Cell Motility and the Cytoskeleton 39:318–330 (1998) Tubulin Polyglycylation in Platyhelminthes: Diversity Among Stable Microtubule Networks and Very Late Occurrence During Spermiogenesis Carlo Iomini,1 Marie-Hélène Bré,2 Nicolette Levilliers,2 and Jean-Lou Justine1* 1Laboratoire de Biologie Parasitaire, Protistologie, Helminthologie, ERS 156 CNRS, Muséum National d’Histoire Naturelle, Paris, France 2Laboratoire de Biologie Cellulaire 4, URA 2227 CNRS, Université de Paris-Sud, Orsay, France The distribution of glycylated tubulin has been analyzed in different populations of stable microtubules in a digenean flatworm, Echinostoma caproni (Platyhelminthes). Two cellular types, spermatozoa and ciliated excretory cells, have been analyzed by means of immunofluorescence, immunogold, and immunoblotting techniques using two monoclonal antibodies (mAbs), AXO 49, and TAP 952, specifically directed against differently glycylated isoforms of tubulin. The presence of glycylated tubulin in the two cell types was shown. However, the differential reactivities of TAP 952 and AXO 49 mAbs with the two axoneme types suggest a difference in their glycylation level. In addition, within a single cell, the spermatozoon, cortical microtubules underlying the flagellar membrane, and axonemal microtubules were shown to comprise different tubulin isoforms, the latter ones only being labelled with one of the antiglycylated tubulin mAbs, TAP 952. Similarly, the antiacetylated (6-11B-1) and polyglutamylated (GT335) tubulin mAbs decorated the two types of axonemal microtubules, but not the cortical ones. From these data, a subcellular sorting of posttranslationally modified tubulin isoforms within spermatozoa, on the one hand, and a cellular sorting of glycylated isoforms inside the whole organism, on the other hand, is demonstrated in the flatworm E. caproni. Last, a sequential occurrence of tubulin posttranslational modifications was observed in the course of spermiogenesis. Acetylation appears first, followed shortly by glutamylation; glycylation takes place at the extreme end of spermiogenesis and, specifically, in a proximo-distal process. Thus in agreement with, and extending other studies [Bré et al., 1996], glycylation appears to close the sequence of posttranslational events occurring in axonemal microtubules during spermiogenesis. Cell Motil. Cytoskeleton 39:318–330, 1998. r 1998 Wiley-Liss, Inc. Key words: posttranslational modifications; polyglycylation; stable microtubules; spermatozoa; Platyhelminthes INTRODUCTION In most eukaryotic cells, the diversity of the various microtubule subpopulations is increased not only by the presence of a- and b- tubulin isotypes encoded by different multigenic families, but also by posttranslational modifications (PTMs) of primary gene products. r 1998 Wiley-Liss, Inc. *Correspondence to: Jean-Lou Justine, Laboratoire de Biologie Parasitaire, Protistologie, Helminthologie, Muséum National d’Histoire Naturelle, 61 rue Buffon, F-75231 Paris cedex 05, France. E-mail: firstname.lastname@example.org Received 16 December 1997; accepted 23 December 1997 Tubulin Polyglycylation in Stable Microtubules These PTMs can be subdivided into two classes. One class includes the addition/removal of (an) amino acid(s), or the chemical modification of a residue of the polypeptide chains. With the exception of acetylation [L’Hernault and Rosenbaum, 1985; LeDizet and Piperno, 1987] and palmitoylation [Caron, 1997; Ozols and Caron, 1997], PTMs, such as detyrosination [Barra et al., 1974; Thompson, 1982], excision of the penultimate glutamate residue [Paturle-Lafanechère et al., 1991], and phosphorylation [Eipper, 1974; see Khan and Ludueña, 1996], occur in the carboxy-terminal region of tubulin. The second class, comprising the polymodifications, has been characterized by the lateral branching of a polyglutamyl [Eddé et al., 1990], or a polyglycyl [Redeker et al., 1994] chain of variable length on a specific glutamate residue near the carboxy-terminal end of the tubulin subunits. A wide panel of isoforms is thus generated by these polymodifications. The role of the tubulin PTMs is poorly known; nevertheless, a general but not universal correlation has been established between the stability of microtubules and the extent of modification [for reviews, see Greer and Rosenbaum, 1989; Gelfand and Bershadsky, 1991]. Accordingly, the polyglutamyl chain length has been assumed to regulate interactions of tubulin with microtubuleassociated proteins (MAPs) [Boucher et al., 1994]. In addition, the polymodifications have been hypothesized to play a role in spermatozoon motility [Bré et al., 1996; Gagnon et al., 1996]. Using specific antibodies, we have investigated the occurrence of tubulin PTMs, and particularly polyglycylation, in Echinostoma caproni, a digenean flatworm belonging to the phylum Platyhelminthes. This biological model allowed us to address the question of the distribution of glycylated isoforms among stable microtubules differently located inside a single organism. The access to germ and excretory cells has permitted analysis of two cell types, either flagellated (spermatozoa) or ciliated (protonephridia), both comprising motile axonemes arranged in different patterns (9 1 ‘‘1’’ and 9 1 2, respectively) [Iomini and Justine, 1997]. In addition, the co-existence in the spermatozoon of E. caproni of axonemes and cortical microtubules allowed us to approach the question of the distribution of polyglycylation comparatively to some other PTMs in two types of stable microtubules—cortical and axonemal—closely located within a same cell. We showed that polyglycylation, as well as acetylation and polyglutamylation, were indeed present, but differently distributed in the three classes of microtubules. Moreover, two complementary tools, specifically directed against distinct levels of glycylation, were found to be differentially reactive with the 9 1 ‘‘1’’ and 9 1 2 axonemes. These results show that the PTMs allow 319 discrimination between different functional classes of stable microtubules inside a single organism or cell. Because in E. caproni the different stages of spermatid maturation are readily distinguishable, we could accurately follow the time occurrence of glycylation, comparatively to acetylation and glutamylation, in the course of spermiogenesis. This analysis revealed the sequential appearance of these PTMs, glycylation being the last one to take place, and delayed to the end of spermiogenesis. The results are discussed in regard to the mechanism of sorting and appearance of tubulin isoforms in a single or in two compartments of one organism. MATERIALS AND METHODS Animals, Tissue, and Cell Isolation Echinostoma caproni, a parasitic species from the intestine of a wild bird [Richard, 1964], has been adapted to laboratory mice and is widely used as a model [Fried and Huffman, 1996]. Specimens were collected from mice experimentally infested with 50 metacercariae. The worms were dissected in phosphate-buffered saline (PBS) at pH 7.4 in order to isolate either the testes or the seminal vesicle. Spermatozoa were extracted from the seminal vesicle with needles. Testes or spermatozoa were collected in PBS containing protease inhibitors [Bré et al., 1996]. Antibodies The monoclonal antibodies, TAP 952 and AXO 49, raised against Paramecium axonemal tubulin [Callen et al., 1994], specifically recognize different levels of tubulin polyglycylation [Bré et al., 1996]. AXO 49 ascitic fluid was prepared by Dr. Jeanmaire-Wolf. GT335 mAb, directed against polyglutamylated tubulin [Wolff et al., 1992], was a kind gift from Dr. P. Denoulet (Collège de France, Paris). 6-11B-1 mAb, directed against acetylated a-tubulin [Piperno and Fuller, 1985], was purchased from Sigma (France). DM1A and DM1B anti-a- and b-tubulin mAbs [Blose et al., 1984] were purchased from Amersham (France). C105 polyclonal antibody (Ab), directed against a carboxy-terminal sequence of b-tubulin [Arevalo et al., 1990], was kindly provided by Dr. J. M. Andreu (Centro de Investigationes Biologicas, Madrid, Spain). The FITC- and TRITC-conjugated goat antimouse and goat antirabbit IgG Abs were purchased from Nordic (France). The gold-conjugated goat antimouse IgG Ab was purchased from BioCell (France). Goat immunoglobulins were purchased from Sigma. The peroxidase-labelled goat antimouse IgG Ab was from Sanofi Diagnostics Pasteur (France). 320 Iomini et al. Immunofluorescence Germ cells were obtained by squashing testes or seminal vesicles in a drop of PBS on a pit slide. Slides were dried for 1 h under a fan, then kept at 4°C and processed within 24 h. The cells were treated with acetone for 10 min at room temperature and rinsed in PBS (3 3 5 min). After rinsing, nonspecific antigens were blocked with 2% bovine serum albumin (Sigma) in PBS (BSA-PBS) for 45–90 min at room temperature. The pits were then covered for 40 min with one antibody diluted in BSA-PBS: AXO 49 cell culture supernatant, 1:2, or ascitic fluid, 1:100; TAP 952, 1:10; 6-11B-1, 1:200; GT335, 1:500; DM1A, 1:200; C105, 1:50. After rinsing in PBS (3 3 5 min), a FITC goat anti-mouse IgG Ab (1:40 in PBS) was applied for 40 min at room temperature. For double labelling, the pits were successively covered for 40 min with one of the antitubulin PTM Abs cited above, then with C105 Ab, 1:50, then with FITCconjugated goat antimouse, and TRITC-conjugated goat antirabbit IgG Abs, both 1:40, followed each time by rinsing in PBS (3 3 5 min). Thereafter, nucleus staining was achieved using DAPI (1 µg/ml in PBS) for 10 min. After final rinsing in PBS (3 3 5 min), mounting was carried out in CITIFLUOR (Citifluor, UK), and slides were sealed with nail enamel. Controls were realized by omitting the primary Abs or by using nonrelevant mouse secondary Abs (they were negative and are thus not further mentioned or illustrated). Observations were made with a Nikon Optiphot epifluorescence microscope. Postembedding Electron Microscope Immunocytochemistry Living specimens cut into pieces were fixed for 1 h at 4°C in 2% glutaraldehyde in 0.1 M sodium cacodylate at pH 7.2. After rinsing in the same buffer (3 3 10 min), the worms were dehydrated in a graded ethanol series and embedded in LR White resin, medium grade (Pelanne Instruments, Toulouse, France). The resin was polymerized in tightly capped gelatin capsules for 10 h at 60°C. Ultrathin sections were placed on nickel or gold grids. Grids were rinsed (PBS, 3 3 5 min), and nonspecific antigens were blocked with goat Igs (1:30 in PBS) for 1 h. The grids were then incubated in a moist chamber for 40 min at room temperature with a primary antibody (AXO 49 ascitic fluid, 1:100; TAP 952, 1:4; 6–11B-1, 1:50; GT335, 1:500; DM1A, 1:100), diluted in PBS with goat Igs (1:100). After rinsing in PBS (3 3 5 min), a gold-conjugated goat antimouse IgG Ab (15 nm gold beads, 1:20 in PBS) was applied for 1 h at room temperature. After final rinsing in PBS (3 3 5 min, followed by distilled water, 3 3 5 min), the grids were dried on Whatman paper, stained according to Daddow , and observed with a Hitachi H600 electron microscope. Gel Electrophoresis and Immunoblotting A whole worm was cut and crushed with a pestle in a plastic tube containing liquid nitrogen and immediately boiled in Laemmli buffer for 4 min. Testes and spermatozoa were resuspended in the same buffer and similarly processed; proteins were separated in 10% polyacrylamide minigel-gels [Laemmli, 1970], containing 0.1% (w/v) SDS, 99% pure (BDH, U.K.) or 70% pure (Sigma), respectively, adjusted to pH 8.8 or 9.3. After electrophoresis, gels were stained with Coomassie brilliant blue R250 or electro-transferred onto nitrocellulose according to Kyhse-Andersen . The blots were stained with Ponceau red, washed in PBS, saturated with PBS containing 3% BSA and 0.1% Tween-20, and then incubated with TAP 952, 1:50, AXO 49 ascitic fluid, 1:5,000, GT335, 1:50,000, 6-11B-1, 1:30,000, DM1A, 1:5,000 or DM1B, 1:1,000, diluted in PBS containing 0.3% BSA and 0,1% Tween-20. After extensive washings with the latter buffer, blots were incubated with peroxidaselabelled goat antimouse IgG Ab, 1:2,000, and processed for enhanced chemiluminescence detection (ECL, Amersham). Some blots were reprobed with another primary antibody and similarly processed. RESULTS Distribution of Polyglycylated Tubulin in Various Classes of Stable Microtubules in Platyhelminthes Although tubulin was a minor polypeptide with respect to the whole protein amount of Echinostoma caproni (Fig. 1, lanes 1 and 2), TAP 952 exhibited a specific reactivity with two polypeptide bands migrating at the same level as the a- and b-tubulin subunits revealed by DM1A and DM1B. In contrast, a very weak staining was obtained with AXO 49 (Fig. 1), an observation that could mean that a- and b-tubulin subunits of E. caproni are only glycylated at low levels. To check this assumption, we have comparatively investigated two distinct axonemal systems expected to be enriched in polyglycylated tubulin, namely, flagella of mature spermatozoa and cilia of protonephridia. The seminal vesicle and the two testes were recovered by dissection in order to separate the mature spermatozoa from the spermiogenesis stages. The spermatozoon of E. caproni is a long and filiform motile cell, in which ultrastructural studies have demonstrated the presence of two parallel axonemes and cortical longitudinal microtubules. The axonemes display the typical 9 1 ‘‘1’’ pattern of parasitic Platyhelminthes, i.e., they are made up of the usual nine microtubule doublets and of a central core 600 nm in diameter [Iomini et al., 1995]. Tubulin Polyglycylation in Stable Microtubules Fig. 1. Immunoblot analyzis of proteins from the flatworm E. caproni. The proteins were submitted to SDS-PAGE and stained with Coomassie Blue (C B: lane 1, E. caproni proteins; lane 2, calf brain tubulin) or transferred onto nitrocellulose and incubated with DM1A (a-tubulin), DM1B (b-tubulin), TAP 952, and AXO 49 (glycylated tubulin). In order to improve the separation between alpha- and beta-tubulin, the migration time of proteins has been increased for immunoblotting compared to C.B. Double staining immunofluorescence (Fig. 2) as well as immunogold labelling (Fig. 3) revealed that spermatozoa were only reactive with TAP 952, but not with AXO 49. Similar reactivity characteristics were observed after immunoblotting of proteins from seminal vesicle spermatozoa (Fig. 4). Close inspection of the distribution of gold particles in transverse sections of spermatozoa showed that the TAP 952 labelling concerned only the peripheral doublets of axonemes but not the cortical microtubules (Fig. 3; see also Fig. 9). This allowed the interpretation of double labelling immunofluorescence observations (Fig. 2). With the C 105 antitubulin sequence Ab, the single heavy line corresponds to the axonemes and the thin more straight line to the cortical microtubules, separated from the axonemes by a fixation artefact. In contrast, with TAP 952, the thin line, corresponding to the cortical microtubules, is not revealed. In order to determine whether the specific PTM distribution found in spermatozoan microtubules concerned glycylation only, the reactivity of mAbs directed against other tubulin PTMs was also investigated by immunofluorescence and immunoelectron microscopy. 6-11B-1 and GT335 mAbs unequivocally labelled the peripheral doublets of the axonemes, but not the cortical microtubules, although the latter were sometimes weakly labelled with GT335 (Table I). It is noteworthy that, with 321 Fig. 2. Double labelling immunofluorescence of E. caproni spermatozoan microtubules with C105 (b-tubulin) (a, c) and TAP 952 (glycylated tubulin) (b) or AXO 49 (glycylated tubulin) (d). The arrow shows the cortical microtubules. Bar 5 50 µm. all antitubulin Abs tested at the ultrastructural level, the central core of the axonemes was never labelled (Fig. 3; see also Fig. 9). This observation strengthens the statement that it does not contain tubulin [Iomini et al., 1995], thus conferring on these axonemes a structural and molecular singularity. We took advantage of the presence in E. caproni of another cell type possessing the classical axonemal structure to check whether the exclusive reactivity of TAP 952 compared to AXO 49 might be a characteristic feature of the whole organism. The excretory system in most Platyhelminthes is constituted by protonephridia [Rohde, 1990]. These cells comprise a variable number of motile cilia arranged according to a 9 1 2 pattern. Postembedding immunocytochemistry experiments revealed a labelling of the peripheral doublets and central microtubules of the axonemes with both antibodies directed against glycylated tubulin, AXO 49 and TAP 952 (Fig. 3a, b, d; Table I). This suggests that a wide panel of glycylated isoforms is present in protonephridia, in contrast to spermatozoa. The exact number of protonephridia in E. caproni is unknown, but these cells represent a small percentage of the total body mass that can account for the weak staining observed after immunoblotting of total protein extracts with AXO 49 (Fig. 1). Nevertheless, one cannot exclude the possibility that some deglycylation might occur during the extract preparation. Concerning the other 322 Iomini et al. Fig. 3. Immunogold labelling of ultrathin sections of protonephridia (a, b, d, f) and spermatozoa (c, e, g) of E. caproni with AXO 49 (glycylated tubulin) (a, b, c), TAP 952 (glycylated tubulin) (d, e), and DM1A (a-tubulin) (f, g). (a) Longitudinal section of a protonephridium, showing its general morphology; inset, higher magnification. (b–g), Transversal sections. Bar 5 2 µm for (a), and 250 nm for inset and (b–g). PTMs, the ciliary axonemes, including the peripheral doublets and the two central microtubules of protonephridia, were likewise decorated with 6-11B-1 and GT335 mAbs (Table I). Occurrence of Glycylation in the Course of Spermiogenesis Figure 4 shows comparative immunoreactivities of protein extracts of testes and of spermatozoa, originating from the seminal vesicles, with the different antitubulin PTM Abs. Since very different protein amounts were present in both extracts (Fig. 4, lanes 1, 2), indicating various numbers of germ cells in the two compartments, their respective loads were adjusted using DM1A and DM1B reactivities as markers of their tubulin content. AXO 49 was unreactive with either tubulin. Whereas 6-11B-1 and GT335 reactivities were approximately similar for testis and seminal vesicle tubulin, TAP 952 exhibited a higher reactivity with the latter tubulin, suggesting that glycylation takes place progressively during spermiogenesis in the testes. Ultrastructural studies [Iomini and Justine, 1997] have shown that spermiogenesis in E. caproni follows a peculiar morphological pattern (diagrammed in Fig. 6). Each spermatid has a conical structure from which three cell processes grow and progressively fuse, a median cytoplasmic one, containing cortical microtubules, and two flagella. Using specific fluorescent markers of tubulin and DNA, the length of these processes and the position of the nucleus can be visualized, thus readily permitting identification of the stages of spermiogenesis. As illustrated in Figure 5, groups of cells with round nuclei (Fig. 5b) and short axonemes (Fig. 5a) correspond to early spermatids (Fig. 6a). The axonemes and the cortical microtubules lengthen (Figs. 5c, 6b) and fuse with the median process (Figs. 5e, 6c) so that two axonemes, surrounded by cortical microtubules, become incorporated within the spermatozoon (Figs. 5i, 6e). Simultaneously, the nuclei lengthen and migrate into the elongating spermatids (Figs. 5e, f, 6c), to be located eventually at the distal extremity of the late spermatids (Figs. 5g, h, 6d) and of the spermatozoa (Figs. 5i, j, 6e). In order to specify the time occurrence of glycylation during spermiogenesis, spermatids from the testes were analyzed by immunofluorescence with the antitubulin PTM antibodies (Fig. 7). Tubulin Polyglycylation in Stable Microtubules 323 ylated isoforms appear first in the proximal part of the axonemes and thereafter along the whole flagella to yield mature spermatozoon. As expected from the immunocytochemistry and immunoblotting data, none of the spermiogenesis stages within the testes was detected by AXO 49. DISCUSSION Fig. 4. Comparison of the extent of tubulin PTMs in protein extracts from testes and from seminal vesicle (SV) spermatozoa of E. caproni. For Coomassie Blue (C B), loads correspond to 0.7 testis (lane 1) or 0.7 vesicle (lane 2); for immunoblotting with DM1A (a-tubulin), DM1B (b-tubulin), 6-11B-1 (acetylated tubulin), GT335 (glutamylated tubulin), TAP 952 and AXO 49 (glycylated tubulin), loads correspond to 0.3 testis or 3 vesicles. The migration of calf brain tubulin is indicated in lane 3. 6-11B-1 mAb (Fig. 7b) revealed the early stages (Fig. 7a, c), and all other stages throughout. GT335 (Fig. 7e) also detected the early stages (Fig. 7d, f), but at that time, the decoration was restricted to the proximal part of the elongating spermatids (compare Fig. 7d, 7e); the whole length of the spermatids became labelled later (Fig. 7h), when the axonemes were elongating (Fig. 7g) and the nuclei migrating (Fig. 7i). The labelling intensity seemed to increase with the progression of spermiogenesis (Fig. 8), yielding fully labelled (Fig. 7k, n) late spermatids (Fig. 7j, l) and spermatozoa (Fig. 7m, o). The TAP 952 epitope (Fig. 7q) was not detected until the last stages of spermiogenesis, namely, when nuclear migration was completed (Fig. 7p, r); only the proximal part of the flagella was visible at this stage (compare p and q, Fig. 7). The whole length of the flagella was labelled in mature spermatozoa (Fig. 7t), once they were released from the common mass (Fig. 7s). The differential distribution of the TAP 952 epitope along the length of late spermatids was checked by immunoelectron microscopy (Fig. 9). In contrast to GT335, which stained the different parts of late spermatids (Fig. 9a–c), only sections of such spermatids devoid of nucleus, interpreted as proximal, were labelled with TAP 952 (Fig. 9d), but sections from the middle part of spermatids (Fig. 9e), as well as distal ones (Fig. 9f), were not labelled. This confirmed that glyc- A series of observations carried out with specific antibodies has allowed to infer that glycylation is widely spread among axonemal tubulins, from ciliated protozoa to Metazoa [Adoutte et al., 1991; Bressac et al., 1995; Levilliers et al., 1995], and is lacking in neuronal tubulin [Callen et al., 1994], in contrast to acetylation and glutamylation, which have been found in both tubulin types [see LeDizet and Piperno, 1991; for a review on acetylation, see Wolff et al., 1992; Bré et al., 1994; Fouquet et al., 1994]. The immunological tools have also permitted recognition of different levels of glutamylation in the mammalian cells [Audebert et al., 1993; Fouquet et al., 1994] and of glycylation in phylogenetically distant species ranging from ciliates to primates [Bré et al., 1996]. In parallel, mass spectrometry analyses have established that the number of glutamyl and of glycyl units added to the tubulin subunits varies considerably according to the organism analyzed [Eddé et al., 1990; Alexander et al., 1991; Redeker et al., 1992, 1994; Rüdiger et al., 1992, 1995; Mary et al., 1994, 1996; Multigner et al., 1996; Weber et al., 1996; Schneider et al., 1997]. This study, revealing the presence of both polymodifications in E. caproni, extends knowledge of their occurrence to the phylum Platyhelminthes. Both a- and b-tubulin subunits have been found to be modified. The distribution of these PTMs, and particularly glycylation, has been analyzed in different populations of stable microtubules in two cellular types of E. caproni, and their time occurrence has been followed during spermiogenesis. Three major results have been obtained: (1) in the spermatozoon, two distinct functional classes of stable microtubules comprise different tubulin isoforms, (2) in the same organism, tubulin constituting two types of axonemes has been shown, for the first time, to possess different levels of glycylation, and (3) tubulin glycylation takes place only at the final step of spermiogenesis, in contrast to the early occurrence of acetylation and glutamylation. PTMs Discriminate Between Two Different Functional Classes of Stable Microtubules Within a Single Cell Numerous studies have shown a correlation between the stability of microtubules and the presence of 324 Iomini et al. TABLE I. Labelling of Microtubules With Antibodies Directed Against Various Tubulin PTMs Cortical microtubules spermatozoa Axonemal microtubules spermatozoa protonephridia AXO 49 (glycylation) TAP 952 (glycylation) 6-11B-1 (acetylation) GT335 (glutamylation) 2 2 2 1/2 2 1 1 1 1 1 1 1 *1 labelling, 2 no labelling, 1/2 weak labelling sometimes observed. Fig. 5. Immunofluorescence staining of spermatid microtubules with DM1A (a-tubulin) (a, c, e, g, i), and DNA labelling (b, d, f, h, j) showing the time course of spermiogenesis in E. caproni. (a, b) Early spermatids; (c, d) elongating spermatids with nuclei still in the central mass; (e, f) distal part of elongating spermatids, showing the two axonemes (arrowheads), partially fused (arrow), and migrating nuclei; (g, h) late spermatids, with completely fused axonemes and distal nuclei; i, j: mature spermatozoon with distal nucleus. Bar 5 50 µm. Tubulin Polyglycylation in Stable Microtubules Fig. 6. Diagram of morphological pattern of spermiogenesis in E. caproni (adapted from Justine, 1991). Early spermatids (a) show three elongating processes, a median cytoplasmic process (M) and two flagella (F). The two flagella fuse with the median cytoplasmic process in a proximo-distal pattern (b, c) and, simultaneously, the nucleus (n) migrates into the distal extremity of the elongating spermatid (b, c, d). Finally, the spermatozoon (e), containing two incorporated axonemes, separates from the spermatid mass. Straight arrows indicate elongation, curved arrows indicate the movement of fusion of the flagella. Possible coiling of nucleus in the distal extremity is not pictured. PTMs, the latter thus reflecting a low microtubule turnover. However, it is worth comparing the distribution of PTMs among the most stable microtubule classes of a same cell in order to determine whether some differentiation can still be found among them. In mammalian spermatids, a differential PTM distribution has in fact been observed between two microtubule arrays, a transient one, that of the manchette, and a permanent one, that of the flagellar axoneme [Fouquet et al., 1994]. The simultaneous presence of cortical and axonemal microtubules in ciliated and flagellated cells provides an opportunity to analyze the PTM distribution among permanent stable microtubules. In the ciliate Paramecium, the three PTMs, acetylation, glutamylation, and glycylation, are located both in the ciliary axonemes and in most cortical microtubules [Bré et al., 1994; Fleury et al., 1995]. In contrast, in E. caproni spermatozoa, comprising two 9 1 ‘‘1’’ axonemes and cortical microtubules, the presence of the three PTMs is restricted to the axonemal doublets, excluding the neighbouring cortical microtubule singlets. Since both 325 microtubular systems are long-lived, their differential modifications cannot be explained by different microtubule dynamics. A masking effect of the three PTM epitopes by MAPs is also unlikely, given the distinct locations of these PTMs on the tubulin molecule. Rather, one mechanism could set in action a subcellular sorting of tubulin isotypes toward the two microtubular networks involved in different functions, as postulated in the multitubulin hypothesis [Fulton and Simpson, 1976]. As a matter of fact, a few cases of segregation and selective utilization of isotypes have been reported for the construction of axonemes. For example, in the Drosophila male germ line, the postmitotic b2 isotype, although multifunctional, i.e., required for the assembly of cytoplasmic microtubules as well as for the sperm axoneme doublet microtubules, has been shown to be indispensable for the right construction of the axoneme [see Raff, 1994]. In the framework of this hypothesis, one would have to consider an isotype-specific occurrence of PTMs, as demonstrated for phosphorylation [Khan and Ludueña, 1996]. As the extreme carboxy-terminal domain is the most divergent part of the tubulin molecule, one could assume, in E. caproni, the existence of nonglycylable and nonglutamylable cortical tubulin isotypes. However, the simultaneous absence of the latter PTMs and of acetylation in cortical microtubules would imply that at least some a-tubulin isotypes possess sequence divergence both in the carboxy-terminal and acetylation regions, as precisely found in a digenean platyhelminth [Duvaux-Miret et al., 1991] and in some other organisms [LeDizet and Piperno, 1987]. This assumption will be warranted only if a diversity of tubulin isotypes is demonstrated in E. caproni, both at the genetic and at the subcellular level. An alternative situation would involve a specific localization of the enzymes. Previous data about the immunoreactivity of AXO 49 mAb and of a related polyclonal Ab (PAT) in different systems such as Paramecium, quail oviduct and Drosophila testis [Adoutte et al., 1991; Bressac et al., 1995; Fleury et al., 1995] have led to postulation of an association of the glycylating enzyme to the membrane. This hypothesis is not in agreement with the present data, since in E. caproni the cortical microtubules that are lying in close contact with the membrane are decorated with none of the antitubulin PTM antibodies tested, in contrast to the axonemal ones. A possibility, then, is that the PTM enzymes might be specifically associated with the axoneme itself. Polyglycylation Discriminates Between Two Axonemal Systems in the Same Organism We have shown that the ciliary axonemes of E. caproni protonephridia, possessing the classical 9 1 2 structure, are reactive with both TAP 952 and AXO 49 mAbs, indicating the presence of low and high levels of Fig. 7. Double labelling immunofluorescence of spermatid microtubules with C105 (b-tubulin) (a, d, g, j, m, p, s) and 6-11B-1 (acetylated tubulin) (b) or GT335 (glutamylated tubulin) (e, h, k, n) or TAP 952 (glycylated tubulin) (q, t), and DNA staining (c, f, i, l, o, r), showing the sequential appearance of PTMs during spermiogenesis in E. caproni. Bar 5 25 µm. Tubulin Polyglycylation in Stable Microtubules 327 shows that evolutionary considerations may not be the sole factor involved in the specification of the glycylation level, contrary to suggestions from previous studies. In order to achieve the covalent binding of the glycine residues, at least two distinct enzymes are expected to be required: one for the linkage of the first glycine to a glutamate residue of the tubulin polypeptide chain by a gCOOH-aNH2 amide linkage, and a second for the formation of the aCOOH-aNH2 peptide bonds between the added glycine residues of the lateral chain. In this context, several types of mechanisms can be considered to account for these results. The most simple explanation involves a differential expression of the glycylating enzymes in the two cell types of E. caproni: the first catalyzing the linkage of the first Gly residue(s) to the tubulin polypeptide chain would be simultaneously present in protonephridia and spermatozoa, whereas the second would be expressed in protonephridia only. Another possibility concerns the regulation of the enzymes. In this context, a reverse enzyme and/or specific inhibitors, such as those previously reported for tubulin acetylation [Maruta et al., 1986] and glutamylation [Audebert et al., 1993], could intervene in the adjustment of the enzymatic equilibrium responsible for the extent of glycylation in each compartment. Therefore, the analysis of polyglycylation, by means of two complementary antibodies, permits postulation of, at least in E. caproni, a differential expression or regulation of the glycylating enzymes involved in two cell types specialized in different functions. In conclusion, whatever the mechanism involved for segregating the right isoforms at the right place, polyglycylation appears to act by two ways, either by its absence/presence or by modulating the number of glycyl units added to the tubulin molecule. Fig. 8. Double labelling immunofluorescence of microtubules from young and late spermatids of E. caproni, visible in the same field, with C105 (b-tubulin) (a) and GT335 (glutamylated tubulin) (b); DNA staining (c). Early (left) and late (right) spermatids show similar intensities of labelling with C105 (a), but with GT335 (b) the labelling intensity is stronger in late spermatids. A few mature strongly labelled spermatozoa, randomly located in the field, can be observed. Bar 5 50 µm. glycylation in this species. However, the lack of reactivity of AXO 49 with the 9 1 ‘‘1’’ flagellar axonemes of spermatozoa suggests the occurrence of only low levels of glycylation in the latter cell. This last result is reminiscent of that found in human spermatozoa [Bré et al., 1996], contrasting with all other species tested, and Sequential Occurrence of PTMs During Flagellar Elongation Could Be Involved in the Morphogenetic Clock of Spermiogenesis The accessibility of the axonemal structure in E. caproni flagella has permitted us to trace the appearance of PTMs during spermiogenesis, both by means of immunofluorescence and immunogold techniques. It is noteworthy that the combined use of TAP 952 and AXO 49 is necessary to follow the time occurrence of glycylation, instead of the exclusive use of AXO 49 [Bressac et al., 1995], specific to a restricted population of glycylated isoforms. Indeed, in Paramecium, a lag time is observed in the detection of the AXO 49 epitope, in contrast to the TAP 952 one, in new microtubular structures assembled in the course of cell division [Fleury et al., 1995]. 328 Iomini et al. Fig. 9. Immunogold labelling of ultrathin sections of late spermatids of E. caproni with GT335 (glutamylated tubulin) (a–c) and TAP 952 (glycylated tubulin) (d–f). Randomly obtained transverse sections at various levels of elongating spermatids were characterized according to Iomini and Justine . For instance, the sections (c) and (f) are recognized as distal by the simultaneous presence of the nucleus and a single axoneme. Bar 5 250 nm. This study in Platyhelminthes shows that various tubulin PTMs take place successively during the process of flagellar elongation, that acetylation and thereafter glutamylation appear precociously, and that these are followed after a great delay by glycylation. The fact that similar conclusions are reached from organisms belonging to evolutionarily very distant phyla—Platyhelminthes (this report) and Arthropoda [Bré et al., 1996]—lead the conclusion that this is a general phenomenon. The sequential aspect of the PTM appearance is clearly demonstrated in E. caproni. In this context, one might ask whether acetylation and glutamylation are prerequisite events for glycylation occurrence. The late detection of the AXO 49 epitope in the course of Drosophila spermiogenesis, at the time of the individualization process, i.e., when the axonemal structure comes in contact with the membrane [Bressac et al., 1995], has largely contributed to the hypothesis, discussed above, of the membrane localization of the enzyme. Echinostoma caproni was expected to be a good model to test this hypothesis because in this organism, the individualization process does not exist. It is therefore possible to dissociate the two events, namely, the coupling of the membrane to the axoneme and the time occurrence of the glycylated epitopes. Indeed, from a structural point of view, it exhibits an opposite situation to that found in Drosophila during spermiogenesis: in the young spermatid, comprising two free flagella (see Fig. 6), the membrane is close to the axoneme, whereas in the late spermatid, after fusion of the three processes (Fig. 6), only a few of the doublets are facing the membrane and are not tightly placed against it [Iomini and Justine, 1997]. Thus the glycylated epitope appears in the late stages even though the axoneme/membrane connection is loose, whereas it is tight in the early stages. In conclusion, the whole set of observations strengthens the previous hypothesis of a developmental regulation of glycylation occurrence during spermiogenesis [Bré et al., 1996]. Moreover, it suggests that this event might be involved in a morphogenetic clock for the end of spermiogenesis rather than being simply the consequence of a coupling of the membrane to the axoneme. In addition to a possible morphogenetic role, glycylation has been postulated to be involved in sea urchin spermatozoa motility [Bré et al., 1996]. When observed out of the seminal vesicle, E. caproni spermatozoa are able to beat but not to swim. As has been undertaken for sea urchin spermatozoa, it would be worth investigating whether TAP 952 is able to inhibit the flagellar beating. Compared to sea urchin spermatozoa, which are deco- Tubulin Polyglycylation in Stable Microtubules rated with both AXO 49 and TAP 952 [Bré et al., 1996], E. caproni spermatozoa are labelled with TAP 952 exclusively. This suggests that, in these sperm, axonemal tubulin would posses an overall lower level of glycylation, possibly arranged in short chains. It may be asked whether the structure of the glycyl chains could explain the ability for the flagella to beat, but the inability for the spermatozoa to swim. This would imply that the length of the glycyl chain could influence, directly or by the way of associated molecules, the swimming parameters of the spermatozoon. It is worth noting that the proximal appearance of glycylated tubulin isoforms in the spermatids of E. caproni is reminiscent of the proximo-distal decreasing gradient of immunoreactivity observed in sea urchin spermatozoa with AXO 49. Therefore, given the polar growth of the axoneme, the extent of glycylation could be considered, at least for some species, as a marker of the oldest part of the flagellar axoneme. One might ask whether these gradients merely reflect the age of the consecutive parts of the axoneme, or provide means to produce polarized associations of microtubules with other molecules. ACKNOWLEDGMENTS The specimens were kindly provided by Sandrine Trouvé and Annie Fournier (University of Perpignan, France). C. I. acknowledges the help of Marina Allary for editing early versions of the manuscript. Prof. B. G. M. Jamieson kindly edited the English. This work was partly supported by a BQR ‘‘Immunocytochimie des spermatozoı̈des de Plathelminthes et Nématodes’’ from the Muséum, the INTAS grant n° 93-2176, ‘‘Ultrastructure and immunocytochemistry of the cytoskeleton of spermatozoa, eggs and fertilization in selected invertebrates species, for the understanding of phylogeny,’’ the CNRS, the Université Paris-Sud, and the Muséum National d’Histoire Naturelle. We are grateful to Prof. A. Adoutte for his interest in this work. We thank the colleagues who provided us with antibodies, Roselyne Tcheprakoff for making the drawing, and L. Elu for help in photographic work. REFERENCES Adoutte, A., Delgado, P., Fleury, A., Levilliers, N., Lainé, M.-C., Marty, M.-C., Boisvieux-Ulrich, E., and Sandoz, D. (1991): Microtubule diversity in ciliated cells: evidence for its generation by post-translational modification in the axonemes of Paramecium and quail oviduct cells. Biol. Cell 71:227–245. Alexander, J.E., Hunt, D.F., Lee, M.K., Shabanowitz, J., Michel, H., Berlin, S.C., MacDonald, T.L., Sundberg, R.J., Rebhun, L.I., and Frankfurter, A. (1991): Characterization of posttranslational modifications in neuron-specific class III beta-tubulin by mass spectrometry. Proc. Natl. Acad. Sci. USA 88:4685–4689. 329 Arevalo, M.A., Nieto, J.M., Andreu, D., and Andreu, J.M. (1990): Tubulin assembly probed with antibodies to synthetic peptides. J. Mol. Biol. 214:105–120. Audebert, S., Desbruyères, E., Gruszczynski, C., Koulakoff, A., Gros, F., Denoulet, P., and Eddé, B. (1993): Reversible polyglutamylation of alpha- and beta-tubulin and microtubule dynamics in mouse brain neurons. Mol. Biol. Cell. 4:615–626. Barra, H.S., Arce, C.A., Rodriguez, J.A., and Caputto, R. (1974): Some common properties of the protein that incorporate tyrosine as a single unit and the microtubule proteins. Biochem. Biophys. Res. Commun. 60:1384–1390. Blose, S.H., Meltzer, D.I., and Feramisco, J.R. (1984): 10 nm filaments are induced to collapse in living cells microinjected with monoclonal and polyclonal antibodies against tubulin. J. Cell Biol., 98:847–858. Boucher, D., Larcher, J.C., Gros, F., and Denoulet, P. (1994): Polyglutamylation of tubulin as a progressive regulator of in vitro interactions between the microtubule-associated protein Tau and tubulin. Biochemistry 33:12471–12477. Bré, M.-H., de Néchaud, B., Wolff, A., and Fleury, A. (1994): Glutamylated tubulin probed in ciliates with the monoclonal antibody GT335. Cell Motil. Cytoskeleton 27:337–349. Bré, M.-H., Redeker, V., Quibell, M., Darmanaden-Delorme, J., Bressac, C., Cosson, J., Huitorel, P., Schmitter, J.-M., Rossier, J., Johnson, T., Adoutte, A., and Levilliers, N. (1996): Axonemal tubulin polyglycylation probed with two monoclonal antibodies: widespread evolutionary distribution, appearance during spermatozoan maturation and possible function in motility. J. Cell Sci. 109:727–738. Bressac, C., Bré, M.-H., Darmanaden-Delorme, J., Laurent, M., Levilliers, N., and Fleury, A. (1995): A massive new posttranslational modification occurs on axonemal tubulin at the final step of spermatogenesis in Drosophila. Eur. J. Cell Biol. 67:346– 355. Callen, A.-M., Adoutte, A., Andreu, J.M., Baroin-Tourancheau, A., Bré, M.-H., Ruiz, P.C., Clérot, J.-C., Delgado, P., Fleury, A., Jeanmaire-Wolf, R., Viklicky, V., Villalobo, E., and Levilliers, N. (1994): Isolation and characterization of libraries of monoclonal antibodies directed against various forms of tubulin in Paramecium. Biol. Cell 81:95–119. Caron, J.M. (1997): Posttranslational modification of tubulin by palmitoylation: I. In vivo and cell-free studies. Mol. Biol. Cell 8:621–636. Daddow, L.Y.M. (1986): An abbreviated method of the double lead stain technique. J. Submicrosc. Cytol. 18:221–224. Duvaux-Miret, O., Baratte, B., Dissous, C., and Capron, A. (1991): Molecular cloning and sequencing of the alpha-tubulin gene from Schistosoma mansoni. Mol. Biochem. Parasitol. 49:337– 340. Eddé, B., Rossier, J., Le Caer, J.-P., Desbruyères, E., Gros, F., and Denoulet, P. (1990): Posttranslational glutamylation of alphatubulin. Science 247:83–85. Eipper, B.A. (1974): Properties of rat brain tubulin. J. Biol. Chem. 249:1407–1416. Fleury, A., Callen, A.-M., Bré, M.-H., Iftode, F., Jeanmaire-Wolf, R., Levilliers, N., and Clérot, J.-C. (1995): Where and when is microtubule diversity generated in Paramecium? Immunological properties of microtubular networks in the interphase and dividing cells. Protoplasma 189:37–60. Fouquet, J.-P., Eddé, B., Kann, M.-L., Wolff, A., Desbruyéres, E., and Denoulet, P. (1994): Differential distribution of glutamylated tubulin during spermatogenesis in mammalian testis. Cell Motil. Cytoskeleton 27:49–58. Fried, B., and Huffman, J.E. (1996): The biology of the intestinal trematode Echinostoma caproni. Adv. Parasitol. 38:311–368. 330 Iomini et al. Fulton, C., and Simpson, P.A. (1976): Selective synthesis and utilisation of flagellar tubulin. The multitubulin hypothesis. In R. Goldman, T. Pollard, and J. Rosenbaum (eds): ‘‘Cell Motility.’’ Cold Spring Harbor, NY: Cold Spring Harbor Laboratory Press, pp. 987–1005. Gagnon, C., White, D., Cosson, J., Huitorel, P., Eddé, B., Desbruyéres, E., Paturle-Lafanechère, L., Multigner, L., Job, D., and Cibert, C. (1996): The polyglutamylated chain of alpha-tubulin plays a key role in flagellar motility. J. Cell Sci. 109:1545–1553. Gelfand, V.I., and Bershadsky, A.D. (1991): Microtubule dynamics: mechanim, regulation and function. Annu. Rev. Cell Biol. 7:93–116. Greer, K., and Rosenbaum, J.L. (1989): Post-translational modifications of tubulin. In F.D. Warner and J.R. McIntosh (eds): ‘‘Cell Movement, Vol. 2: Kinesin, Dynein and Microtubule Dynamics.’’ New York: Alan R. Liss, pp. 47–66. Iomini, C., and Justine, J.-L. (1997): Spermiogenesis and spermatozoon of Echinostoma caproni (Platyhelminthes, Digenea): Transmission and scanning electron microscopy, and tubulin immunocytochemistry. Tissue & Cell 29:107–118. Iomini, C., Raikova, O., Noury-Sraı̈ri, N., and Justine, J.-L. (1995): Immunocytochemistry of tubulin in spermatozoa of Platyhelminthes. In B.G.M. Jamieson, J. Ausio, and J.-L. Justine (eds): ‘‘Advances in Spermatozoal Phylogeny and Taxonomy.’’ Mém. Mus. Natn. Hist. Nat., 166:97–104. Justine, J.-L. (1991): Phylogeny of parasitic Platyhelminthes: A critical study of synapomorphies proposed on the basis of the ultrastructure of spermiogenesis and spermatozoa. Can. J. Zool. 69:1421– 1440. Khan, I.A., and Ludueña, R.F. (1996): Phosphorylation of beta III-tubulin. Biochemistry 35:3704–3711. Kyhse-Andersen, J. (1984): Electroblotting of multiple gels: a simple apparatus without buffer tank for rapid transfer of proteins from polyacrylamide to nitrocellulose. J. Biochem. Biophys. Methods 10:203–209. L’Hernault, S.W., and Rosenbaum, J.L. (1985): Chlamydomonas alpha-tubulin is posttranslationally modified by acetylation on the Sigma-amino group of a lysine. Biochemistry 24:473–478. Laemmli, U.K. (1970): Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227:680–685. LeDizet, M., and Piperno, G. (1987): Identification of an acetylation site of Chlamydomonas alpha-tubulin. Proc. Natl. Acad. Sci. USA 84:5720–5724. LeDizet, M., and Piperno, G. (1991): Detection of acetylated alphatubulin by specific antibodies. Meth. Enzymol. 196:264–274. Levilliers, N., Fleury, A., and Hill, A.-M. (1995): Monoclonal and polyclonal antibodies detect a new type of post-translational modification of axonemal tubulin. J. Cell Sci. 108:3013–3028. Maruta, H., Greer, K., and Rosenbaum, J.L. (1986): The acetylation of alpha-tubulin and its relationship to the assembly and disassembly of microtubules. J. Cell Biol. 103:571–579. Mary, J., Redeker, V., Le Caer, J.-P., Promé, J.-C., and Rossier, J. (1994): Class I and IVa beta-tubulin isotypes expressed in adult mouse brain are glutamylated. FEBS Lett. 353:89–94. Mary, J., Redeker, V., Le Caer, J.-P., Rossier, J., and Schmitter, J.-M. (1996): Posttranslational modifications in the C-terminal tail of axonemal tubulin from sea urchin sperm. J. Biol. Chem. 271:9928–9933. Multigner, L., Pignot-Paintrand, I., Saoudi, Y., Job, D., Plessmann, U., Rüdiger, M., and Weber, K. (1996): The A and B tubules of the outer doublets of sea urchin sperm axonemes are composed of different tubulin variants. Biochemistry 35:10862–10871. Ozols, J., and Caron, J.M. (1997): Posttranslational modification of tubulin by palmitoylation: II. Identification of sites of palmitoylation. Mol Biol. Cell 8:637–645. Paturle-Lafanechère, L., Eddé, B., Denoulet, P., Van Dorsselaer, A., Mazarguil, H., Le Caer, J.P., Wehland, J., and Job, D. (1991): Characterization of a major brain tubulin variant wich cannot be tyrosinated. Biochemistry 30:10523–10528. Piperno, G., and Fuller, M.T. (1985): Monoclonal antibodies specific for an acetylated form of alpha-tubulin recognize the antigen in cilia and flagella from a variety of organisms. J. Cell Biol. 101:2085–2094. Raff, E.C. (1994): The role of multiple tubulin isoforms in cellular microtubule function. In J.S. Hyams and C.W. Lloyd (eds): ‘‘Microtubules.’’ New York: Wiley-Liss, pp. 85–109. Redeker, V., Levilliers, N., Schmitter, J.-M., Le Caer, J.-P., Rossier, J., Adoutte, A. and Bré, M.-H. (1994): Polyglycylation of tubulin: A posttranslational modification in axonemal microtubules. Science 266:1688–1691. Redeker, V., Melki, R., Promé, D., Le Caer, J.-P., and Rossier, J. (1992): Structure of tubulin C-terminal domain obtained by subtilisin treatment. The major alpha and beta tubulin isotypes from pig brain are glutamylated. FEBS Lett. 313:185–192. Richard, J. (1964): Trématodes d’oiseaux de Madagascar (Note III) Espèces de la famille Echinostomatidae Poche 1926. Ann. Parasitol. Hum. Comp. 34:607–620. Rohde, K. (1990): Phylogeny of Platyhelminthes, with special reference to parasitic groups. Int. J. Parasitol. 20:979–1007. Rüdiger, M., Plessman, U., Kloppel, K.D., Wehland, J., and Weber, K. (1992): Class II tubulin, the major brain beta tubulin isotype is polyglutamylated on glutamic acid residue 435. FEBS Lett. 308:101–105. Rüdiger, M., Plessmann, U., Rüdiger, A.H., and Weber, K. (1995): Beta tubulin of bull sperm is polyglycylated. FEBS Lett. 364:147– 151. Schneider, A., Plessmann, U., and Weber, K. (1997): Subpellicular and flagellar microtubules of Trypanosoma brucei are extensively glutamylated. J. Cell Sci. 110:431–437. Thompson, W.C. (1982): The cyclic tyrosination/detyrosination of alpha tubulin. Meth. Cell Biol. 24:235–255. Weber, K., Schneider, A., Müller, N., and Plessmann, U. (1996): Polyglycylation of tubulin in the diplomonad Giardia lamblia, one of the oldest eukaryotes. FEBS Lett. 393:27–30. Wolff, A., de Néchaud, B., Chillet, D., Mazarguil, H., Desbruyères, E., Audebert, S., Eddé, B., Gros, F., and Denoulet, P. (1992): Distribution of glutamylated alpha and beta tubulin in mouse tissues using a specific monoclonal antibody, GT335. Eur. J. Cell Biol. 59:425–432.