Cell Motility and the Cytoskeleton 43:72–81 (1999) Centrin-Like Filaments in the Cytopharyngeal Apparatus of the Ciliates Nassula and Furgasonia: Evidence for a Relationship With Microtubular Structures Bernard Vigues,* Marie-Pierre Blanchard, and Philippe Bouchard UPRES A CNRS 6023, Laboratoire de Biologie Comparée des Protistes, Université Blaise Pascal Clermont-Ferrand II, Aubière Cedex, France The cytopharyngeal apparatus in the Nassulinid ciliates Nassula and Furgasonia is a highly specialized microtubular/filamentous organelle designed for ingestion of organisms such as filamentous bacteria. From studies on living cells, it was previously shown that this organelle, also called ‘‘feeding basket,’’ guides the filamentous bacteria and manipulates them to some extent during the early steps of ingestion. This results in a complex sequence of movements where the basket is successively dilated and constricted in its upper part. Whereas some of these movements (dilation) seem to be intrinsic to the microtubular components of the basket, others (constriction) are believed to be mediated by contractile filamentous structures [Tucker, 1968: J. Cell Sci. 3:493–514]. In this study, we have used antibodies raised against ciliate centrins to demonstrate these proteins by Western blot and immunocytochemical methods in Nassula and Furgasonia. In both ciliates, a 20-kDa centrin immunoanalog was localized in the upper (contractile) part of the cytopharyngeal apparatus. Immunoelectron microscopy revealed that cytopharyngeal centrin is engaged in filamentous material, forming a sphincterlike structure possibly involved in the movements of contraction. Interestingly, physical links were noted between filaments labeled for centrin and cytopharyngeal microtubules. The mechanistic implications of these findings are discussed. Cell Motil. Cytoskeleton 43:72–81, 1999. r 1999 Wiley-Liss, Inc. Key words: centrin; contractility; microtubules; cell motility; endocytosis; ciliates; Protozoa INTRODUCTION Centrin is a 20-kDa Ca⫹⫹-binding protein that belongs to the EF-hand protein superfamily. Despite its structural similarities to other EF-hand proteins, centrin is functionally unique in that it is believed to be directly responsible for contraction of calcium-sensitive filamentous systems mediating various cell motility processes in unicellular eukaryotes. Centrin-based filamentous systems were initially discovered in the flagellar apparatus of unicellular green algae where they connect basal bodies to one another and to the nuclear envelope [for a review see, e.g., Melkonian et al., 1992]. Since this pionnering work, molecular cloning and/or immunolocalization experiments have indicated that centrin is a ubiquitous protein in eukaryotic cells. A common property of arrays r 1999 Wiley-Liss, Inc. of centrin in a number of cell types is their association with microtubule organizing centers (MTOCs). Immunolocalization experiments indicate centrin in the spindle pole body of yeast [Spang et al., 1993], mammalian centrosome [see, e.g., Baron et al., 1992; Lee and Huang, 1993; Errabolu et al., 1994; Middendorp et al., 1997], and MTOCs associated with basal bodies of cilia and eukary- Contract grant sponsor: Centre National de la Recherche Scientifique. *Correspondence to: Bernard Vigues, UPRES A CNRS 6023, Laboratoire de Biologie Comparée des Protistes, Université Blaise Pascal Clermont-Ferrand II, 63177 Aubière Cedex, France. E-mail: firstname.lastname@example.org Received 23 November 1998; accepted 14 January 1999 Centrin in the Feeding Basket of Nassulinid Ciliates otic flagella [Salisbury, 1989; Levy et al., 1996]. In mammalian centrosomes, centrin has been localized in the core of centrioles and in the pericentriolar matrix which organizes the network of cytoplasmic microtubules and mitotic spindle during cell division. Attempts to find some contribution, even indirect, of centrin to the process of centrosome-dependent assembly of microtubules were, however, unsuccessful [Moudjou et al., 1991]. From more recent studies, further colocalizations with microtubular structures were noted, notably with microtubules of the mid-body, a transient structure assembled in the intercellular bridge of dividing cells [Paoletti et al., 1996]. In no case, however, has the functional significance of centrin-microtubule colocalizations been elucidated. In ciliates, centrin or closely related proteins have been reported in basal bodies [Klotz et al., 1997] and in cortical filamentous networks displaying calcium-dependent contractile properties [Garreau de Loubresse et al., 1991; David and Vigues, 1994; Levy et al., 1996]. We previously showed that some of these networks may be part of cytoskeletal complexes that include microtubular structures [Vigues and Grolière, 1985; Vigues and David, 1994]. It seemed, therefore, that more detailed studies of such complexes might be pertinent for our understanding of centrin-microtubule relationships in general. In the present study, we have investigated the problem of centrin-microtubule relationships in the cytopharyngeal apparatus of Nassula aurea and Furgasonia blochmanni, two ciliates that feed on filamentous bacteria. The fine structure of the cytopharyngeal apparatus, also called the ‘‘feeding basket,’’ has been described in detail by others [Tucker, 1968; Eisler, 1988]. In both ciliates, the feeding baskets are similar in organization, consisting mostly of microtubular structures. These include (1) the nemadesmata, which are arranged in a tube through which the filamentous bacteria enter the organism and (2) crest-shaped structures attached to the nemadesmata in the upper part of the basket and that spiral around the outside of the tube in the lower part. During ingestion, the basket guides the filamentous bacteria and manipulates it to some extent. This results in a complex sequence of movements, the so-called ‘‘feeding movements,’’ where the basket is successively dilated and constricted in its upper part (see insets in Fig. 3). Whereas some of these movements (dilation) seem to be intrinsic to the microtubular components of the basket, others (constriction) are believed to be mediated by contractile filamentous structures [Tucker, 1968, 1978]. We report evidence suggesting that a 20-kDa centrin immunoanalog may be functionally important for constriction of the upper part of the basket. An immunoelectron microscopic study of this region is presented, showing the 73 20-kDa protein is a component of a sphincter-like filamentous structure that interconnects adjacent nemadesmata. Accordingly, we suggest that colocalization of centrinbased filaments and microtubules, as shown in a number of cells, may be more than circumstantial and might reflect physical and possibly functional interactions, as explored in the Discussion. MATERIALS AND METHODS Cells and Cell Fractions The ciliates Nassula aurea and Furgasonia blochmanni were used as cell models for this study. The method used for culturing both ciliates and their food, the cyanobacteria Phormidium inundatum, have been described by others [Tucker, 1968; Eisler, 1988]. For cell ghost preparation, cells were starved for 2 days in sterile Volvic water and then collected by three successive low-speed centrifugations in a GGT oil centrifuge (500g for Nassula; 1,500g for Furgasonia; 5-min runs). Cell pellets were suspended and gently shaken for 10 min in 5 volumes of ice cold SEMT solution (1M sucrose, 1 mM EDTA, 0.1% 2-mercaptoethanol in 10 mM Tris-HCl buffer, pH 8.6). The non-ionic detergent Triton X-100 (0.25 vol of a 10% [v/v] stock solution) and protease inhibitors (1 mM PMSF, 10 mM TAME, 1 mM leupeptin, 0.1 mM pepstatin-A) were added in sequence with continuous agitation. Cilia generally were disrupted at this stage. Removal of endoplasmic material was monitored by phase contrast light microscopy. The released cell ghosts were washed using two successive centrifugations (3,500g, 15 min) in PE buffer (10 mM sodium phosphate, pH 6.9, 1 mM EDTA) containing 1 mM PMSF. When larger amounts of basket material were required, cells from large-scale cultures were concentrated by centrifugation and Triton X-100 was 2% (v/v) in SEMT buffer instead of 0.6% in the usual conditions, resulting in noticeable enrichment in the feeding basket due to disintegration of pellicular structures. Following two successive centrifugations (4,000g; 15 min) in PE buffer, pellets were dissolved in 6M urea, 0.2% Nonidet P-40, 10 mM Tris-HCl, pH 8.6. Extractable proteins were concentrated in a microprodicon vacuum concentrator (Bioblock Scientific, Illkirch, France) filled with 6M urea. The final concentrate was diluted in an equivalent volume of 8M urea, 0.4% Nonidet P-40, and ampholines at the concentration required for isoelectric focalization experiments (see electrophoresis). Antibodies Anti-EEB antiserum was raised against 22/23 kDaCa⫹⫹-binding proteins of the ectoendoplasmic boundary, 74 Vigues et al. a major centrin-like filamentous system in the ciliate Isotricha prostoma. Specificity of this antiserum for ciliate centrins was reported elsewhere [Vigues et al., 1984; Garreau de Loubresse et al., 1991, see also Madeddu et al., 1996 for molecular characterization of the antigens as centrins]. A69 monoclonal antibody (mAb A69) was raised against calmyonemin, a 23-kDa centrin homolog involved in various cell motility processes in entodiniomorphid ciliates [David and Vigues, 1994; Vigues and David, 1994]. Anti-tubulin antibody, raised against ␤-tubulin from Tetrahymena pyriformis, has been described elsewhere [Bouchard et al., 1998]. In all labeling experiments, i.e., immunocytochemistry and immunoblottings (see below), anti-EEB antiserum was diluted 1/250 and monoclonal antibodies were used as crude hybridomas culture supernatants. Secondary antibodies used in this study were as follows: fluorescein-conjugated goat anti-mouse or rabbit antibodies were from Sigma (Sigma Chemical Co., St. Louis, MO) and used at a 1/200 dilution. Horseradish peroxidase goat anti-mouse or rabbit antibodies were from Promega (Promega Corporation, Madison, WI) and used at a 1/5,000 dilution. Goat anti-mouse or rabbit antibodies conjugated with 15-nm colloidal particles were purchased from Sigma and used at 1/200. Electrophoresis and Immunoblotting Techniques SDS-PAGE was performed using 15% acrylamide gels under standard conditions (Laemmli, 1970). Onedimensional gels were routinely stained using Coomassie Blue. For 2-D gels, isoelectric focusing was performed by the method of O’Farell  on 4% acrylamide gels containing 0.8% of 3–10 ampholines and 1.2% of 4–6 ampholines (Pharmacia, Uppsala, Sweden). Seconddimension gels were silver stained according to Merril et al. . Western transfers of material from gels to PVDF membranes (Millipore, Bedford, MA) followed the protocol of Towbin et al. . Before immunolabeling, membranes were shaken for 30 min in PBS (10 mM sodium phosphate, pH 7.2, 0.17 M NaCl) containing 5% w/v dry milk, then briefly washed in PBS plus 0.5% dry milk before incubation for 1 h with primary antibodies. Excess of primary antibody was removed by washes in PBS-0.5% nonfat dry milk before incubation for 1 h with horseradish peroxidase-conjugated second antibody. After removal of excess secondary antibody by copious washes, the reaction was either developed using 0.001% hydrogen peroxide and 0.5 mg/ml diamino benzidine as substrates or processed for chemoluminescence labeling using an ECL kit according to the instructions of the manufacturer (Amersham International, Amersham, England). Immunofluorescence For immunofluorescence microscopy, starved cells were deciliated by successive transfers in 2% (w/v) manganese chloride and permeabilized for 2 min in PHEM buffer (60 mM Pipes, 25 mM Hepes, 10 mM EGTA, 2 mM MgCl2) containing 1% Triton X-100. Cells were fixed for 30 min with 2% (w/v) paraformaldehyde in PHEM buffer, copiously washed in PHEM and then incubated in TBS-T (10 mM Tris-HCl, pH 7.4, 0.15 M NaCl, 0.1% [v/v] Tween 20) containing 3% (w/v) bovine serum albumin (BSA). TBS-T buffer was used in all steps: incubation with primary antibody, washes, incubation with secondary antibody, and final washes. Cells were finally mounted in 50% (v/v) glycerol and observed under an Olympus BH2 epifluorescence microscope. Electron Microscopy Immunolocalizations at the ultrastructural level were performed using the colloidal gold method. Cells were aldehyde fixed for 1 h in 3.6% glutaraldehyde plus 0.4% paraformaldehyde in 0.1M sodium phosphate buffer pH 7.2, at 4°C in the presence (Nassula) or absence (Furgasonia) of 0.2% tannic acid [Begg et al., 1978]. Cells were then dehydrated and infiltrated with the polar embedding medium Lowicryl K4M at ⫺35°C. The resin was then polymerized overnight by U.V. irradiation at ⫺35°C. Thin sections were collected on nickel grids, saturated with 3% (w/v) BSA in PBS and processed for immunogold labeling as described elsewhere [Vigues and Grolière, 1985]. For conventional electron microscopy, cells were fixed for 30 min in 2% glutaraldehyde and post-fixed in 1% OsO4 both at 4°C in phosphate buffer. Samples were embedded in Epon 812. Sections were examined under a Jeol JEM-1200X transmission electron microscope. RESULTS Immunofluorescence Localization of a Cytopharyngeal Centrin-Immunoanalog in the Ciliates Nassula and Furgasonia Indirect immunofluorescence labeling was performed on aldehyde-fixed cells previously deciliated and permeabilized in the presence of the non-ionic detergent Triton-X100. Figure 1A shows the typical fluorescence pattern given by anti-tubulin antibody in these conditions. Cilia have disintegrated but basal bodies and the feeding basket persist following this treatment. The staining pattern of Nassula cells by A69 monoclonal antibody is shown in Figure 1B. Immunoreactive material is restricted to the anterior part of the basket. This material forms adjacent polygonal meshes that seem precisely fitted to the transverse section of nemadesmata. Identical Centrin in the Feeding Basket of Nassulinid Ciliates 75 Fig. 1. Immunofluorescence microscopy of Triton-X100 permeabilized, deciliated cells of Nassula (A, B, and C) and Furgasonia (D and E) stained with anti-tubulin (A), mAb A69 (B and D), and anti-EEB antiserum (C and E). Tubulin labeling mainly decorates basal bodies and microtubular components of the basket (*). mAb A69 and anti-EEB antiserum provide similar staining patterns, resulting in a collar of adjacent polygonal meshes near the top of the basket. Scale bars: A ⫽ 40 µm; B ⫽ 15 µm; C ⫽ 7 µm; D and E ⫽ 10 µm. results were obtained when cells were treated by antiEEB antiserum (Fig. 1C). From a series of micrographs of N. aurea cells processed for immunofluorescence microscopy, we calculated a number of adjacent polygonal meshes in agreement with the number (variable; from 25 to 34) of nemadesmata in this species. The basket structure, which reacted in Nassula, was also successfully identified in cells of Furgasonia using the two antibodies (Fig. 1D, E). The size of this structure is different due to the lower size of Furgasonia cells. Fifteen to eighteen polygonal adjacent meshes could be counted on the more characteristic micrographs, which, here again, corresponds to the number of nemadesmata in F. blochmanni. For each ciliate, control experiments were performed in which the primary antibody was omitted (mAb A69) or replaced by the corresponding preimmune serum (anti-EEB antiserum). Under these conditions the secondary antibodies failed to detect any cell structure (data not shown). oped. In the procedure used, cells of Nassula were aldehyde-fixed in the presence of tannic acid, a method that allows a better visualization of filamentous structures [Begg et al., 1978]. For quantitative evaluation of the antigen(s), a positive control was run using an antitubulin antibody previously found to afford strong labeling of microtubular structures in ciliates. When thin sections of the anterior part of the basket were treated with this antibody, gold particle deposits were found on microtubules of nemadesmata (Fig. 2). Crest microtubules, basal bodies and cilia were also stained by the antibody (not shown). Figure 3A and B shows sections of the same region of the basket stained by anti-EEB antiserum and mAb A69, respectively. With both antibodies, the label was confined to the area occupied by filamentous material of the apical contractile annulus, a filamentous structure initially described by Tucker , and thought to mediate constriction of the basket during feeding movements. It is noteworthy that labeling of the filamentous annulus was comparable in intensity to that of nemadesmata stained with anti-tubulin antibody. We also note that many filaments abut against the wall of nemadesmata. No significant staining was observed in control experiments in which the first antibody was left out or replaced by preimmune serum (not shown). Similar ultrastructural labelings were obtained in Furgasonia cells (Fig. 4). In this case, tannic acid was absent from the fixative solution, resulting in noticeable loss of definition with respect to apical annulus substructure. The discrepancy is particularly evident when comparing, for instance, ultrastructure of the apical annulus in either of the two ciliates (compare, e.g., Fig. 3B with Fig. Immunoelectron Localization of Anti-EEB Antiserum and mAb A69 Immunoreactivity A variety of methods were experimented with to localize anti-EEB/mAb A69 antibody immunoreactivity at the ultrastructural level. All attempts to use preembedding labeling procedures such as those described in previous studies [see, e.g., Vigues et al., 1984; Garreau de Loubresse et al., 1988] were unsatisfactory because gold particles, even of very small diameter, were unable to penetrate into the bulk of cytoskeletal structures found in the anterior part of the basket. For this reason, a thin section post-embedding labeling procedure was devel- 76 Vigues et al. Fig. 2. Immunocytochemical labeling of the anterior part of Nassula feeding basket using anti-tubulin antibody. The section cuts through the top of nemadesmata (nd) and shows the filamentous system termed the apical filamentous annulus (fa). Nemadesmata are heavily stained; the latter pass through the annulus and are connected to each other by the filamentous material. Note that most of the filaments are attached by end-on associations to the sides of nemadesmata (arrows). Inset: Schematic representation of a resting basket showing the position of the annulus (arrow) with regard to the palisade of nemadesmata. Scale bar ⫽ 0.5 µm. 4B). On the other hand, electron micrographs obtained in these conditions revealed an interesting aspect of microtubule-filament structural interactions. As shown in Figure 4B, plates of electron dense material are interposed between nemadesmata and filaments of the annulus. These plates were consistently found in the zones of contact and absent lower down in the basket, suggesting a role of the dense material in the attachment of filaments to the wall of nemadesmal microtubules. All the other bands were faintly stained by conventional procedures. Small-sized polypeptides were notably present in fairly low proportions, in a narrow MW range between 16 and 20 kDa. When Western blotting was carried out on Nassula cell ghost proteins, anti-EEB antiserum and mAb A69 both reacted with the same polypeptide at 20 kDa (Fig. 5C). The signal given by chemoluminescence was slightly less intense in the case of the monoclonal antibody. In both cases, an additional band was also evident, though weakly stained, near the bottom of the gel. For further biochemical characterization of the 20-kDa antigen, 2% Triton X-100 was used in the extraction buffer, a method that generally disintegrates the pellicle leading to an enrichment of oral structures [Wolfe, 1970]. Feeding basket–enriched preparations obtained in such conditions were solubilized in 6 M urea-2% Nonidet P-40. Soluble proteins were then concentrated by vacuum dialysis, resuspended in IEF sample buffer and processed for 2-D gel electrophoresis as described in Materials and Methods. In these conditions, the 20-kDa antigen was split into two spots easily distinguishable by their isoelectric point (Fig. 5D,1). Biochemical Characterization of Filamentous Annulus Antigen As shown in Figure 5A, cell ghosts obtained in the presence of 1% Triton X-100 are devoid of cilia but retain the feeding basket. The electrophoretic pattern of Nassula cell ghosts is shown in Figure 5B. Even though cilia have disintegrated, tubulins occur in fairly high proportion, probably owing to the persistence of basal bodies, associated fibers, and microtubules of the basket. The band of highest intensity migrates with an apparent MW of 33 kDa and probably corresponds to a major protein component of the pellicle. Five bands of medium intensity are evident with MW of 60, 50, 45, 43, and 35 kDa. Fig. 3. Immunogold labeling of the anterior part of the Nassula feeding basket using anti-EEB antiserum (A) and A69 monoclonal antibody (B). Both antibodies produce intense decoration of the apical filamentous annulus (fa and arrows). Nemadesmata are not stained, so that the pattern provided could be ‘‘negatively’’ imaged in electromicrographs obtained with anti-tubulin antibody (compare with Fig. 2). Scale bars ⫽ 0.3 µm. Inset: Schematic representations of the feeding process steps in Nassula [Tucker, 1968]. The top of the basket dilates (1) as it is closely applied to the filamentous bacteria (fb). Subsequent feeding movements include bending of the filamentous bacteria into a hair-pin shape (2), and then constriction of the top of the basket, mediated by the apical annulus (3). 78 Vigues et al. Fig. 4. Apical filamentous annulus in Furgasonia as viewed by conventional (A) and immuno-electron microscopy (B,C) following staining with anti-EEB (B) and mAb A69 (C). As shown in Nassula, gold particles are found in regions occupied by filaments of the annulus (fa). Labeling of the annulus is slightly less intense with mAb A69. Note that microtubules of nemadesmata (nd) are embedded in electron dense material (arrowheads) as they pass through the filament annulus. Accumulation of such dense material is evident in the zones of contact with filaments of the apical annulus. Scale bars: A ⫽ 0.5 µm; B and C ⫽ 1 µm. Centrin in the Feeding Basket of Nassulinid Ciliates 79 Fig. 5. Anti-EEB antiserum and mAb A69 are both reactive with the same 20-kDa protein. A: Light micrograph of a cell ghost obtained from Nassula. Cilia have disintegrated. The feeding basket is conserved (arrow) and retains most if not all the nemadesmata (arrowheads). Scale bar ⫽ 35 µm. B: SDS-PAGE, gels were stained with Coomassie Brilliant Blue. Lane 1: Standard proteins from Bio-Rad. Lane 2: Nassula cell ghost protein components. The main bands are represented by tubulins (tu) and a 33-kDa polypeptide (*). C: Western blot analysis. Cell ghost proteins were transferred to filter membranes then reacted with anti-EEB antiserum (lane 1) or mAb A69 (lane 2). In the two cases, a 20-kDa band was antibody-stained. D: Portions of 2-D Western blots of the 20-kDa antigen. 1: Silver staining. Note that the antigen is revealed as two spots with acidic pI between 4.8 and 4.9. 2: Immunodetection with anti-EEB antiserum. The two spots are reactive to the antiserum. 3: Immunodetection with mAb A69. Only spot a, which is the more acidic one, is reactive to the monoclonal antibody. Spot b is not stained. Immunoblots performed on replicates of 2-D electrophoregrams gave different results depending on the antibody tested. Using anti-EEB antiserum, isoelectric variants A and B were both immunoreactive as shown in Figure 5D,2. Using mAb A69 (Fig. 5D,3), immunoreactivity was restricted to variant A, i.e., the more acidic isoelectric variant of the 20-kDa protein antigen, suggesting that for mAb A69 immunoreactivity, variants A and B are antigenically distinct polypeptides. mAb A69 revealed two main arrays of centrin in Entodinium cytoalimentary system: a contractile lip-like structure and a filamentous sheath extending between and parallel to the nemadesmata [Vigues and David, 1994]. The latter location was not found in this study. The filamentous annulus might be the functional equivalent of the contractile lip that contributes to ingestion of large plant tissue fragments by entodiniomorphids. We also note that centrin fibers may be associated with the phagocytic process in some unicellular algae. However, this cannot be directly correlated to the situation found in Nassulinids since the role of these fibers seems limited to food particle capture by flagella or more specialized organelles as shown for instance in Ochromonas and Pseudopedinella [Melkonian et al., 1992]. The biological function of the 20-kDa protein is not known. It is tempting to speculate that it mediates contraction of the filamentous annulus, possibly in response to an increase in local calcium concentration. From observation of living feeding cells of Nassula and ultrastructural studies of the basket, it has been proposed that feeding movements result from antagonistic action of crest microtubules and filaments of the apical annulus [Tucker, 1968, 1978]. Whereas sliding of adjacent microtubules of the crests increases the distance between adjacent nemadesmata and thereby dilates the top of the basket, contraction of the fibrous annulus has the opposite effect. If this interpretation is correct, our results indicate the fibrous annulus is a specialization of centrin-mediated cell motility designed to act antagonistically to a microtubule-sliding type of movement, a dual system that might DISCUSSION In this study we identified a 20-kDa antigen shared by the two ciliates Nassula and Furgasonia and immunoreactive to antibodies raised against and specific for non-centriolar forms of centrin in ciliated protozoans [see, e.g., Garreau de Loubresse et al., 1988, 1991; David and Vigues, 1994]. By indirect immunofluorescence microscopy, it is shown that the 20-kDa protein is restricted to the cytopharyngeal basket, a complex microtubular-filamentous organelle specialized for ingestion of filamentous cyanobacteria. Immunoelectron microscopy shows the 20-kDa protein in the apical filamentous annulus that interconnects microtubular fibers (nemadesmata) in the upper part of the basket. A centrin immunoanalog in association with the cytoalimentary system of ciliates has been found in previous studies on Entodinium and its close relatives. In the so-called entodiniomorphids, a cytopharyngeal specialization occurs below the cytostome [see, e.g., Furness and Butler, 1988], reminiscent of the feeding basket in Nassulinids. Labeling with 80 Vigues et al. account for extreme changes in cell body length in the large Heterotrich ciliates Stentor and Spirostomum [Huang and Mazia, 1975]. In this respect, it is worth noting that the contractile fibers (myonemes) that effect shortening of Stentor and Spirostomum have been shown to be composed of spasmin [Ochiai et al., 1989], another calciumsensitive protein closely related to algal and mammalian centrins [Melkonian et al., 1992; Levy et al., 1996]. Detectable amounts of the 20-kDa protein are not present in regions occupied by basal bodies, which are sites where one would also expect to find centrins in ciliated Protozoa (see below). The simplest interpretation that basal bodies of Nassula and Furgasonia lack centrin is in obvious conflict with current concepts of evolutionary conservation of centriolar proteins. Since kinetosomal centrin has been demonstrated in other ciliates, including when cells were prepared for IF as in this study [see, e.g., Klotz et al., 1997], it is conceivable that the failure of our antibodies to react with kinetosomes is due to their specificity for non-kinetosomal forms of this protein. Several reports describe heterogeneity in the current centrin protein family; some based on the observation of multiple bands or spots on one- or two-dimensional protein gels [Vigues and Grolière, 1985; Garreau de Loubresse et al., 1991; Levy et al., 1996; Paoletti et al., 1996], others based on direct examination of centrin genes themselves [Madeddu et al., 1996; Middendorp et al., 1997]. On the basis of sequence comparison, two divergent subfamilies of centrin proteins have been proposed in mammalian cells possibly linked to their implication in distinct centrosome-associated functions [Middendorp et al., 1997]. The discriminative properties of our antibodies could well reflect a similar dichotomy between centrins involved in filamentous contractile structures and kinetosomal forms of the protein. Consistent with this hypothesis, centrins previously characterized by mAb A69 and anti-EEB antibody have been found in myonemes and/or related contractile fiber systems [Vigues et al., 1984; Vigues and Grolière, 1985; Garreau de Loubresse et al., 1988; Hulays et al., 1991; David and Vigues, 1994] and were absent from kinetosomes as shown in the present study. With the increasing availability of centrin genes in ciliates, it would be interesting to know where the epitopes reside on nonkinetosomal forms of centrin since this might provide further insight into the molecular mechanisms that contribute to the functional diversity of these proteins in a singled-cell organism. Two isoeletric variants were disclosed in the case of the 20-kDa protein, a situation that is more reminiscent of that reported for centrin in algal flagellar apparatus. As far as we know, algal centrin is a single gene product that may exist as two isoforms that differ by the presence of a phosphate group, Ca⫹⫹-binding being able to induce dephosphorylation of centrin in vivo [Salisbury et al., 1984]. Microheterogeneity of the 20-kDa protein could well correspond to post-translational modification such as phosphorylation, possibly associated with Ca⫹⫹-binding activity. The suggestion of a unique type of posttranslationally regulated centrin in Nassulinid feeding basket, however, conflicts with the following observations: first, Salisbury and coworkers  showed that the electrophoretic mobility of centrin is substantially increased by the coupled effect of Ca⫹⫹-binding and dephosphorylation, whereas in our experiments, variants A and B were consistently found to be indistinguishable otherwise than by their net charge as revealed by isofocalization. Second, for mAb A69 immunoreactivity, we show here that variants A and B are antigenically distinct polypeptides. Variant A stains with mAb A69 but variant B does not. It cannot, however, be inferred that mAb A69 epitope is absent from variant B: the epitope could be modified and/or masked in a special mode upon phosphorylation. On the other hand, assuming one of the two variants is a phosphorylated form of the 20-kDa protein, one would expect pretreatment of feeding baskets with dephosphorylating agents to restore immunoreactivity of variant B to mAb A69. We verified that pretreatment with alkaline phosphatase, for instance, had no apparent effect either on immunoreactivity or on the 2-D gel pattern of the 20-kDa protein bands (data not shown). The relationship of the two proteins disclosed on 2-D gels is, therefore, still unknown and further work is required to determine whether the two proteins are derived one from the other via post-translational modifications, or correspond to different gene products coexpressed in the same cell. Finally, the results presented in this study extend our knowledge about the occurrence and functions of centrins in different organisms. More interestingly, they raise the question of whether physiological processes relying on centrin-microtubule interactions may be a more widespread phenomenon, particularly in mammalian centrosomes. This possibility is especially intriguing since studies by different groups have identified centrin as a permanent component of the pericentriolar matrix, which is the main nucleating and anchoring site for centrome-assembled microtubules. A good way for mammalian cells to control the number and orientation of microtubules of the cytoplasmic microtubule complex may be to influence and possibly modify the distribution of their anchoring sites in the centrosome region. The centrin array found in the pericentriolar matrix, whose overall conformation can be altered by calcium in vitro may have properties compatible with such a function [Moudjou and Bornens, 1992; Paintrand et al., 1992; Baron et al., 1994]. Whether physical links between centrin containing filaments and microtubules, which are Centrin in the Feeding Basket of Nassulinid Ciliates particularly evident in the ciliate feeding basket, have structural counterparts in mammalian centrosome remains to be determined. Biochemical and immunological characterization of electron-dense cytopharyngeal material is in progress to explore this possibility. ACKNOWLEDGMENTS We gratefully acknowledge Dr. J. C. Roumagoux for his kind help in the large-scale preparation of ciliate cultures, Mrs M. Bérard and M. Dion for excellent technical assitance, and Mr. J. L. Vincenot for the photographic work. REFERENCES Baron AT, Greenwood T, Bazinet C, Salisbury JL. 1992. Centrin is a component of the pericentriolar lattice. Biol Cell 76:383–388. Baron AT, Suman V, Nemeth E, Salisbury JL. 1994. The pericentriolar lattice of ptK2 cells exhibits temperature and calciummodulated behavior. J Cell Sci 107:2993–3003. Begg DA, Rodewald R, Rebhun LI. 1978. The visualization of actin filament polarity in thin sections. evidence for the uniform polarity of membrane-associated filaments. J Cell Biol 79:846–852. Bouchard P, Vigues B, Ruchaud MH, Ravet V. 1998. The membrane skeleton of Tetrahymena contains immunoanalogs of GFAP, the intermediate filament protein expressed in astrocytes and cells of glial origin. Eur J Protistol 34:138–147. David C, Vigues B. 1994. Calmyonemin: a 23 kDa analogue of algal centrin occuring in contractile myonemes of Eudiplodinium maggii (ciliate). Cell Motil Cytoskeleton 27:169–179. Eisler K. 1988. Electron microscopical observations on the ciliate Furgasonia blochmanni Fauré-Frémiet, 1967. Eur J Protistol 24:75–93. Errabolu R, Sanders MA, Salisbury JL. 1994. Cloning of cDNA encoding human centrin, and EF-hand protein of centrosome and mitotic spindle poles. J Cell Sci 107:9–16. Furness DN, Butler RD. 1988. The functional and evolutionary significance of ultrastructure of the Ophryoscolecidae (Order Entodiniomorphida). J Protozool 35:34–38. Garreau De Loubresse N, Keryer G, Vigues B, Beisson J. 1988. A contractile network of Paramecium: the infraciliary lattice. J Cell Sci 90:351–364. Garreau De Loubresse N, Klotz C, Vigues B, Rutin J, Beisson J. 1991. Ca⫹⫹-binding proteins and contractility of the infraciliary lattice in Paramecium. Biol Cell 71:217–225. Huang B, Mazia D. 1975. Microtubules and filaments in ciliate contractility. In: Inoué S, Stephens RE, editors. Molecules and cell movement. New York: Raven Press. p 389–410. Hulays S, Grain J, Vigues B. 1991. Changes in the organization of the microfilamentous cortical cytoskeletal system during oral regeneration in Climacostomum virens (ciliate). Biol Cell 72:67–74. Klotz C, Garreau De Loubresse N, Ruiz F, Beisson J. 1997. Genetic evidence for a role of centrin-associated proteins in the organization and dynamics of the infraciliary lattice in Paramecium. Cell Motil Cytoskeleton 38:172–186. Laemmli UK. 1970. Cleavage of structural protein during the assembly of the head of bacteriophage T4. Nature 227:680–685. Lee VD, Huang B. 1993. Molecular cloning and centrosomal localization of human caltractin. Proc Natl Acad Sci USA 90:11309–11043. 81 Levy YY, Lai EY, Remillard SP, Heintzelman MB, Fulton C. 1996. Centrin is a conserved protein that forms diverse associations with centrioles and MTOCs in Naegleria and other organisms. Cell Motil Cytoskeleton 33:298–323. Madeddu L, Klotz C, Le Caer JP, Beisson J. 1996. Characterization of centrin genes in Paramecium. Eur J Biochem 238:121–128. Melkonian M, Beech PL, Katsaros C, Schulze D. 1992. Centrinmediated cell motility in algae. In: Melkonian M, editor. Algal cell motility. New York: Chapman and Mall. p 179–221. Merril CR, Dunau ML, Goldman D. 1981. A rapid sensitive silver stain for polypeptides in polyacrylamide gels. Anal Biochem 110:201–207. Middendorp S, Paoletti A, Schiebel E, Bornens M. 1997. Identification of a new mammalian centrin gene, more closely related to Saccharomyces cerevisiae CDC31 gene. Proc Natl Acad Sci USA 94:9141–9146. Moudjou M, Paintrand M, Vigues B, Bornens M. 1991. A human centrosomal protein is immunologically related to basal-body associated proteins from lower eukaryotes and is involved in the nucleation of microtubules. J Cell Biol 115:129–140. Moudjou M, Bornens M. 1992. Le centrosome est-il une structure dynamique? C R Acad Sci Paris 315:527–534. Ochiai T, Kato M, Ogawa T, Asai H. 1988. Spasmin-like proteins in various ciliates revealed by antibody to purified spasmins of Carchesium polypinum. Experientia 44:768–771. O’Farell PM. 1975. High resolution two-dimensional electrophoresis of proteins. J Biol Chem 250:4007–4021. Paintrand M, Moudjou M, Delacroix H, Bornens M. 1992. Centrosome organization and centriole architecture: their sensitivity to divalent cations. J Struct Biol 108:107–128. Paoletti A, Moudjou M, Paintrand M, Salisbury JF, Bornens M. 1996. Most centrin in animal cells is not centrosome-associated and centrosomal centrin is confined to the distal lumen of centrioles. J Cell Sci 109:3089–3102. Salisbury JL. 1989. Centrin and the algal flagellar apparatus. J Phycol 25:201–206. Salisbury JL, Baron AT, Surek B, Melkonian M. 1984. Striated flagellar roots: isolation and partial characterization of a calciummodulated contractile organelle. J Cell Biol 107:641–653. Spang A, Courtney I, Facker U, Matzner M, Schiebel E. 1993. The calcium binding protein cell division cycle 31 of Saccharomyces cerevisiae is a component of the half bridge of the spindle pole body. J Cell Biol 123:405–416. Towbin H, Staehelin T, Gordon J. 1979. Electrophoretic transfer of proteins from polyacrylamide gels to nitrocellulose sheets: procedure and some applications. Proc Natl Acad Sci USA 76:4350–4354. Tucker JB. 1968. Fine structure and function of the cytopharingeal basket in the ciliate Nassula. J Cell Sci 3:493–514. Tucker JB. 1978. Endocytosis and streaming of highly gelated cytoplasm alongside rows of arm-bearing microtubules in the ciliate Nassula. J Cell Sci 29:213–232. Vigues B, David C. 1994. Calmyonemin: Identification and distribution during the cell cycle in Entodinium bursa (ciliate). Biol Cell 82:121–127. Vigues B, Grolière CA. 1985. Evidence for a Ca⫹⫹-binding protein associated to non-actin microfilamentous systems in two ciliated protozoans. Exp Cell Res 159:366–376. Vigues B, Metenier G, Groliere CA. 1984. Biochemical and immunological characterization of the microfibrillar ecto-endoplasmic boundary in the ciliate Isotricha prostoma. Biol Cell 51:67–78. Wolfe J. 1970. Structural analysis of basal bodies of the isolated oral apparatus of Tetrahymena pyriformis. J Cell Sci 6:670–700.