Vinculin talin integrin О61 and laminin can serve as components of attachment complex mediating contraction force transmission from cardiomyocytes to extracellular matrixкод для вставкиСкачать
Cell Motility and the Cytoskeleton 41:341–352 (1998) Expression of Glycylated Tubulin During the Differentiation of Spermatozoa in Mammals Marie-Louise Kann,1* Yann Prigent,1 Nicolette Levilliers,2 Marie-Hélène Bré,2 and Jean-Pierre Fouquet1 1Laboratoire de Biologie Cellulaire, Spermatogenèse et maturation du spermatozoı̈de, Université Paris V, U.F.R. Biomédicale, Paris, France 2Laboratoire de Biologie Cellulaire 4, U.R.A. 1134 C.N.R.S., Université Paris-Sud, Orsay, France Using quantitative immunogold analyses of tubulin isoforms we previously demonstrated a unique differential expression of glutamylated tubulin in the flagellum of mouse and man spermatozoa [Fouquet et al., 1997: Tissue Cell 29:573–583]. We have performed similar analyses for glycylated tubulin using two monoclonal antibodies, TAP 952 and AXO 49, directed to mono- and polyglycylated tubulin respectively. Glycylated tubulin was not found in centrioles and cytoplasmic microtubules (manchette) of germ cells. In mouse and man, axonemal tubulin was first monoglycylated and uniformly distributed in all doublets at all levels of the flagellum in elongating spermatids. In human mature spermatozoa axonemal microtubules were enriched in monoglycylated tubulin from the base to the tip of the flagellum. In mouse sperm flagellum a similar gradient of monoglycylated tubulin was also observed in addition to an opposite gradient of polyglycylated tubulin. In both species, monoglycylated tubulin labeling predominated in doublets 3-8 whereas glutamylated tubulin labeling [Fouquet et al., 1997] predominated in doublets 1-5-6. These differential labelings were suppressed after motility inhibition of mouse spermatozoa by sodium azide treatment and in non-motile human spermatozoa lacking dynein arms. The unique distribution of these tubulin isoforms and the known inhibition of motility induced by their specific antibodies are consistent with a complementary role of tubulin glycylation and glutamylation in the regulation of flagellar beating in mammalian spermatozoa. Cell Motil. Cytoskeleton 41:341–352, 1998. r 1998 Wiley-Liss, Inc. Key words: axoneme; microtubules; glycylated tubulin; spermatozoa; immunogold INTRODUCTION In the mammalian testis, germ cell microtubules are made of three major tubulin isotypes: m␤ 3 and m␣ 3/7 [Lewis and Cowan, 1988] and pRD␣TT1 [Hecht et al., 1988; Simerly et al., 1993]. These isotypes are either expressed as unmodified polypeptides or as posttranslationally modified isoforms. For ␣ tubulin the identified modifications are the N-terminal acetylation at lys 40, the detyrosylation of C-terminal and excision of the penultimate glutamyl residue after detyrosylation [for review see MacRae, 1997; Ludueña, 1998]. In addition, both ␣ and ␤ tubulin can be subjected to polyglutamylation and polyglycylation. The glutamylation consists of r 1998 Wiley-Liss, Inc. the C-terminal addition, of one to six glutamate units at glu 445 for ␣ tubulin [Eddé et al., 1990] and at glu 435 for ␤ tubulin [Redeker et al., 1992; Rüdiger et al., 1992]. The glycylation consists of the C-terminal addition of three to 34 glycine units as originally reported in Paramecium cilia [Redeker et al., 1994] or one to 23 glycine units in the bull sperm flagellum [Plessmann and Weber, 1997] at glu 445 for ␣ tubulin and at glu 437 for ␤ tubulin. During *Correspondence to: Dr M.L. Kann, Laboratoire de Biologie Cellulaire, Université Paris V, U.F.R. Biomédicale, 45 rue des Saints-Pères, 75270 Paris cedex 06, France. Received 9 July 1998; accepted 8 September 1998 342 Kann et al. spermiogenesis the spermatid build up two microtubular structures: the manchette, a transient bundle of microtubules encasing the posterior region of the nucleus, and axonemal microtubules of the flagellum. In the flagellum all post-translational modifications described above have been characterized whereas in the manchette the only reported modification is the detyrosylation of ␣ tubulin [Fouquet et al., 1994, 1997]. Recently it has been demonstrated that among a panel of site directed antibodies to tubulin only those directed to glutamylated and glycylated tubulin were potent inhibitors of flagellar motility [Bré et al., 1996; Gagnon et al., 1996], and only those directed to glutamylated tubulin evidenced a differential labeling of the mammalian sperm flagellum using quantitative immunoelectron microscopy [Fouquet et al., 1996, 1997; Prigent et al., 1996]. Both a proximo-distal decrease of the labeling and its predominance in axonemal doublets 1-5-6 corresponding to the plane of the flagellar wave were reported. Together, these results have suggested that glutamylated tubulin could be involved in a functional heterogeneity of peripheral doublets of the sperm flagellum, i.e., flagellar beating regulation. Since polyglycylated tubulin is also expressed in the mammalian sperm flagellum and since glycylation and glutamylation occur on the same and/or adjacent C-terminal glutamate residues of ␣ and ␤ tubulin [Bré et al., 1996; Plessmann and Weber, 1997], it appears that the glycylation of flagellar microtubules also might be involved in their functional heterogeneity. In the present work the distribution of glycylated tubulin was studied in the sperm flagellum of various mammals by immunoelectron microscopy using two monoclonal antibodies, AXO 49 and TAP 952, directed to glycylated tubulin [Bré et al., 1996]. MATERIALS AND METHODS Biological Samples Testes and epididymides of sexually active mouse (Swiss) were removed under anesthesia. Mouse motile epididymal spermatozoa were collected in phosphate buffered saline (PBS), pH 7.4 at 37°C within 15 min [Fouquet et al., 1994]. In some experiments motility was inhibited with sodium azide 0.1% before further investigation. Testicular biopsies and ejaculates from human donors were also used. Ejaculated spermatozoa were obtained after an abstinence period of 2–4 days from 10 fertile donors assessed as normal from their spermogram and spermocytogram; sperm counts were ⬎20 ⫻ 106/ml, ⬎50% of spermatozoa were motile and ⬍50% were morphologically abnormal. In addition, non-motile ejaculated spermatozoa, lacking dynein arms, from two nonfertile donors were also studied. Testicular samples of other species including hamster, rat, rabbit and monkey (Maccaca fascicularis) as previously prepared [Fouquet et al., 1994] were also investigated. Antibodies Two monoclonal antibodies, TAP 952 and AXO 49, raised against axonemal tubulin of Paramecium [Callen et al., 1994] and directed to glycylated tubulin [Bré et al., 1996] were used. TAP 952 mAb recognizes monoglycylated tubulin bearing single glycine unit on one or several carboxyterminal glutamate residues. AXO 49 mAb recognizes polyglycine chains from three residues upwards [Bré et al., 1998]. Neither TAP 952 mAb nor AXO 49 mAb recognizes biglycylated peptides. Indirect Immunofluorescence (IIF) As in a previous study [Kann et al., 1995] smears of mouse epididymal spermatozoa and mechanically isolated testicular germ cells were fixed-permeabilized in acetone (⫺20°C) before immunostaining. In some cases, germ cell samples diluted in phosphate buffer saline (PBS) were frozen (1 h, ⫺20°C) before the preparation of smears to enhance antigen accessibility. Immunostainings were performed for 45 min at room temperature with TAP 952 or AXO 49 mAb, then with FITC conjugated anti-mouse IgG (ICN) diluted in PBS–0.2% bovine serum albumin (BSA). The preparations were mounted in PBS-glycerol (10/90, v/v) before epifluorescence observations. Immunoelectron Microscopy (IEM) All biological samples were fixed for 1 h in 1% glutaraldehyde–0.1 M cacodylate buffer pH 7.3 and embedded in Lowicryl K4M. Thin sections collected on uncoated nickel grids were incubated for 2 h in different antibodies followed by incubation in goat antimouse secondary antibodies conjugated either to 10 nm (Sigma, Aldritchchimie, France) or 15 nm (Biocell, U.K.) gold particles. All antibodies were diluted in Tris buffered saline (TBS) pH 7.9 containing 0.2% bovine serum albumin. The sections were contrasted with aqueous saturated uranyl acetate before electron microscope observations. Quantitative Immunogold Analysis The number of gold particles was determined per cross-section of mouse and human sperm flagellum. Five regions were considered: part I, middle piece; part II, proximal part of the principal piece with 9-6 outer dense fibres (ODF); part III, principal piece with 4-2 ODF; part IV, principal piece devoid of ODF; part V, terminal piece. The results are presented as the mean ⫾ SEM for peripheral doublets, central pair and total microtubules (central pair plus peripheral doublets) of the axoneme in Sperm Glycylated Tubulin 343 Fig. 1. IIF of mouse epididymal spermatozoa, after freezing pretreatment, with TAP 952 (a) and AXO 49 (b) mAb, and corresponding phase contrast (a8, b8). Bar, 10 µm. ⫻640. 50 cross-sections per region. Gold particle counts were also performed on individual peripheral doublets to search for a possible differential labeling. As in a previous report [Fouquet et al., 1997], three groups of doublets were compared: 1-5-6, 3-8 and 2-4-7-9. One-way analysis of variance was used for the statistical determination of significant differences in the labeling patterns of various parts of the flagellum; the Duncan multiple range test was used for specific comparisons. The Wilcoxon test was used for doublet comparisons in each region of the flagellum. RESULTS IIF of Glycylated Tubulin During Mammalian Spermiogenesis IIF was used to detect glycylated tubulin in spermatozoa and different germ cell types isolated from mouse testes. The TAP 952 mAb did not label spermatogonia, spermatocytes or round spermatids. A uniform labeling was first detected in the flagellum of early elongating spermatids. During the maturation phase of spermatids, the labeling showed a greater intensity at the tip than at the base of the flagellum, i.e., a proximo-distal increase of labeling. In late spermatids with a well differentiated middle piece, i.e., testicular spermatozoa, and in epididymal spermatozoa a similar labeling gradient was observed, but it was less intense than in previous steps of differentiation. As in a previous work [Kann et al., 1995] the less intense labeling observed in terminally differentiated cells than in younger cells was attributed to the presence of compact periaxonemal sheaths which constitute physical barriers to the diffusion of antibodies despite permeabilization treatments. Indeed, freezing of spermatozoa before the preparation of smears resulted in a fluorescence intensity similar to that observed in maturing spermatids (Fig. 1a, a8). The AXO 49 mAb did not label any germ cell types except the flagellum of maturing spermatids, testicular and epididymal spermatozoa. This labeling was also enhanced after freezing pretreatment. A decreasing labeling was observed from the middle piece to the end of the principal piece of the flagellum, i.e., a proximo-distal decrease, with a predominant labeling of the flagellum tip (Fig. 1b, b8). In man, only ejaculated spermatozoa could be studied. The labeling of the flagellum with the TAP 952 mAb showed a proximo-distal increase as described 344 Kann et al. above for mouse spermatozoa. No labeling could be detected with the AXO 49 mAb (data not shown). IEM of Glycylated Tubulin During Mammalian Spermiogenesis In all studied species, mouse (Fig. 2a, b, c), hamster, rat, rabbit, monkey and man, the TAP 952 mAb labeled the whole axoneme with an increasing intensity from the middle piece to the terminal piece (region I-V) of the flagellum of epididymal and ejaculated spermatozoa. During spermiogenesis the axoneme of round spermatids was not yet labeled. The labeling appeared in the whole axoneme of elongating spermatids with a uniform distribution and an increasing intensity. Gold particles were first detected from the beginning of the elongation phase in mouse spermatids but at later steps of spermiogenesis (mid-elongation phase) in other species. The manchette was never labeled (Fig. 2d). The proximo-distal differential labeling of the flagellum described above for mature spermatozoa became detectable from the mid-maturation phase of spermatids in all species. In spermatozoa of rodents, as shown for the mouse (Fig. 3) and the rabbit, the AXO 49 mAb produced a decreasing labeling from the middle piece (region I) to the end of the principal piece (region IV) of the flagellum with a predominant labeling, however, in the terminal piece (region V), including the junctional area of region IV with V. In contrast, in primate spermatozoa, monkey and man, no labeling could be detected at any level of the flagellum. During spermiogenesis in rodents and rabbit, the axoneme of round and elongating spermatids and the manchette were not labeled. The labeling became detectable during the maturation phase of spermatids and immediately exhibited the differential distribution described above for mature spermatozoa. In summary, the results observed by IEM fit well with those observed by IIF. In addition, it should be noted that centrioles were never labeled either with TAP 952 (Fig. 2d) or with AXO 49 mAb (data not shown). Quantitative Immunogold Analyses of Glycylated Tubulin in Mammalian Spermatozoa The distribution of gold particles was analysed at the five levels of sperm flagellum in mouse and man to determine more accurately the differential labeling patterns described above. Examples of pictures used in these analyses are shown in Figure 2b, c and Figure 3 for mouse spermatozoa and Figure 4 for human spermatozoa. The results obtained with motile human spermatozoa (dynein-positive) labeled with TAP 952 mAb are presented in Figure 5a. This graph shows that the end of the principal piece and the terminal piece of the flagellum (regions IV and V) are respectively twofold and 2.5-fold more labeled than the middle piece (region I). From the beginning of the principal piece (region II) the labeling intensity was sufficient to perform a quantitative analysis per doublet. This analysis (Fig. 5a8) shows in region II a predominant labeling in doublets 3-8, whereas the labeling is twice less in doublets 1-5-6 and three to four times less in doublets 2-4-7-9. In more distal regions of the principal piece (regions III and IV) the labeling intensity of doublets 3-8 is similar whereas that of other doublets increases progressively but remains significantly less (P ⬍ 0.001). A similar study was also performed on non-motile human spermatozoa devoid of dynein arms (dyneinnegative). At first sight these spermatozoa also exhibited a proximo-distal increase of the labeling along the flagellum. However, the predominant labeling of doublets 3-8 was not as evident as in motile (dynein-positive) spermatozoa (compare panels a and b of Fig. 4). Gold particle counts fully confirmed the simple observation. In fact, the same proximo-distal labeling gradient was demonstrated in both dynein-positive and dyneinnegative spermatozoa (compare panels a and b of Fig. 5). However, the differential labeling of peripheral doublets was dramatically reduced (compare panels a8 and b8 of Fig. 5). More precisely at each level the doublets 3-8 and 2-4-7-9 were now uniformly labeled whereas the labeling intensity was significantly less (P ⬍ 0.01) in doublets 1-5-6 (regions II and IV). The expected differential labeling of mouse spermatozoa with TAP 952 mAb was also confirmed but the profile was slightly different of that presented above for human spermatozoa. Thus, the labeling increased slowly from the region I (middle piece) to the mid principal piece (region III) then increased sharply at the end of the principal piece (region IV) and in the terminal piece (region V) (Fig. 6a). In the mouse as in man, the differential labeling of doublets was characterized by a predominant labeling of doublets 3-8 in regions II, III, whereas this labeling was almost uniform in region IV (Fig. 6a8). An analysis of TAP 952 mAb labeling was also performed with sodium azide immobilized mouse spermatozoa. This treatment did not modify the profile of the proximo-distal labeling gradient reported above (compare panels a and b of Fig. 6). However, in this case, the predominant labeling of doublets 3-8 was observed only in the region II (compare panels a8 and b8 of Fig. 6). On the other hand the labeling of doublets 2-4-7-9 began more intensely than did that of doublets 1-5-6, thus corresponding to an inversion of the labeling intensity as observed in motile spermatozoa. The quantitative analysis with AXO 49 mAb was performed only on mouse spermatozoa since as indicated Sperm Glycylated Tubulin Fig. 2. Monoglycylated tubulin immunogold labeling with the mAb TAP 952. a, b, c: Mouse epididymal spermatozoa labeled with 15 nm gold particles. II, III, IV and V refer to successive regions of flagellum. d: Elongated spermatid labeled with 10 nm gold particles. The axoneme is labeled but the centriole (arrow) and manchette (m) are not. Bar, 0.25 µm. a, d, ⫻40,000; b, c, ⫻80,000. 345 346 Kann et al. Fig. 3. Polyglycylated tubulin immunogold labeling of mouse epididymal spermatozoa with the mAb AXO 49 and 15 nm gold particles (a, b). I, II, III, IV and V as in Figure 2. Note the highest labeling at the end of region IV and in region V (a). Bar, 0.25 µm. ⫻60,000. above this mAb did not react with human spermatozoa. The proximo-distal decrease of the labeling previously described was confirmed. Thus (Fig. 7a), the middle piece (region I) was seven times more labeled that the end of the principal piece (region IV). However, the most intense labeling was observed in the short terminal piece (region V). Moreover, at each level, the labeling of each doublet was similar (Fig. 7a8). After sodium azide motility inhibition of mouse spermatozoa, the general profile of AXO 49 labeling was not modified (compare panels a and b of Fig. 7). Likewise, the uniform labeling of doublets remained essentially unchanged (compare panels a8 and b8 of Fig. 7). DISCUSSION The post-translational modifications (PTM) of tubulin isoforms and particularly polyglutamylation and polyglycylation are universally expressed from ciliates to mammals [for review see MacRae, 1997; Ludueña, 1998] specifically in axonemes, i.e., in the most stable microtubules. Mammalian spermatids contain three microtubular structures: the flagellar axoneme, itself arising from a centriole, and the manchette, a transient bundle of cytoplasmic microtubules involved in nuclear shaping [for review see Fouquet and Kann, 1994]. As expected flagella were labeled with antiglutamylated tubulin [Fou- Sperm Glycylated Tubulin 347 Fig. 4. Monoglycylated tubulin immunogold labeling with the mAb TAP 952 and 15 nm gold particles. Cross sections in the region II of flagellum in motile (dynein-positive) (a) and non-motile (dyneinnegative) (b) human spermatozoa. Bar, 0.25 µm. ⫻80,000. quet et al., 1994, 1997] and antiglycylated tubulin antibodies (present results) whereas the manchette was not labeled with the same antibodies. In spermatogenic cells of all species studied here centrioles appeared glutamylated as previously reported [Fouquet et al., 1994, 1997] and as also described for somatic cells [Bobinnec et al., 1998] but they were not glycylated. Taken together these differential expressions of tubulin isoforms in the centriole and axoneme suggest that centriole microtubules might be less stable than axoneme microtubules. Indeed, centrioles and their associated pericentriolar material are dynamic structures during the cell cycle [Moudjou et al., 1996; Bobinnec et al., 1998]. In the axoneme of mammalian sperm flagellum all ␣-tubulin isoforms and a majority of ␤-tubulin isoforms are glutamylated [Fouquet et al., 1994]. Their labeling assumes a decreasing proximo-distal gradient and a prevalence in doublets 1-5-6 in rodents and man [Fouquet et al., 1996; Prigent et al., 1996; Fouquet et al., 1997]. In this work we also confirm the presence of glycylated tubulin in mammalian spermatozoa [Bré et al., 1996; Plessmann and Weber, 1997]. In addition our results clearly demonstrate a differential expression of glycylation in particular peripheral doublets of the flagellum. Results of IIF and IEM show that polyglycylation is a late PTM of tubulin isoforms as compared to acetylation, detyrosylation and glutamylation during mammalian spermatogenesis [Fouquet et al., 1997] and as already observed in Drosophila [Bré et al., 1996]. Indeed, a majority of PTM are observed from the round spermatid stage in naked still growing axonemes whereas glycylation is only detectable in elongating spermatids with axonemes of definitive length but which are building their periaxonemal sheaths. In elongated spermatids, in mouse as in man, the glycylated tubulin is, at first, detected with the TAP 952 mAb. This allows the conclusion that axonemal tubulin is, at first, monoglycylated and uniformly distributed in all doublets at all levels of axoneme. During the following steps of spermatid differentiation, i.e., maturing spermatids, the levels of glycylation are different in the mouse and man. In the mouse during the last steps of spermatid differentiation into testicular spermatozoa, as well as in epididymal spermatozoa, two opposite labeling gradients were observed: a proximo-distal decreasing labeling with AXO 49 mAb and a proximo-distal increasing labeling with TAP 952 mAb. These gradients and antibody specificity suggest that in proximal regions of the flagellum (I and II) microtubules carry a majority of polyglycylated chains with three or more glycine units whereas in more distal region of the flagellum (III and IV) the microtubules are essentially monoglycylated. In fact, the labeling, both with TAP 952 and AXO 49 mAbs, indicates the coexistence of both monoglycylated and polyglycylated chains in any region of the flagellum. In addition to these linear labeling gradients, antibodies against glycylated tubulin revealed a differential labeling among peripheral doublets. Thus, in regions II, III and probably I of the flagellum, doublets 3-8 are richer in monoglycylated tubulin than doublets 1-5-6 and 2-4-7-9 respectively. In contrast, polyglycylated tubulin seems uniformly distributed in all microtubule doublets. On the whole, these results suggest that in mouse sperm flagellum, axonemal microtubules have numerous glycylatable sites bearing glycine chains of different length. In man, the flagellum of spermatids and spermatozoa was never labeled with AXO 49 mAb. At first sight this result indicates that axonemal tubulin is monoglycylated only as recognized by TAP 952 mAb, and predominantly in doublets 3-8. However, in human spermatozoa, as in mouse, a proximo-distal increase of labeling was observed in the whole flagellum. To explain this labeling 348 Kann et al. Fig. 5. Quantitative analyses of monoglycylated immunogold labeling along dynein-positive (a) and dynein-negative (b) human sperm flagellum with the mAb TAP 952. The mean number ⫾ SEM of gold particles is indicated for the central pair, peripheral doublets and whole microtubules in cross sections of the middle piece (I), principal piece (II, III, IV) and terminal piece (V). Right panels (a8, b8): the mean number ⫾ SEM of gold particles per doublet in groups 2-4-7-9, 1-5-6 and 3-8 is indicated for the regions II–IV of the flagellum. Sperm Glycylated Tubulin Fig. 6. Quantitative analyses of monoglycylated immunogold labeling along the flagellum of motile (a) and non-motile (azide treated) (b) mouse epididymal spermatozoa. Legends as in Figure 5a, b. Right panels (a8, b8): legends as in Figure 5a8, b8. 349 350 Kann et al. Fig. 7. Quantitative analyses of polyglycylated immunogold labeling along the flagellum of motile (a) and non-motile (azide treated) (b) mouse epididymal spermatozoa. Legends as in Figure 5a, b. Right panels (a8, b8): legends as in Figure 5a8, b8. Sperm Glycylated Tubulin gradient which appeared at the end of spermiogenesis two interpretations can be proposed. The simplest one is that the number of monoglycylated tubulin units increases progressively from proximal to distal regions of the flagellum. Another possibility is that a polyglycylation limited to two glycyl units occurs in proximal regions of the flagellum at least in some glycylatable sites. As a consequence a decrease of TAP 952 mAb labeling can be expected in corresponding regions, thus creating a labeling gradient with this mAb, but no reactivity with AXO 49 mAb, which do not recognize biglycylated chains [Bré et al., 1998]. To conclude, in mouse as in man the monoglycylated tubulin predominates in distal regions of the flagellum and in doublets 3-8 despite differences in glycylation levels. Since we have previously reported [Fouquet et al., 1997] that glutamylated tubulin expression is modified in abnormal or motility inhibited flagella we also investigated these models for glycylated tubulin expression. In mouse, sodium azide inhibition of sperm motility did not change the labeling of polyglycylated tubulin as revealed with AXO 49 mAb. In contrast, with TAP 952 mAb, the changes in monoglycylated tubulin labeling which occurred in various doublets tended to make uniform their labeling intensity. Similar changes were also observed in non-motile human spermatozoa lacking dynein arms. Thus in both species the major change was the loss of the predominant labeling in doublets 3-8. These two models suggest a functional role for monoglycylated tubulin, but not for polyglycylated tubulin, in doublets other than 1-5-6 and particularly in doublets 3-8 of the axoneme in mammalian sperm flagellum. Indeed, the decrease of TAP 952 mAb labeling may be interpreted as an accessibility decrease of this antibody for monoglycylated tubulin in doublets 3-8, i.e., an increase of interactions of these doublets with adjacent structures. These changes are just opposite to those previously observed for glutamylated tubulin labeling with the GT335 mAb [Fouquet et al., 1997]. Thus, in motile spermatozoa, glutamylated tubulin labeling is much more intense in doublets 1-5-6 than in other doublets whereas in non-motile spermatozoa the labeling is uniform in all doublets because it increases in doublets other than 1-5-6, particularly in 3-8 doublets. Therefore in the same sperm models the loss of flagellar motility seems to be correlated with an unmasking of glutamylated tubulin and a masking of glycylated tubulin particularly in doublets 3-8 of the axoneme, i.e., a decrease and an increase of interactions respectively with adjacent molecules. This interpretation in terms of differential interactions of glycylated and glutamylated tubulin with adjacent structures is supported by the fact that with both tubulin isoforms [present results and Fouquet et al., 1997] the highest labelings were observed in the terminal 351 piece of the flagellum which is devoid of periaxonemal structures and dynein arms. Taking also into account that among a panel of site-directed antibodies to C-terminal tubulin epitopes only those directed to glutamylated epitopes [Gagnon et al., 1996] and glycylated epitopes [Bré et al., 1996] were able to inhibit sperm motility in man and sea urchin it is tempting to speculate that glutamylation and glycylation are two tubulin PTM which might play a complementary role in the regulation of flagellar beating. This hypothesis is consistent with the plane of morphological symmetry and flagellar beating of mammalian spermatozoa as defined by doublets 1-5-6 and the plane of fibrous sheath symmetry as defined by doublets 3-8. Obviously, there is a need to determine axonemal and/or periaxonemal molecules which might specifically interact with glutamylated and/or glycylated tubulin in the flagellum of mammalian spermatozoa. This last point is important since the periaxonemal structures are believed to be involved in the flagellar beating regulation [Eddy and O’Brien, 1994]. To conclude, both glycylation and glutamylation are the only PTM of tubulin which are differentially expressed in the flagellum of mammalian spermatozoa and for which a functional significance can be proposed. ACKNOWLEDGMENTS The technical assistance of A. Gonzales for tissue embedding and sectioning and the photographic work of E. Prieto are appreciated. We are indebted to Mrs. D. Bligny for excellent secretariat assistance. REFERENCES Bobinnec, Y., Moudjou, M., Fouquet, J.P., Desbruyères, E., Eddé, B., and Bornens, M. (1998): Glutamylation of centriole and cytoplasmic tubulin in proliferating non-neuronal cells. Cell Motil. Cytoskeleton 39:223–232. 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. Bré, M.H., Redeker, V., Vinh, J., Rossier, J., and Levilliers, N. (1998): Tubulin polyglycylation: differential posttranslational modification of dynamic cytoplasmic and stable axonemal microtubules in Paramecium. Mol. Biol. Cell., 9:2655–2665. Callen, A.M., Adoutte, A., Andreu, J.M., Baroin-Tourancheau, A., Bré, M.H., Ruiz, P.C., Clérot, J.C., Delgado, P., Fleury, A., JeanmaireWolf, 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. Eddé, B., Rossier, J., Le Caer, J.P., Desbruyères, E., Gros, F., and Denoulet, P. (1990): Posttranslational glutamylation of alphatubulin. Science 247:83–85. 352 Kann et al. Eddy, E.M., and O’Brien, D.A. (1994): The spermatozoon. In Knobil, E., and Neil, J.D. (eds): ‘‘The Physiology of Reproduction,’’ vol. 1. New York: Raven, pp. 29–77. Fouquet, J.P., and Kann, M.L. (1994): The cytoskeleton of mammalian spermatozoa. Biol. Cell 81:89–93. Fouquet, J.P., Eddé, B., Kann, M.L., Wolff, A., Desbruyères, E., and Denoulet, P. (1994): Differential distribution of glutamylated tubulin during spermatogenesis in mammalian testis. Cell Motil. Cytoskeleton 27:49–58. Fouquet, J.P., Prigent, Y., and Kann, M.L. (1996): Comparative immunogold analysis of tubulin isoforms in the mouse sperm flagellum: unique distribution of glutamylated tubulin. Mol. Reprod. Dev. 43:358–365. Fouquet, J.P., Kann, M.L., Péchart, I., and Prigent, Y. (1997): Expression of tubulin isoforms during the differentiation of mammalian spermatozoa. Tissue Cell 29:573–583. Gagnon, C., White, D., Cosson, J., Huitorel, P., Eddé, B., Desbruyères, E., Paturle-Lafanechère, L., Multigner, L., Job, D., and Cibert, C. (1996): The polyglutamylated lateral chain of alpha-tubulin plays a key role in flagellar motility. J. Cell Sci. 109:1545–1553. Hecht, N.B., Distel, R.J., Yelick, P.C., Tanhauser, S.M., Driscoll, C.E., Goldberg, E., and Tung, K.S.H. (1988): Localization of a highly divergent mammalian testicular ␣-tubulin that is not detectable in brain. Mol. Cell Biol. 8:996–1000. Kann, M.L., Prigent, Y., and Fouquet, J.P. (1995): Differential distribution of glutamylated tubulin in the flagellum of mouse spermatozoa. Tissue Cell 27:323–329. Lewis, S.A., and Cowan, N.J. (1988): Complex regulation and functional versatility of mammalian ␣ and ␤-tubulin isotypes during the differentiation of testis and muscle cells. J. Cell Biol. 106:2023–2033. Ludueña, R.F. (1998): Multiple forms of tubulin: different gene products and covalent modifications. Int. Rev. Cytol. 178:207– 275. MacRae, T.H. (1997): Tubulin post-translational modifications, enzymes and their mechanisms of action. Eur. J. Biochem. 244:265–278. Moudjou, M., Bordes, N., Paintrand, M., and Bornens, M. (1996): ␥-tubulin in mammalian cells: the centrosomal and the cytosolic forms. J. Cell Sci. 109:875–887. Plessmann, U., and Weber, K. (1997): Mammalian sperm tubulin: an exceptionally large number of variants based on several posttranslational modifications. J. Protein Chem. 16:385–390. Prigent, Y., Kann, M.L., Lach-Gar, H., Péchart, I., and Fouquet, J.P. (1996): Glutamylated tubulin as a marker of microtubule heterogeneity in the human sperm flagellum. Mol. Hum. Reprod. 2:573–581. Redeker, V., Melki, R., Promé, D., Le Caer, J.P., and Rossier, J.P. (1992): Structure of tubulin c-terminal domain obtained by subtilisin treatment. The major ␣ and ␤-tubulin isotypes from pig brain are glutamylated. FEBS Lett. 313:185–192. 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 of axonemal microtubules. Science 226:1688–1691. Rüdiger, M., Plessmann, U., Klöppel, K.D., Wehland, J., and Weber, K. (1992): Class II tubulin, the major brain ␤-tubulin isotype, is polyglutamylated on glutamic acid residue 435. FEBS Lett. 308:101–105. Simerly, C.R., Hecht, N.B., Goldberg, E., and Schatten, G. (1993): Tracing the incorporation of the sperm tail in the mouse zygote and early embryo using an anti-testicular ␣-tubulin antibody. Dev. Biol. 158:536–548.