and Kuriyama, 1992), nantly restricted to the structure of centrioles with their circular such as centrin (Baron and Salisbury, 1988), centractin, an arrangement of nine triplet microtubules (reviewed by actin homolog (Clark and Meyer, 1992), MAP-1 (Mascardo et Wheatley, 1982). The PCM, on the other hand, appears as a al., 1982; Sato et al., 1983), protein kinase II (Nigg et al., less-defined, osmiophilic material in close association with the 1985), and phosphoproteins (Vandré et al., 1984; Kuriyama, centrioles, and its biochemical composition is just beginning 1989); however, little is known of the importance of these com- to be clarified (for reviews see Bornens and Karsenti, 1986; ponents for centrosomal functions. Certainly, an important 2096 W. Steffen, E. A. Fajer and R. W. Linck

finding is the discovery of γ-tubulin (Horio et al., 1991; Oakley precipitating them out of cell culture supernatant with 50% and Oakley, 1989; Oakley et al., 1990; Stearns et al., 1991; ammonium sulfate, by purifying them over a Protein G column or by Zheng et al., 1991). γ-Tubulin shares significant sequence injecting hybridoma cells into mice to obtain ascites fluid. In this homology with both α- and β-tubulin, it is localized in micro- study, except when indicated otherwise, antibodies obtained only tubule organizing centers (MTOCs) and centrosomes, and from cell culture supernatant were employed. The antibody class of although it appears not to be present along microtubules in monoclonals were determined using the isotyping kit from Boehringer Mannheim (Indianapolis, IN). vivo, it is required for nucleation of microtubules from centrosomes. Cell culture In an earlier study we noticed that affinity-purified, poly- Chinese hamster ovary (CHO) cells were cultured in F-10 medium clonal antibodies raised against sea urchin tektins cross-reacted containing 10% fetal bovine serum at 10% CO2 and 37¡C. HeLa cells with basal bodies and centrioles from various cell lines (Steffen were cultured in DMEM medium containing 10% fetal calf serum. and Linck, 1988). Tektins are filamentous proteins that form Mouse myeloma (X63-Ag 8.653) and mouse hybridoma cells were polymers in the walls of axonemal doublet microtubules from cultured in DMEM containing 10% fetal calf serum supplemented sea urchin sperm flagella and presumably in triplet micro- with 10 mM hypoxanthine and 1.6 mM thymidine. For immunofluo- tubules of centrioles. By sequence analysis tektins have been rescence microscopy cells were cultured on poly-L-lysine-coated cov- shown to be an independent class of filamentous proteins erslips. (Chen et al., 1993; Norrander et al., 1992), but they share bio- Isolation of centrosomes and spindles chemical and immunological properties with intermediate Centrosomes were isolated from CHO cells according to Mitchison filament proteins (Chang and Piperno, 1987; Linck and and Kirschner (1986). Briefly, confluent cells from 30 cm × 10 cm Stephens, 1987; Steffen and Linck, 1989a), and tektins have a culture dishes of confluent cells were treated with 10 µg/ml nocoda- predicted secondary structural motif similar to that of inter- zole and 5 µg/ml cytochalasin B for 90 minutes to disintegrate the mediate filament (IF) proteins (Norrander et al., 1992). In this cytoskeleton. Cells were harvested in PBS, washed first with diluted study we present the characterization of a set of monoclonal PBS (1/10 the original concentration) containing 8% sucrose and then antibodies raised against axonemal proteins. Using these with 8% sucrose in water. Cells were lysed in 1 mM Tris-HCl, pH 8.0, containing 8 mM β-mercaptoethanol and 0.5% Nonidet P40. The tektin-specific monoclonal and polyclonal antibodies, we β demonstrate the presence of a tektin-like component in cen- lysate was brought to 10 mM PIPES, pH 7.2, 1 mM EGTA, 8 mM - mercaptoethanol, and fractionated on a sucrose gradient. trosomes of molluscan and mammalian cells. Mitotic spindles were isolated from CHO cells according to Kuriyama et al. (1984). Semi-confluent cells were synchronized by MATERIALS AND METHODS arresting them first in S-phase with 3 mM thymidine and then in M- phase with 0.1 µg/ml nocodazole. Harvested cells were freed from the Antibodies culture dishes and washed free of the nocodazole. Mitotic spindles Rabbit polyclonal antibodies were raised against SDS-PAGE-purified were isolated 18 minutes after the removal of the drug by resuspend- tektin A, B and C derived from sperm axonemal doublet microtubules ing the pelleted cells in 2 mM PIPES, pH 6.8, 20 µg/ml taxol, 0.25% of two sea urchin species: Lytechinus pictus (L.p.) and Strongylocen- Triton X-100. trotus purpuratus (S.p.). A detailed characterization of the polyclonal anti-tektins is provided elsewhere (Linck et al., 1987). All polyclonal Isolation of cytoplasmic microtubules and acetylation of anti-tektins employed in this study were affinity purified with tektin tubulin filaments from L. pictus. Eggs from female Strongylocentrotus purpuratus sea urchins were To raise monoclonal antibodies against components of axonemal obtained by shedding them into artificial sea water (423 mM NaCl, 9 microtubules, the antigen was prepared as follows: doublet micro- mM KCl, 23 mM MgCl2, 25 mM MgSO4, 2 mM NaHCO3, 9 mM tubules of S. purpuratus sperm were extracted twice with 0.5% CaCl2, 5 mM Tris-HCl, pH 8.0) at 4¡C. They were washed with arti- Sarkosyl in 50 mM Tris, 1 mM EDTA, 1 mM DTT, pH 8.0, on ice, ficial sea water, filtered through cheesecloth and a 95 µm nitex filter resulting in preparations of axonemal ribbons consisting of three to strip off their egg jelly, and then washed three times with PIPES protofilaments each. Protofilament ribbons consisting principally of buffer (100 mM PIPES, 5 mM EGTA, 2 mM MgCl2, 0.1 mM GTP, α- and β-tubulin, tektin A, B and C, and two polypeptides of about pH 6.9). After a final wash the eggs were resuspended in an equal 77 kDa and 83 kDa were dissolved in SDS sample buffer (2% SDS, volume of PIPES buffer containing 1 mM PMSF and homogenized 0.5% β-mercaptoethanol, 2.5 mM Tris, 19.2 mM glycine) and then on ice in a glass Potter homogenizer. The homogenization was dialyzed extensively against 2 mM Tris, 0.5 mM EDTA, 0.1 mM DTT monitored by phase-contrast microscopy. After centrifugation at to remove most of the SDS. The antigen was mixed with Freund’s 15,000 gav for 30 minutes, the cytoplasmic phase was collected complete adjuvant and injected into mice subcutaneously and (avoiding the lipid layer) and was centrifuged again at 100,000 gav for intraperitoneally. Mice were boosted with antigen suspended in 60 minutes. The supernatant was brought up to 1 mM GTP and Freund’s incomplete adjuvant at days 14, 21 and 28 after initial immu- incubated at 37¡C for 30 minutes. The polymerization of tubulin was nization. A 100 µg sample of antigen per mouse was used for each monitored by turbidity measurements at A350. Microtubules were injection. The immuno-response was tested by SDS-PAGE depolymerized and polymerized a second time and then stabilized immunoblot. Spleen cells were isolated and fused with X63-Ag 8.653 with 20 µg/ml taxol. Taxol-stabilized microtubules were centrifuged mouse myeloma cells according to Goding (1986). Fused cells were at 100,000 gav for 30 minutes and resuspended in 2 M sodium acetate, selected by culturing the cells in 96-well plates in Dulbecco’s pH 8.1. Up to 1% acetic anhydride was added to the microtubule sus- modified Eagle’s medium (DMEM) medium containing 10% fetal pension. After an incubation of about 30 minutes at room tempera- bovine serum, 10 mM hypoxanthine, 40 µM aminopterin and 1.6 mM ture microtubules were centrifuged at 100,000 gav and used for gel thymidine. To identify antibody-producing hybridomas, culture electrophoresis and immunoblotting. Acetylation of cytoplasmic supernatant of the colonies were screened on SDS-PAGE microtubules was carried out according to Piperno and Fuller (1985). immunoblots of S. purpuratus protofilament ribbons. Immunoblot- positive clones were cultured twice under limited dilution to ensure SDS-PAGE immunoblotting that they were monoclonal cell lines. Antibodies were obtained by Protein samples were separated by SDS-PAGE according to Laemmli Anti-tektin staining of centrosomes 2097

(1970) and by 2-dimensional IEF/SDS-PAGE as described earlier an EIKONIX 78/99 image-scanning system. A computer program (Steffen and Linck, 1989b). Electrophoresis purity reagent SDS from (developed by Dr E. H. Egelman, Department of Cell Biology and Bio-Rad (Richmond, CA) was used for SDS-PAGE and LKB ampho- Neuroanatomy, University of Minnesota) was utilized to render pixels lines from Pharmacia (Piscataway, NJ) were used for isoelectric with highest black level (gold particles) in yellow. The program focussing (IEF). Transblotting was performed under SDS conditions according to Towbin et al. (1979) or under the non-SDS conditions of Dunn (1986) or Matsudaira (1987). For the Dunn procedure the gel was initially incubated in 20% glycerol, 50 mM Tris, pH 7.4, for 60 Table 1. Characterization of monoclonal antibodies minutes, and then electroblotted in a buffer containing 10 mM Production no. Antibody Isotype Antigen Crossreaction NaHCO3, 3 mM Na2CO3, 20% methanol, pH 9.9. For the Matsudaira B5-2-2 tekC1 IgG1, k 47 kDa Centrioles, procedure the gel was transferred to nitrocellulose in 10 mM 3-[cyclo- centrosomes* hexylamino]-1-propanesulfonic acid (CAPS), pH 11.0, 10% 26-1-2 tekC2 IgG1, k 47 kDa methanol. For immunodetection a peroxidase-conjugated goat anti- 3E9-1-1 tekC3 IgM 47 kDa Centrosomes† mouse (Bio-Rad, Richmond, CA) and 4-chloro-2-naphthol were 14H7-C-3 tekB1 IgG1, k 51 kDa Centrosomes‡ employed. Double-labeling immunoblotting was performed as 33-4-2 tekB2 IgG1, k 51 kDa described earlier (Steffen and Linck, 1989b) by using 52-3-3 tekB3 IgG1, k 51 kDa Spindle poles* 214-4-9 tekB4 ND 51 kDa peroxidase/luminescence and alkaline phosphatase as detection α α systems. 8E11-6-1 -tub IgG1, k -Tubulin Subset of MTs 23-1-4 ND 47, 51, 55, 77, Immuno-microscopy 83 kDa ND 53-2-3 ND 55, 77, 83 kDa IFs, vimentin- For immunofluorescence microscopy cells were cultured on poly-L- type lysine-coated coverslips and fixed for 10 minutes with one of the 205-3-2 IgM 55, 77, 83 kDa ND following fixatives: (1) −20¡C methanol; (2) 3.7% formaldehyde in 227-11-7 IgG1, k 77 kDa ND PIPES buffer (50 mM PIPES, 2 mM EGTA, 1 mM MgCl2), pH 6.8; 246-1-1 ND 77 kDa ND (3) 1% glutaraldehyde and 2% formaldehyde in PIPES buffer con- 138-1-9 ND 77 kDa ND taining 0.5% Triton X-100; (4) 5 mM ethylene glycol-bis(succinic 30-1-2 IgG1, k 83 kDa ND acid N-hydroxysuccinimide ester) (EGS) in PIPES buffer containing 202-1-2 IgG1, k 83 kDa ND 70-2-2 IgG , k 83 kDa ND 0.5% Triton X-100. Some of the cells were extracted prior to fixation 1 with 0.1% Triton X-100, 100 mM PIPES, 2 mM EGTA, 1 mM MgCl2, *CHO and HeLa cells, Spisula eggs. 10 µg/ml taxol, pH 6.8 for 30 seconds. Glutaraldehyde-fixed cells †CHO cells, Spisula eggs. were treated with 0.5 mg/ml NaBH4 in 100 mM PIPES, 140 mM ‡Spisula eggs. NaCl, pH 6.8, on ice for 3× 10 minutes each. Immunostaining was ND, not determined. performed according to Steffen and Linck (1989a). A biotin-conjugated goat anti-mouse or goat anti-rabbit (Sigma, St Louis, MO) and Texas Red-conjugated streptavidin (Molecular Probes, Eugene, OR) was used to detect the primary antibodies. Immunofluorescence images were photographed with hypersensitized Kodak Technical Pan film (Sliva, 1981a,b). Labeling for immuno-electron microscopy was conducted as follows: CHO cells were cultured on H2SO4-treated, carbon-coated coverslips. EM-finder grids were employed to generate a recognizable pattern with the carbon coating. Semi-confluent coverslips were rinsed with PBS and cells were then extracted with 0.1% Triton X- 100 in PIPES buffer, pH 6.8, containing 100 mM PIPES, 2 mM EGTA, 1 mM MgCl2, 10 µg/ml taxol for ~10 seconds at room temperature. After blocking non-specific antibody binding sites with 1% BSA in PIPES buffer, pH 7.2, containing 10 µg/ml taxol for 20 kDa minutes, cells were incubated with primary antibody (same buffer as kDa above) for 60 minutes at room temperature, washed 3× 10 minutes each with BSA-PIPES buffer and fixed with 0.3% glutaraldehyde, 2% kDa formaldehyde in PIPES buffer for 10 minutes. Cells were rinsed kDa several times with PBS and then treated with 0.5 mg/ml NaBH4 in PBS, pH 8.0, 3× 10 minutes each and incubated with 1% BSA in PTN (PBS containing 0.05% Tween-20 and 0.02% NaN3) for 20 minutes. Cells were then incubated with a 5 nm or 1 nm gold-conjugated secondary antibody (Ted Pella, Redding, CA) at a concentration of about 4-8 µg/ml in 1% BSA, 0.5% fish gelatine (Birrell et al., 1987) in PTN overnight at room temperature and washed extensively for several hours. In some experiments cells were incubated with rabbit anti-goat biotin-conjugated and streptavidin/Texas Red to detect the gold-conjugated secondary antibody prior to enhancement of the gold Fig. 1. SDS-PAGE immunoblot of Sarkosyl-resistant protofilament particles with a silver enhancer (Sigma, St Louis, MO). Light micro- ribbons of sea urchin sperm flagellar microtubules. Proteins were graphs were taken of appropriate cells before they were embedded transferred under renaturing conditions according to Dunn (1986) with Vestopal (Polaron, San Jose, CA). Ultrathin sections were cut on and stained with monoclonal antibodies raised against Sarkosyl- a Reichert Ultracut S, stained with uranyl acetate and lead citrate, and resistant protofilaments of S. purpuratus axonemal microtubules. examined with a JEOL 100CX at 80 kV. Top: identification of monoclonal antibodies. Right-hand side: To pseudocolor the gold particles, EM negatives were digitized on indication of molecular masses of relevant components. 2098 W. Steffen, E. A. Fajer and R. W. Linck

Fig. 2. Double-labeled immunoblots of 2-dimensional IEF/SDS-PAGE replicas of Sarkosyl-resistant protofilaments from flagellar microtubules. (a) Ponceau S protein staining of 2-D and 1-D gels; (b) staining with polyclonal anti-(L.p.)-tektin B visualized with a phosphatase assay; (c) staining with monoclonal tekB3 visualized with a peroxidase-luminescence assay; arrowheads indicate the corresponding spots in all three panels. (d) Protein staining with Ponceau S; (e) staining with polyclonal anti-(L.p.)-tektin C visualized with a phosphatase assay; (f) staining with monoclonal tekC2 visualized by a peroxidase-luminescence assay; arrowheads indicate the corresponding spots in all three panels. Numbers at the top of the panel indicate the pH range. Polypeptides in a and d are indicated as α- and β-tubulin, tektins A, B and C, and 77 and 83 (kDa). provided an objective distinction between gold and other round antibodies. As demonstrated by this SDS-PAGE immunoblot, particles of lesser density, such as ribosomes. several monoclonal antibodies were monospecific for tektins B and C: four antibodies (tekB1 to tekB4) recognized tektin B, RESULTS and three antibodies (tekC1 to tekC3) recognized tektin C. Three antibodies (138-1-9, 246-1-1 and 227-11-7) were found Characterization of monoclonal antibodies against to be monospecific for the 77 kDa polypeptide band, and three axonemal ribbons antibodies (30-1-2, 70-2-2 and 202-1-2) were monospecific for By using 0.5% Sarkosyl-resistant protofilament ribbons of S. the 83 kDa polypeptide band. In addition to these monospe- purpuratus sperm axonemes as the antigen, a total of 16 cific antibodies, several monoclonal antibodies were found to antibody-producing monoclonal cell lines were isolated. The crossreact with multiple components of axonemal ribbons: characterization of the antibodies is summarized in Table 1. As antibody 23-1-4 recognized all the tektins as well as the 77 kDa described earlier, Sarkosyl-resistant protofilament ribbons and 83 kDa polypeptide bands, and antibodies 53-2-3 and 205- consist principally of α- and β-tubulin, tektins A, B and C, and 3-2 recognized tektin A, and the 77 kDa and 83 kDa polypep- polypeptides of ~53 kDa, ~77 kDa and ~83 kDa (Linck and tide bands. So far, we were not able to isolate an antibody- Langevin, 1982). producing cell line monospecific for tektin A (cf. Chang and Fig. 1 summarizes the immunoblot data of our monoclonal Piperno, 1987). Anti-tektin staining of centrosomes 2099

Fig. 3. Immunofluorescence of centrosomes in isolated spindles of artificially activated eggs of the surf clam Spisula solidissima. (a-f) Phase- contrast images; (a′-f′) corresponding immunofluorescence images stained with monoclonal antibody specific for acetylated α-tubulin, α-tub (a′); monoclonal antibodies specific for tektin B, tekB1 (b′), tekB3 (c′) and tekC3 (d′); and polyclonal anti-(S.p.)-tektin A (e′) and anti-(S.p.)- tektin B (f′). Bar in f, 20 µm.

The immunoblot shown in Fig. 1 was performed on essentially identical antibody specificities to those shown in axonemal ribbon proteins transferred to nitrocellulose under Fig. 1 and Table 1. the ‘renaturing’ conditions of Dunn (1986). Compared to the As reported earlier (Linck et al., 1987), tektins separate into SDS-’denaturing’ transblot method (Towbin et al., 1979), a 10- several polypeptide spots by 2-dimensional IEF/SDS-PAGE: to 100-fold higher sensitivity of monoclonal antibodies was tektin A splits into one major and two minor spots (pI ~6.9), observed for ribbon proteins electro-transferred under the tektin B splits into several spots (pI ~6.2), and tektin C splits Dunn-renaturing conditions. In addition to the reduced sensi- into one major and three minor polypeptide spots (pI ~6.15). tivity of the monoclonal antibodies for SDS-transblotted In the earlier study of Linck et al. (1987), it was not possible polypeptides, three antibodies (23-1-4, 53-2-3 and 227-11-7) to determine whether the associated polypeptide spots recognized only renatured polypeptides. Since the original represent tektins or unrelated polypeptides, since the poly- antigen consisted of SDS-denatured microtubule proteins (see clonal anti-tektins were raised against tektins purified by only Materials and Methods), and since the binding of some the 1-D SDS-PAGE. However, when monoclonal and polyclonal antibodies depends on the transblot conditions, we assume that anti-tektins were used in a double-labeling 2-D immunoblot- there was some renaturation of the antigens after injection into ting, it was observed that the appropriate monoclonal and poly- the rabbits. Using the CAPS transfer conditions of Matsudaira clonal anti-tektins recognized the same polypeptides, demon- (1987), we also tested the antibodies against whole axonemes, strated here for tektin B (Fig. 2a-c) and tektin C (Fig. 2d-f). containing ~200 polypeptides (Piperno et al., 1977), and found These results indicate that the multiple spots (for a given 2100 W. Steffen, E. A. Fajer and R. W. Linck tektin) represent closely related isoforms or post-translation- ally modified versions of the same protein. For example, there is the possibility that these multiple spots represent phosphorylated tektins; however, treatment of the microtubule fraction containing tektins with alkaline phosphatase prior to SDS- PAGE did not decrease the number or positions of the polypeptide spots (data not shown). Tubulin contributes about 60-70% of the protein sample used for immunization. However, in spite of the high content of tubulin, only one antibody (α-tub) was found to be specific for tubulin. By immunoblot analysis of sea urchin axonemal ribbons this antibody revealed a specificity for α-tubulin (Fig. 1, lane α-tub). By immunofluorescence microscopy the α-tub antibody stained sperm tails of all echinoderms tested (not shown). α-tub also stained centrosomes (i.e. spindle poles) in oocytes of the surf clam Spisula solidissima (Fig. 3a,a′); however, it did not recognize spindle microtubules in Spisula (Fig. 3a,a′) or in embryos of the sea urchins Arbacia punctu- lata and S. purpuratus. On the other hand, in cultured cell lines the α-tub antibody labeled spindle microtubules and a subset of cytoplasmic microtubules (see Steffen and Linck, 1992). The observed staining pattern with antibody α-tub indicated a possible specificity of this antibody for acetylated α-tubulin (Piperno and Fuller, 1985). Cytoplasmic tubulin of sea urchin eggs is known to be non-acetylated (Greer and Rosenbaum, 1989). When twice polymerized-depolymerized cytoplasmic tubulin of S. purpuratus was acetylated with acetic anhydride, a strong immunoblot reaction was obtained, whereas no immunoreaction was noted in preparations of non-acetylated tubulin (not shown). A monoclonal antibody, 6-11B-1, previously characterized as specific for acetylated α-tubulin (Piperno and Fuller, 1985) gave an identical immunoblot pattern. Thus, we conclude that our monoclonal antibody α- Fig. 4. Immunofluorescence of mitotic CHO cells with monoclonal tub is specific for acetylated α-tubulin, and is similar or anti-tektin. CHO cells were fixed with −20¡C methanol and stained perhaps identical in its specificity to the monoclonal 6-11B-1 with tekB3. tekB3 only recognized the centrosomes from early of Piperno and Fuller (1985). In retrospect, the anti-acetylated- prometaphase (a, a′) until anaphase (c, c′), with the highest intensity ′ ′ µ α-tubulin staining of centrosomes in Spisula oocytes is remi- of the immunosignal during metaphase (b, b ). Bar in c , 10 m. niscent of the staining reported for the spindle poles in mouse oocytes (Schatten et al., 1988). centrosomes throughout the cell cycle. The tekB3-specific antigen became visible in early prometaphase and disappeared Centrosomes and anti-tektins again in late anaphase. The intensity of the immunofluores- It was demonstrated previously that polyclonal anti-tektins cence signal increased during prometaphase, until it reached its recognize components in basal bodies of sea urchin sperm and highest level at metaphase, then decreased during anaphase and centrioles of various cell lines (Steffen and Linck, 1988). An vanished in late telophase. In some cases (Fig. 4c) the tekB3- initial screen by immunofluorescence microscopy indicated staining pattern was fan-like, with fluorescence extending that certain polyclonal anti-tektins might label centrosomal slightly in the direction of the spindle fibers. components in addition to centriolar components. In this report By immunofluorescence microscopy several anti-tektin anti- we have examined the idea of tektin-related centrosomal com- bodies (both monoclonal and polyclonal) were found to specifi- ponents by immuno-localization, using both polyclonal anti- cally label the centrosomes of whole, methanol-fixed CHO cells. bodies and the monoclonal antibodies characterized above. As At the ultrastructural level the question of whether anti-tektins summarized in Table 1, several monoclonal anti-tektins were recognize components of centrioles or of the pericentriolar found to label centrosomes. In isolated meiotic spindles of material (PCM), or both structures, can best be answered by Spisula anti-tektins tekB1, tekB3, tekC1 and tekC3 stained the immuno-electron microscopy. While glutaraldehyde fixative is spindle poles (Fig. 3). No distinct signal could be detected in needed for good structural preservation of the cells, we found that whole Spisula oocytes (perhaps because of poor penetration of our monoclonal and polyclonal antibodies were not able to the antibodies through the fixed cytoplasm); however, in recognize glutaraldehyde-fixed material as tested by immuno- mammalian cell lines (CHO and HeLa cells) tekB3, tekC1 and light microscopy. Furthermore, the antibodies did not label tekC3 did recognize centrosomes (e.g. see Figs 4-6, see below). ultrathin sections of Lowicryl- or LR-White-embedded CHO In whole, methanol-fixed CHO cells one monoclonal anti- cells. Immunogold labeling of the centrosomes was finally tektin, tekB3, revealed a cell-cycle-specific staining pattern achieved when Triton X-100-extracted CHO cells were (Fig. 4), while the polyclonal anti-(L.p.)-tektin B labeled the incubated with the anti-tektins prior to fixation with a glu- Anti-tektin staining of centrosomes 2101

Fig. 5. Immuno-electron microscopy of a CHO metaphase cell labeled with the monoclonal anti-tektin tekB3. Mitotic cells were extracted with 0.1% Triton X-100 under microtubule-stabilizing conditions, labeled with the primary antibody prior to fixation with glutaraldehyde/ formaldehyde, and then labeled with 1 nm gold-conjugated goat anti-mouse, followed by silver enhancement of the gold. (a and b) Two ultrathin sections from a section series through a half-spindle. Note the antibody-gold labeling of pericentriolar material. Inset in a, demonstrates the same cell by bright-field light microscopy prior to embedment. Arrow in inset indicates labeled spindle pole shown in a and b. Bar in b, 0.5 µm. taraldehyde/formaldehyde mixture. Fig. 5 shows immuno-EM spindles the monoclonal antibody tekB3 recognized a polypep- images of a mitotic CHO cell labeled with the monoclonal tekB3. tide band of 48-50 kDa (Fig. 7b). In semi-purified preparations The tekB3-specific epitope appeared only in association with of interphase centrosomes the tekB3 did not recognize any material surrounding the centrosome, whereas in centrioles this such components (data not shown), presumably due to the cell epitope was either not present or not accessible to the antibody, cycle dependency of the epitope; however, a similar sized as very few gold particles could be found in close proximity to polypeptide band (of 48-50 kDa) was stained by the polyclonal the centrioles. (Note: the estimated distance from epitope to gold tektin B antibody (Fig. 7d). particle for indirect immuno-labelling is ~40 nm.) On the other In order to correlate further the tekB3-specific epitope with hand, in interphase cells the polyclonal anti-tektin B did label the centrosome, the pericentriolar material (PCM) was disso- centrioles as well as material surrounding the centrosome (Fig. ciated from the centrioles by drug treatment. It has been shown 6); i.e. gold particles were within 40 nm of the centriole. Impor- previously that the PCM can be dissociated by incubating cells tantly, there was no staining of either the surrounding cytoplas- in microtubule drugs (Sellitto and Kuriyama, 1988). As a result mic intermediate filaments or the nuclear lamina. of this drug treatment, cells may form multipolar spindles. The The biochemical identity of the centrosome component(s) distribution of the tekB3-specific antigen was altered when was investigated by SDS-PAGE immunoblot analysis of CHO cells were treated with nocodazole for 90 minutes. The fractions from CHO cells (Fig. 7). In preparations of whole integrity of mitotic centrosomes (spindle poles) was disrupted cells, no centrosomal antigens were detected, presumably and the drug treatment resulted in the formation of multipolar because of their low abundance. However, in purified mitotic spindles (Fig. 8). Each of the spindle poles was recognized by 2102 W. Steffen, E. A. Fajer and R. W. Linck

Fig. 6. Immuno-electron microscopy of a CHO interphase cell labeled with the polyclonal anti-(L.p.)-tektin B. After incubation in primary antibody, the specimen was stained with 1 nm gold-conjugated anti-rabbit IgG and silver-enhanced. The gold labeling is specifically associated with components of the centrosomes, as shown by bright-field light microscopy (inset in a) and in ultrathin sections (a and b). For clarity and objectivity, gold particles were pseudo-colored after computer selection of all areas above a certain gray level; this method allows a distinction to be made between solid black (gold) particles and other dark gray particles, such as ribosomes (compare with Fig. 5). Note that the adjacent intermediate filaments and nuclear lamina are not labeled. Bars in a and b, 1 µm. the tekB3 antibody. As demonstrated previously, polyclonal (~55-, 51 and 47 kDa, respectively) are proteins associated anti-tektin C recognizes the centrioles (see Steffen and Linck, with a highly stable set of protofilaments in the A-tubules of 1988). When nocodazole-treated interphase cells were labeled ciliary and flagellar microtubules (Linck and Langevin, with a monoclonal or polyclonal anti-tektin C, only one or two 1982; Linck et al., 1985; Steffen and Linck, 1989c; Stephens nucleus-associated dots were observed (not shown), confirm- et al., 1989), and that tektins have some similarities to inter- ing the specificity of anti-tektin C for centrioles and support- mediate filament proteins (Norrander et al., 1992; Steffen ing the conclusion that the tekB3 antigen resides in the material and Linck, 1989a,c). Recent studies now suggest that tektins surrounding the centrosome. exist as heteropolymeric, longitudinal filaments associated with the A-tubules (Nojima et al., 1994; Pirner and Linck, DISCUSSION 1993). Furthermore, anti-tektin antibodies label basal bodies and centrioles (Steffen and Linck, 1988), suggesting that Previous studies have demonstrated that tektins A, B and C tektins may also form specialized filaments in centriolar Anti-tektin staining of centrosomes 2103

nents of 48-50 kDa (Fig. 7), suggesting that these components are similar, if not identical. The temporal staining with the monoclonal anti-tektin could be explained by a mitosis-specific, post-translational modification of the tekB3-related epitope, such as acetylation, gutamy- Fig. 7. SDS-PAGE/immunoblots of lation, phosphorylation, tyrosination or others. When cells enter isolated spindles (a and b) and isolated mitosis, centrosomes undergo a structural and functional centrosomes (c and d) from CHO cells, change; the duplicated centrosomes move to opposite sides of after staining with anti-tektin antibodies. the nucleus and are largely responsible for organizing the (a and c) Ponceau staining of the spindle. Mitosis-specific phosphorylation has been demon- nitrocellulose replicas; (b) staining with monoclonal anti-tektin tekB3; and (d) strated for several centrosomal components (Kimble and staining with the polyclonal anti-tektin B. Kuriyama, 1992; Vandré et al., 1984), and phosphorylation or Arrowheads indicate principal stained some other modification might very well account for the bands of 48-50 kDa. appearance of the tekB3 labelling during cell division. In contrast, the lack of centriolar labeling with tekB3 might be explained by a lack of such a modification of tektins that are assembled in centriolar microtubules. We attempted to investi- microtubules. This idea would fit with the templated gate whether flagellar tektins or the centrosomal components assembly of doublet microtubules from the plus ends of basal are phosphorylated; so far, however, we have not been able to body microtubules (Allen and Borisy, 1974). Given these label the relevant polypeptides in vivo (with 32P), and we have features of tektins, the immunological crossreactions of anti- not been able to alter the number or position of spots present in tektin antibodies with centrosomes reported here may be 2D IEF-SDS/PAGE by treating the specimens by alkaline phos- important in understanding the structure and function of phatase. Further studies are required to determine whether post- these organelles. translational modifications might explain the epitope recogni- The staining of centrosomes by monoclonal versus poly- tion during the cell cycle. As indicated above, centrosomes clonal anti-tektin B are similar and yet different. Both types of undergo structural rearrangements at the interphase-mitosis antibodies stain centrosomes in whole cells fixed for immuno- transition; thus, the possibility cannot be excluded that a lack fluorescence microscopy (Fig. 4), indicating that the epitopes of centrosome staining during interphase is due to the lack of are localized at the centrosome, not redistributed during cell accessibility of this particular epitope. lysis for immuno-EM (Figs 5 and 6). Furthermore, the staining In an early electron microscopical study Kalnins and Porter patterns following nocodazole treatment strongly suggests that (1969) suggested that a filamentous material surrounds the the epitopes are associated with the PCM. On the other hand, centrioles. Although there is still no conclusive evidence for the monoclonal anti-tektin B staining is mitosis-specific, a centrosomal filament matrix, a filamentous material has increasing from early prometaphase and decreasing in late been demonstrated for the rootlet structure associated with anaphase, while the polyclonal anti-tektin B staining occurred basal bodies (Salisbury, 1983), and a filamentous connection throughout the cell cycle (Figs 3-6). Although it is not certain between the centrioles has been described in isolated centro- whether monoclonal and polyclonal anti-tektin B recognize the somes of human lymphoblastic cells (Bornens et al., 1987). same or different components in centrosomes and mitotic Now, our immuno-localization studies lend support to the spindles, the immunoblot data indicate polypeptide compo- possibility of a filamentous centrosomal matrix, given that

Fig. 8. Immunofluorescence of a nocodazole-treated mitotic CHO cell, stained with monoclonal anti-tektin tekB3. Cells were treated with 0.1 µg/ml nocodazole for 90 minutes prior to extraction with 0.5% Nonidet P-40 for 30 seconds and fixed with −20¡C methanol for 10 minutes. tekB3 recognized multiple spots (arrowheads) in close proximity of the chromosomes. (a) Phase-contrast image; (b and c) immunofluorescence images at two different focal planes. Bar in c′, 10 µm. 2104 W. Steffen, E. A. Fajer and R. W. Linck tektins are coiled-coil polymers (Norrander et al., 1992; and 8811015, NSF RTG grant DIR-9113444, and the Marine Biological see below). In related studies of spindle pole bodies of Sac- Laboratory (Woods Hole, MA). charomyces, Kilmartin et al. (1993) have shown that a 110 kDa, predicted coiled-coil protein (SPC110) forms a filamentous bridge between the central and inner plaques of the REFERENCES spindle pole body; interestingly, SPC110 is also cell cycle regulated. Allen, C. and Borisy, G. (1974). Structural polarity and directional growth of The properties and structure of tektins may provide some microtubules of Chlamydomonas flagella. J. Mol. Biol. 90, 381-402. Bao, L., Zimmer, W. E. and Balczon, R. (1993). Molecular cloning and insight into the structure and function of the centrosomal autoepitope analysis of PCM-1, a novel centrosome autoantigen. Mol. Biol component(s) recognized by the anti-tektin antibodies. By Cell 4, 395a. cDNA sequence analysis tektins have been shown to be a Baron, A. T. and Salisbury, J. L. (1988). Identification and localization of a distinct class of proteins, sharing only nine residues with the novel, cytoskeletal centrosome-associated protein in PtK2 cells. J. Cell Biol. main consensus sequence of IF proteins (Norrander et al., 107, 2669-2678. Berns, M. W. and Richardson, S. M. (1977). Continuation of mitosis after 1992). Yet, tektins and IF proteins appear to have a similar selective laser microbeam destruction of the centriolar region. J. Cell Biol. structural motif, the long-range pattern of coiled-coil and non- 75, 977-982 helical linkers domains (Norrander et al., 1992). Furthermore, Birrell, G., Hedberg, K. K. and Griffith, O. (1987). Pitfalls of immunogold tektins and IF proteins possess numerous immunological sim- labeling: Analysis by light microscopy, transmission electron microscopy, ilarities, i.e. a variety of both polyclonal and monoclonal anti- photoelectron microscopy. J. Histochem. Cytochem. 35, 843-853. Bornens, M. and Karsenti, E. (1986). The centrosome. In Membrane tektin antibodies crossreact with desmin, vimentin, cytoker- Structure and Function (ed. E.E. Bittar), pp. 99-171. John Wiley & Sons, atins and nuclear lamins (Chang and Piperno, 1987; Steffen New York. and Linck, 1989a). Besides these studies, several others have Bornens, M., Paintrand, M., Berges, J., Marty, M. C. and Karsenti, E. reported the crossreaction of monoclonal antibodies between (1987). Structural and chemical characterization of isolated centrosomes. centrosomes and IF or IF-associated proteins (antibody D73.3, Cell Motil. Cytoskel. 8, 238-249. Brenner, S., Branch, A., Meredith, S. and Berns, M. W. (1977). The absence Ring et al., 1980; antibody AH-6, Schatten et al., 1987, of centrioles from spindle poles of rat kangaroo (PtK2) cells undergoing 1988a,b; antibody CTR2611, Buendia et al., 1990; Perret et al., meiotic-like reduction division in vitro. J. Cell Biol. 72, 368-379. 1991; antibody CHO4, Sellitto et al., 1992). Despite these Buendia, B., Antony, C., Verde, F., Bornens, M. and Karsenti, E. (1990). A immunological similarities, we found that the monoclonal anti- centrosomal antigen localized on intermediate filaments and mitotic spindle tektin tekB3 and the polyclonal anti-(L.P.)-tektin B specifically poles. J. Cell Sci. 97, 259-271. Chang, X. J. and Piperno, G. (1987). Cross-reactivity of antibodies for stained the centrosome, and did not stain the surrounding IFs flagellar tektins and intermediate filament subunits. J. Cell Biol. 104, 1563- or nuclear lamina (Figs 5 and 6). Conceivably then, previous 1568. reports of centrosome staining by IF antibodies might be attrib- Chen, R., Perrone, C.A., Amos, L. A. and Linck, R. W. (1993). Tektin B1 utable to tektin-like rather than IF-like components. More from ciliary microtubules: primary structure as deduced from the cDNA recently, a number of centrosomal or MTOC components have sequence and comparison with tektin A1. J. Cell Sci. 106, 909-918. Clark, S. W. and Meyer, D. I. (1992). Centractin is an actin homologue been reported and some have been sequenced (e.g. see Bao et associated with the centrosome. Nature 359, 246-250. al., 1993; Doxsey et al., 1994; Kilmartin et al., 1993; Lange Doxsey, S. J., Stein, P., Evans, L., Calarco, P. D. and Kirschner, M. (1994). and Gull, 1993; Oegema et al., 1993; Paul and Quaroni, 1993; Pericentrin, a highly conserved centrosome protein involved in microtubule Tassin et al., 1993). As one example, centrosomin has several organization. Cell 76, 639-650. short stretches (each containing 4-5 amino acid residues) of Dunn, S. D. (1986). Effect of the modification of transfer buffer composition and the renaturation of proteins in gels on the recognition of proteins on similarity with nuclear lamins but has little other structural western blots by monoclonal antibodies. Anal. Biochem. 157, 144-153. similarity to IF proteins (Joswig et al., 1991) or to tektins Goding, J. W. (1986). Monoclonal Antibodies: Principles and Practice. (Norrander et al., 1992). While IFs have a diameter of ~10 nm Academic Press, London. and are observed to terminate at the outer border of the cen- Gould, R. R. and Borisy, G. G. (1977). The pericentriolar material in Chinese trosome (Starger et al., 1978; Vogl et al., 1983), isolated tektin hamster ovary cells nucleates microtubule formation. J. Cell Biol. 73, 601- 615. filaments measure only 2-3 nm in diameter (Linck and Greer, K. and Rosenbaum, J. L. (1989). Post-translational modification of Langevin, 1982) and might be difficult to visualize in ultrathin tubulin. In Cell Motility, vol. 2 (ed. F. D. Warner and J. R. McIntosh), pp. 47- section through centrosomes, due to insufficient orientation or 66. Alan R. Liss, New York. preservation. It would be interesting if the putative centroso- Horio, T., Uzawa, S., Jung, M. K., Oakley, B. R., Tanaka, K. and Yanagida, mal matrix, in addition to its role in nucleating microtubule M. (1991). The fission yeast γ-tubulin is essential for mitosis and is localized assembly, also provided a link with the intermediate filament at microtubule organizing centers. J. Cell Sci. 99, 693-700. Joswig, G., Petzelt, C. and Werner, D. (1991). Murine cDNAs coding for the cytoskeleton. The centrosomal component(s) recognized by centrosomal antigen centrosomin A. J. Cell Sci. 98, 37-43. anti-tektin antibodies reported here might have the properties Kalnins, V. J. and Porter, K. R. (1969). Centriole replication during to provide this dual role. ciliogenesis in the chick tracheal epithelium. Z. Zellfors. Mikrosk. Anat. 100, 1-30. We thank Kathy Ensrud for introducing us to the technique of Kilmartin, J. V., Dyos, S. L., Kershaw, D. and Finch, J. T. (1993). A spacer producing monoclonal antibodies, Rich Erickson for providing protein in the Saccharomyces cerevisiae spindle pole body whose transrcript technical support in culturing the hybridomas and Dr Gianni Piperno is cell cycle-regulated. J. Cell Biol. 123, 1175-1184. Kimble, M. and Kuriyama, R. (1992). Functional components of for the anti-acetylated tubulin antibody (6-11B-1). The work on microtubule-organizing centers. Int. Rev. Cytol. 136, 1-50. Spisula oocytes was conducted during a stay at the Marine Biologi- Kuriyama, R., Keryer, G. and Borisy, G. G. (1984). The mitotic spindle of cal Laboratory, Woods Hole, in the laboratory of Dr Sidney S. Tamm Chinese hamster ovary cells isolated in taxol-containing medium. J. Cell Sci. and Raymond E. Stephens to whom we are very grateful. We also 66, 265-275. thank Gerald Sedgewick for assistance with the photography. This Kuriyama, R. (1989). 225-kilodalton phosphoprotein associated with mitotic study was supported by USPHS grant GM 35648, NSF grant DCB- centrosomes in sea urchin eggs. Cell Motil. Cytoskel. 12, 90-103. Anti-tektin staining of centrosomes 2105

Laemmli, U. K. (1970). Cleavage of structural proteins during the assembly of Schatten, H., Walter, M., Mazia, D., Biessmann, H., Paweletz, N., Coffe, G. the head of bacteriophage T4. Nature 227, 680-685. and Schatten, G. (1987). Centrosome detected in sea urchin eggs with a Lange, B. M. H. and Gull, K. (1993). Cenexin, a protein defining one of the monoclonal antibody against Drosophila intermediate filaments: two centrioles in a mammalian cell. Mol. Biol Cell 4, 394a. Characterization of stages of the division cycle of centrosomes. Proc. Nat. Linck, R. W. and Langevin, G. L. (1982). Structure and chemical composition Acad. Sci. USA 84, 8488-8491. of insoluble filamentous components of sperm flagellar microtubules. J. Cell Schatten, G., Simerly, C., Asai, D. J., Szoke, E., Cooke, P. and Schatten, H. Sci. 58, 1-22. (1988a). Acetylated α-tubulin in microtubules during mouse fertilization and Linck, R. W., Amos, L. A. and Amos, W. B. (1985). Localization of tektin early development. Dev. Biol. 130, 74-86. filaments in microtubules of sea urchin sperm flagella by immunoelectron Schatten, H., Walter, M., Biessmann, H. and Schatten, G. (1988b). microscopy. J. Cell Biol. 100, 126-135. Microtubules are required for centrosome expansion and repositioning while Linck, R. W., Goggin, M. J., Norrander, J. M. and Steffen, W. (1987). microfilaments are required for centrosome separation in sea urchin eggs Characterization of antibodies as probes for structural and biochemical during fertilization and mitosis. Cell Motil. Cytoskel. 11, 248-259. studies of tektins from ciliary and flagellar microtubules. J. Cell Sci. 88, 453- Sellitto, C. and Kuriyama, R. (1988). Distribution of pericentriolar material in 466. multipolar spindles induced by colcemid treatment in Chinese hamster ovary Linck, R. W. and Stephens, R. E. (1987). Biochemical characterization of cells. J. Cell Sci. 89, 57-65. tektins from sperm flagellar doublet microtubules. J. Cell Biol. 104, 1069- Sellitto, C., Kimble, M. and Kuriyama, R. (1992). Heterogeneity of 1075. microtubule organizing center components as revealed by monoclonal Mascardo, R. N., Sherline, P. and Wheatherbee, J. (1982). Localization of antibodies to mammalian centrosomes and to nucleus-associated bodies from high molecular weight microtubule-associated protein (MAP1 and MAP2) in Dictyostelium. Cell Motil. Cytoskel. 22, 7-24. HeLa microtubule-organizing center. Cytobios 35, 113-127. Sliva, R. (1981a). Hypersensitizing, part 1. Astronomy, April, pp. 39-42. Matsudaira, P. (1987). Sequence from picomole quantities of proteins Sliva, R. (1981b). Hypersensitizing, part 2. Astronomy, May, pp. 48-50. electroblotted onto polyvinylidene difluoride membranes. J. Biol. Chem. Starger, J. M., Brown, W. E., Goldman, A. E. and Goldman, R. D. (1978). 262, 10035-10038. Biochemical and immunological analysis of rapidly purified 10-nm filaments Mitchison, T. J. and Kirschner, M. W. (1986). Isolation of mammalian from baby hamster kidney (BHK-21) cells. J. Cell Biol. 78, 93-109. centrosomes. Meth. Enzymol. 134, 261-268. Stearns, T., Evans, L. and Kirschner, M. (1991). γ-tubulin is a highly Nigg, E. A., Schaefer, G., Hilz, H. and Eppenberger, H. M. (1985). Cyclic- conserved component of the centrosome. Cell 65, 825-836. AMP-dependent protein kinase type II is associated with the Golgi complex Steffen, W. and Linck, R. W. (1988). Evidence for tektin-related components and with centrosomes. Cell 41, 1039-1051. in axonemal microtubules and centrioles. Proc. Nat. Acad. Sci. 85, 2643- Nojima, D., Linck, R. W. and Egelman, E. H. (1994). Evidence suggests 2647. that at least one of the 13 protofilaments of the A-tubule from doublet Steffen, W. and Linck, R. W. (1989a). Relationship between tektins and microtubules is not composed of tubulin. Biophys. J. 66, A313. intermediate filaments: an immunological study. Cell Motil. Cytoskel. 14, Norrander, J. M., Amos, L. A. and Linck, R. W. (1992). Primary structure of 359-371. tektin A1: comparison with intermediate-filament proteins and a model for its Steffen, W. and Linck, R. W. (1989b). Multiple immunoblot: A sensitive association with tubulin. Proc. Nat. Acad. Sci. USA 89, 8567-8571. technique to stain proteins and detect multiple antigens on a single 2-D Oakley, B. R., Oakley, C. E., Yoon, Y. and Jung, M. K. (1990). γ-tubulin is a replica. Electrophoresis 10, 714-718. component of the spindle pole body that is essential for the microtubule Steffen, W. and Linck, R. W. (1989c). Tektins in ciliary and flagellar function in Aspergillus nidulans. Cell 61, 1289-1301. microtubules and their association with other cytoskeletal systems. In Cell Oakley, C. E. and Oakley, B. R. (1989). Identification of γ-tubulin, a new Movement, vol. 2 (ed. J. R. McIntosh and F. D. Warner), pp. 67-81. Alan R. member of the tubulin superfamily encoded by mipA gene of Aspergillus Liss, Inc., New York. nidulans. Nature 338, 662-664. Steffen, W. and Linck, R. W. (1992). Evidence for a non-tubulin spindle Oegema, K., Whitfield, W. G. and Alberts, B. M. (1993). Characterization of matrix and for spindle components immunologically related to tektin the functional domains of a Drosophila centrosomal protein. Mol. Biol. Cell filaments. J. Cell Sci. 101, 809-822. 4, 228a. Stephens, R. E., Oleszko-Szuts, S. and Linck, R. W. (1989). Retention of Paul, E. C. A. and Quaroni, A. (1993). Identification of a 102 kDa protein ciliary ninefold structure after removal of microtubules. J. Cell Sci. 92, 391- (cytocentrin) immunologically related to keratin 19, which is a 402. cytoplasmically derived component of the mitotic spindle pole. J. Cell Sci. Tassin, A. M., Celati, C., Paintrand, M., Kilmartin, J. V. and Bornens, M. 106, 967-981. (1993). Identification of a human centrosomal 100 kd protein homologue of Perret, E., Albert, M., Bordes, N., Bornens, M. and Soyer-Gobillard, M.-O. the 110 kd SPB protein from Saccharomyces cerevisiae. Mol. Biol. Cell 4, (1991). Microtubular spindle and centrosome structure during the cell cycle 395a. in a dinoflagellate Crypthecodinium cohnii. Biosystems 25, 53-65. Toriyama, M., Otha, K., Endo, S. and Sakai, H. (1988). 51-kD protein, a Piperno, G., Huang, B. and Luck, D. J. L. (1977). Two-dimensional analysis component of microtubule-organizing granules in the mitotic apparatus of flagellar proteins from wild-type and paralyzed mutants of involved in aster formation in vitro. Cell Motil. Cytoskel. 9, 117-128. Chlamydomonas reinhardtii. Proc. Nat. Acad. Sci. USA 74, 1600-1604. Towbin, H., Staehlin, T. and Gordon, J. (1979). Electrophoretic transfer of Piperno, G. and Fuller, M. T. (1985). Monoclonal antibodies specific for an proteins from polyacrylamide gels to nitrocellulose sheets: procedure and acetylated form of α-tubulin recognize the antigen in cilia and flagella from a some applications. Proc. Nat. Acad. Sci. USA 76, 4350-4354. variety of organisms. J. Cell Biol. 101, 2085-2094. Vandré, D. D., Davis, F. M., Rao, P. N. and Borisy, G. G. (1984). Pirner, M. A. and Linck, R. W. (1993). Structural associations in tektin A-B Phosphoproteins are components of mitotic microtubule organizing centers. filaments. Mol. Biol. Cell 6, 272a. Proc. Nat. Acad. Sci. USA 81, 4439-4444. Ring, D., Hubble, R., Caput, D. and Kirschner, M. W. (1980). Isolation of Vogl, A. W., Linck, R. W. and Dym, M. (1983). Colchicine-induced changes microtubule organizing centers from mouse neuroblastoma cells. In in the cytoskeleton of the golden-mantled ground squirrel (Spermophilus Microtubules and Microtubule Inhibitors (ed. M. DeBrabander and J. lateralis) sertoli cells. Amer. J. Anat. 168, 99-108. DeMey), pp. 297-309. Elsevier/North-Holland Biomedical Press, Wheatley, D. N. (1982). The Centriole: a Central Enigma of Cell Biology. Amsterdam. Elsevier, New York. Salisbury, J. L. (1983). Contractile flagellar roots: the role of calcium. J. Zheng, Y. X., Jung, M. K. and Oakley, B. R. (1991). γ-tubulin is present in Submicrosc. Cytol. 15, 105-110. Drosophila melanogaster and Homo sapiens and its associated with the Sato, C., Nishizawa, K., Nakamura, H., Komagoe, Y., Shimada, K., Ueda, centrosome. Cell 65, 817-823. R. and Suzuki, S. (1983). Monoclonal antibody against microtubule associated protein-1 produces immunofluorescent spots in the nucleus and centrosome of cultured mammalian cells. Cell Struct. Funct. 8, 245-254. (Received 30 January 1994 - Accepted 26 April 1994)