Ultrastructural analysis of sperm flagella in two clitellates (Annelida), plasma membrane and periaxonemal area

TISSUE AND CELL, 1991 23 (4) 537-546 0 1991 Longman Group UK Ltd.

M. FERRAGUTI”,

M. BONDI*

and S. LANZAVECCHlAt

ULTRASTRUCTURAL ANALYSIS OF SPERM FLAGELLA IN TWO CLITELLATES (ANNELIDA), PLASMA MEMBRANE AND PERIAXONEMAL AREA Keywords:

Spermatozoa,

Oligochaeta,

zipper lines, ultrastructure,

freeze-fracture,

flagella

ABSTRACT. The region between axonemes and plasma membrane in the sperm tails of the tubificid oligochaetes Tubifex tubifex and Monopylephorus limosus has been studied by means of thin sections of conventionally and tannic acid lixed material. and of freeze-fracture replicas. The main portion of the flagellum in both species showed prominent, regularly repeating bridges connecting doublets to plasma membrane. In correspondence to the doublets, characteristic double rows of intramembrane particles are present, with an arrangement reminiscent of the ‘zipper lines’ described in other species. A well-developed cortical web with a honeycomb appearance underlies the plasma membrane. Glycogen granules are regularly arranged within the cells. An outstanding difference between the two species is to be found in the presence only in Monopylephorus of complex muff-like structures apparently formed by membrane particles and series of teeth embedded in the cortical web. Nothing similar has been found in Tubifex. This difference may be related to the fact that spermatozoa in Monopylephorus are not enclosed in spermatozeugmata as are those of Tubifex.

Introduction In the oligochaete family Tubificidae the spermatozoa are stored after copulation in the spermathecae and form, in some species, peculiar aggregates (spermatozeugmata) ‘characterized by repetitive order of the spermatozoa and the presence of some cementing agent’ (Ferraguti et al., 1989). Among the various species so far studied, those belonging to the subfamily Tubificinae show a constant rod-like grouping of spermatozeugma in which two types of spermatozoa are present: the atypical ones, oligopyrenes and connected by a complex series of cell junctions constitute the external layer (cortex), *Dipartimento di Biologia, Universita di Milano. Italy. tlstituto di Chimica Strutturistica Inorganica. Universita di Milano, Italy. Correspondence to: Marco Ferraguti, Dipartimento di Biologia, Universita di Milano, 26. Via Celoria, I-20133 Milano, Italia. Received 14 January 1991 Revised 11 April 1991

whereas the fertilizing sperm is stored in the innermost portion of the structure, called axial cylinder (Ferraguti et al., 1988). The fertilizing spermatozoa are not connected by any obvious structure, but just held packed and parallel in the space delimited by the cortex. In the other subfamilies of Tubificidae, no double sperm line and no evidence of cell junctions connecting the sperm tails has been found up to now, even though highly ordered sperm bundles are observed in some species. The spermatozeugmata of Tubificinae appear thus as extremely specialized structures whose phylogenetic origin is, at present, not easy to explain. The genus Monopylephorus (subfamily Rhyacodrilinae) may supply some interesting observations in this respect, since I) it is considered quite primitive within Tubificidae (ErsCus, 1990) and II) spermatozoa are present in the different species of the genus in various stages of aggregation, from loose sperm (Baker and Brinkhurst, 1981) to compact bundles (ErsCus and Paoletti, 1986). 537

FERRAGUTI

538

In the present paper we compare the membrane specializations of the flagella in the sperm bundles from the spermatheca of the euryhaline species Monopylephorus limosus to those of the eupyrene fertilizing spermatozoa from the axial cylinder of Tubifex spermatozeugmata. Materials and Methods Monopylephorus limosus is a tubificid annelid originally reported from highly polluted brackish waters of Japan and China and recently found in the Lambro river, a highly polluted watercourse close to Milan0 (Erseus and Paoletti, 1986). In the muddy sediments of the Lambro, Monopylephorus forms extremely dense populations with other species of tubificids, amongst which are Tubifex tubifex.

Worms belonging to the two species were selected and bred in the laboratory to supply sexually mature animals throughout the year. Various fixation techniques for transmission electron microscopy have been employed: I) (TAGO) Tubifex tubifex spermatozeugmata were dissected from the spermathecae and immediately fixed in a 0.1 M cacodylate buffered (pH 7.3) glutaraldehyde-parafomaldehyde mixture containing 5% tannic acid; after 6 hr the specimens were washed in 0.2 M cacodylate buffer (pH 7.3), and post-fixed for 2 hr in cacoylate buffered 1% osmium tetroxide. II) (TAGU) Monopylephorus limosus seminal vesicles were fixed with a mixture containing 2% glutaraldehyde, 1% tannic acid, and 1.8% sucrose in 0.1 M phosphate buffer (pH 7.2) for 2 hr, washed in distilled water, block stained for 1 hr with 2% uranyl acetate (Afzelius, 1988). III) (PAFG) Monopylephorus limosus spermathecae were fixed for 2 hr in a mixture prepared by dissolving 2 g of paraformaldehyde in 15 ml of a saturated solution of picric acid, then bringing to 100 ml with cacodylate buffer 0.15 M at pH 7.2 and finally mixing equal volumes of the above solution and of 6% aqueous glutaraldehyde (Ermak and Eakin, 1976), washed in buffer and post-fixed for 2 hr in 1% osmium tetroxide in 0.1 M cacodylate buffer (pH 7.2). The specimens were dehydrated in a graded ethanol series, embedded in Spurr’s resin, sectioned with an LKB Ultrotomes III and V and stained with lead

ET AL.

citrate-uranyl acetate-lead citrate (Daddow, 1982). For freeze-fracture, the spermatozeugmata of Tubifex and the spermathecae of Monopylephorus were processed as previously reported (Ferraguti et al., 1988). Some sections of suitably fixed Tubifex spermatozeugmata and Monopylephorus spermathecae were selectively stained for polysaccharides using periodic acid-thyocarbohydrazide-silver proteinate (Thiery, 1967). All the specimens were observed in Jeol 100 SX or 1200 electron microscope. Some electron micrographs of Monopylephorus limosus TAGU fixed sperm tails and others of tails stained for polysaccharides have been studied by computer analysis for comparison. Images were recorded in digital form by a TV camera (Sony XC77e) connected to a frame grabber (Matrox VIP 640) and elaborated by a Motorola system based on the MVME 147 single board computer (68030 microprocessor). Images of specimens fixed in the two different ways were recorded at the same enlargement and then elaborated separately. A reconstruction of each section was obtained by filtering the Fourier transform of the numerical array, according to its periodicity and symmetry; before performing the filtration, images were resampled in order to eliminate lattice distortions. Resampling was performed following the dictates of a pattern of deformations which was detected by a correlation technique; this procedure is being fully described in another paper (Lanzavecchia et al., submitted). The two periodical were finally reconstructions superimposed to verify the agreement among the different patterns. Observations The sperm tails of Monopylephorus Cimosus and of Tubifex tubifex conform respectively

to the two different models of clitellate axoneme (Ferraguti, 1984): Monopylephorus has a prominent central sheath (Figs 1, 2, 3), like the other most primitive microdriles, the branchiobdellids and the hirudineans; Tubifex has two dense fibers disposed in a plane perpendicular to the one of the central tubules (tetragon fibers of Henley, 1973); the two fibers are of different lengths, so cross

IJLTRASTRUCTURAL

ANALYSIS

OF SPERM

FLAGELLA

sections of the flagella often reveal only one of them (Fig. 9). Both species show nine double rows of glycogen granules external to the axoneme and located between neighbouring doublets (Figs 12, 16) (Anderson and Personne, 1970). Tannic acid fixed specimens show, in the principal portion of the tail, clean connections (hereafter: ‘bridges’, following the nomenclature of Dentler, 1990) between axonemal doublets and the flagellar plasma membrane (Figs 5, 10, 14). In cross sections the bridges appear as electron dense structures starting from a point close to the connection between A and B tubules of the axonemes and ending in the dense cortical web underlying plasma membrane (Figs 1.3, 9). Neither conventionally fixed material nor tannic acid preparations reveal the exact point of emergence of bridges from the doublets. Longitudinal sections show an obvious periodicity of the bridges, repeating each 45 nm, with small variations depending on the fixation techniques (Figs 4, 10, 14). The bridges appear to terminate within the cortical web. This last has a peculiar honeycomb shape around the whole periphery of the flagellum, particularly evident in grazing sections of tannic acid fixed material (Figs 10,14). In oblique sections the worps of the honeycomb appear to be in register with the bridges and with the doublets (Fig. 14). The empty spaces, the cells of the honeycomb, have a center-tocenter distance of about 4.5 nm. The arrangement of the glycogen granules as revealed by Thiery technique appears, in both species (Figs 13, 16), similar to that of the honeycomb empty spaces (cells) (Figs. 10, 14). This similarity suggests that glycogen granules occupy the cells of the honeycomb. However, it is impossible to stain, at the same time, glycogen and cortical web: glycogen is barely visible in conventionally stained specimens. Thus we have performed a computerassisted superimposition of averaged images of Monopylephorus honeycomb-shaped cortical web and glycogen granules. Figure 15 shows the complemeutarity of the two patterns. In Monopylephorus sperm tails fixed with PAFG, another periodical structure is sometimes visible: rows of short (about 9 nm long) teeth starting from the plasma membrane and directed towards the interior of the axon-

539

eme. These rows of short teeth may be arranged to form ‘muffs’ winding around the axonemes (periodicity of 15 nm: Fig. 2) or in short lines rarely visible in the longitudinal sections of the tails (periodicity of 15 nm: Fig. 4). These teeth are barely visible in TAGU fixed spermatozoa, because of the dense stain of the cortical web material (Fig. 3). PAFG fixed spermatozoa sometimes reveal the presence of two teeth in correspondence with each bridge (Fig. 1, arrows), or double rows of teeth in the longitudinal sections of the muffs (Fig. 4). Where the muffs are present, the honeycomb structure of the cortical web is interrupted. The bridges are less clearly visible because of the extremely dense tannic acid staining in the muff region. It is possible that other structures, not revealed by our techniques, are present in this area. In the Thiery-stained sections, it is evident that glycogen granules are extremely small at the level of the muffs (Fig. 16). Muffs of neighbouring sperm tails are often seen in register (Fig. 17). No muffs are visible in Tubifex tubifex sperm tails. Freeze-fractured sperm tails of Monopylephorus and of Tubifex eupyrene spermatozoa show arrays of regularly arranged 13.5 nm particles (Figs 6, 7, 8, 11). The particles, mainly on the E-face, are disposed in short double rows longitudinally arranged along the flagellar plasma membrane (Fig. 8). In Tubifex the double rows are formed by a mean of 29.57 particles (N = 57; S.D. = 7.21). In Monopylephorus the double rows are formed by a mean of 23.82 particles (N = 114; S.D. = 9.83). In both species the longitudinal double rows of particles are not in register with one another and are separated from the neighbours by a distance of about 96 nm. This figure is in good agreement with that of a chord substanding an arc of 40” (one ninth of 360”) on the circumference formed by the plasma membrane. A maximum of four parallel double rows is visible at any level in each fractured flagellum (Fig. 6). In Monopylephorus only there is also a different arrangement of the double rows of particles: as many as four or five double rows are disposed perpendicularly to the major axis of the flagellum to form muffs similar in shape and disposition to the ones observed in longitudinal sections of conventionally fixed tails (Figs 6,7). The number and distance of muffs in a single flagellum is variable.

ULTRASTRUCTURAL

ANALYSIS

OF SPERM

Discussion Structures connecting the axonemes to the plasma membranes have been known for a long time: an example are the so-called Y links, structures with the shape of a champagne glass connecting the doublets to the plasma membrane in the basal region of many cilia and flagella (see Dustin, 1984 for a review). In the limited area in which Ylinks are present (a transition area between centriole and axoneme) no central apparatus is present, giving rise to a 9+0 pattern in cross sections. There, the Y-links are connected to characteristically ordered membrane particles (necklaces). Similar connections between doublets and plasma membrane have been described in the main portion of the axonemes only in a few instances (for a review see Dentler, 1990). Curiously enough, two of the best examples of these connections are supplied

Figs l-8. Monopylephorus

541

FLAGELLA

by 9+0 axonemes, thus devoid of any central apparatus: the chordotonal sensillae of an insect (Crouau, 1980) and the sperm flagellum of the eel (Todd, 1976). On the other hand, Y-links have been reported for a long tract of Tubifex fertilizing sperm flagella by Braidotti and Ferraguti (1982). Other, more simple, axonemal-membrane connections (bridges) are reported in the literature (Dentler, 1990), but in many cases the bridges are seen only in short tracts of longitudinal sections. This may be due to osmium post-fixation (Dentler, 1981). Thus, we have treated Monopylephorus spermatozoa with the technique suggested by Afzelius (1988) which suggests that the specimens are not post-fixed in osmium, but only pre-stained in uranyl acetate, and we verified that bridges are clearly seen for the main portion of the axoneme. However, bridges were also seen, albeit more faintly, in Tubifex post-fixed with osmium but not pre-stained in uranyl acetate.

limosus spermatozoa

Fig. I. Cross vection of a PAFG fixed sperm tail. Note the prominent central sheath ‘embedding the two central tubules, barely visible. Bridges connect doublets to plasma membrane. Between neighbouring bridges the membrane is swollen. Double teeth are visible in connection with some bridges (arrows). X 135.ooO. Fig. 2. Cross section of a PAFG fixed tail at the level of a muff. Teeth are regularly all around the axoneme and the plasma membrane does not swell. x 135.000.

arranged

Fig. 3. Appearance of a section similar to the one in Figure 1 in TAGU fixed material. The preservation of the cortical web due to tannic acid probably avoids membrane swelling between the discrete bridges. x 135.000. Fig. 4. Longitudinal section of a PAFG fixed tail shows the aspect of a cross-sectioned muff (asterisk) and small teeth (short lines) between the bridges (long lines). The prominent central sheath has a serrated appearance. x 135,000. Fig. so the plasma dense

5. As in Figure 4, but TAGU fixed spermatozoa. The section is not completely sagittal, link heads crowd the center of the axoneme. The bridges regularly connect doublets and membrane. The muff (asterisk) is less resolved than in Figure 4 because of the electron marginal web. x135.000.

Fig. 6. Low magnification picture of a freeze-fractured tail bundle. Two different muffs are visible (asterisks) as well as many double rows of particles of differing lengths. In one area (arrowheads) four parallel double rows are present. Particles are mainly in the E-face. ~55,000. Fig. 7. A muff and a double row at higher magnification show a similar aspect of the particles composing the two structures, suggesting that the muff may be formed by the winding of a double row around the tail. The muff is formed by ten rows of particles, two terminating together (arrow). ~135,090. Fig. 8. Two muffs are present in the same tail at a close distance, alternating rows. Some particles are also present in the P-face (short arrowheads) together left by the E-face double rows (long arrowheads). x78,ooO.

with double with the pits

015

ULTRASTRUCTURAL

ANALYSIS

OF SPERM

It is thus evident that the bridges in the sperm tails of the two species are per se particularly evident. We have measured a longitudinal periodicity of 45 nm for the bridges in both species, but electron microscopic measures are distorted by fixation shrinkage. In fact, in a computer-assisted analysis of the flagellar structure of Monopylephorus (Ferraguti et al., in preparation), the bridges repeat every other dynein arm, i.e. with a 48 nm periodicity. This figure is greater than the figures reported in the literature [ranging from 15 nm-24 nm following the species (reviewed in Dentler, 1990)]. Dentler et al. (1980) have shown that bridges, at least in Tetrahymena and Equipecten cilia are partly formed by a dynein similar to that forming the outer arms. Our finding of a periodicity of 48 nm for the bridges could be further evidence of their nature. Freeze fracture of Monopylephorus sperm tails, as well as of Tubifex fertilizing sperm flagella, show regular arrays of particles similar to the so-called ‘zipper lines’, i.e.: ‘. . . a

Figs 9-13. Tubifex rubifex fertilizing

543

FLAGELLA

longitudinal double row of staggered paraxonemal ticles along one of the doublets . . .’ (Dallai and Afzelius, 1982). The appearance of the zipper lines in the spermatozoa of the various animal species in which they have been described varies considerably, thus prompting Millard de Montrion (1986) to question that all the structures called ‘zipper lines’ perform in fact the same function. The main inter-taxa differences concern: 1) the number of rows forming each zipper-line: from two in guinea pig (Friend and Fawcett, 1974) and mouse (Stackpole and Devorkin, 1974) to five in some hemipteran insects (Dallai and Afzelius, 1982), or even a variable number between two and five in the opossum (Olson et al., 1977); 2) the existence of interruptions in the particle rows, well evident for example in opossum (Olson et al. , 1977), occasionally present in Loligo (Olson and Link, 1980), but absent in other species (guinea pig: Friend and Fawcett, 1974); 3) the number of zippers present in a given flagellar portion: from one in many mammalian and hem-

spermatozoa.

Fig. 9. A cross section of a conventionally fixed sperm tail reveals the presence of bridges regularly connecting doublets and plasma membrane. The tetragon fiber facing doublet one is visible in the central apparatus (arrowhead). x 135,OW. Fig. 10. Longitudinal section of two TAG0 fixed neiehbourine tails: that on the rieht is paraiagittally sectioned and shows regularity of bridges (l&es); tha;on the left is grazed b; the plane of section and show the honeycomb aspect of the cortical web. x 135,ooO. Fig. 11. In longitudinal fractures, the double rows of particles are very similar to those of Monopykphorus. but no muff is visible. Some particles are present on the P-face (arrowheads). x70,OGQ. Figs 12, 13. Transverse (Fig. 12) and longitudinal tails. The glycogen granules are regularly arranged.

(Fig. 13) sections of Thitry stained sperm Fig. 12 x 135,COO. Fig. 13 x 70.000.

Fig. 14. Longitudinal oblique section of TAGU fixed Monopylephorus sperm tail. The section passes through, on top, the central sheath and two opposing doublets, then two neighbouring doublets and finally the cortical web. At the bottom of the figure the honeycomb structure of the cortical web is evident. x 105,000. Fig. 15. Computer-assisted superimposition of average images of Monopylephorus honeycomb cortical web and glycogen granules. The glycogen granules fit very well the cells of the cortical web. ~230,000. Fig. 16. A longitudinal section of a conventionally fixed Monopylephorur sperm tail has been stained with the Thi6ry method for polysaccharides. The arrangement of glycogen granules is similar to that of Tubifex with the exception of the muff area in which the amount of glycogen is extremely reduced (asterisk). x73.ooO. Fig. 17. A grazing section of two neighbouring

tails showing two muffs in register.

x 100,000.

544

FERRAGUTI

ET AL.

b

La.

Fig. 18. This schematic drawing represents author’s view of the arrangement and mutual connections of the various parts of Monopylephoruc periaxonemal area. Only two neighbouring outer doublets are shown: one is folded towards the interior, and the other one is devoid of most of the A tubule to show the periaxonemal area. Dynein arms are sketched only to show a consistent difference between inner and outer arms that will be described elsewhere (Ferraguti er al., in preparation). Radial links have been omitted. b = bridges; t = teeth; d = double rows of particles; * = muff; g = glycogen; w = cortical web; B, A = doublet tubules; i.a. = inner arms; 0.a. = outer arms

ipteran spermatozoa (Dallai and Afzelius, 1982) to (probably) nine in the cephalopod Loligo (Olson and Link, 1980) and nine in the gastropod Onchidella celtica (Selmi et al., 1988) and in the boar terminal piece (Suzuki and Nagano, 1980). Two different types of function have been proposed for the zipperlines, in relation with their different aspect: mechanical connections between plasma membrane and various underlying structures, and transport of substances across the axonemal membrane. A possible model for bridges and flagellar membrane particles (Fig. 18)

We propose

here a model for the regular

arrays of particles in Tubifex and MonThis model is peculiar, even if some of its features are present in other species. We hypothesize that there are nine parallel double rows of membrane particles for each flagellum: in fact, even if we have rarely observed a maximum of four double rows in each longitudinal fracture of the flagella, the distance between two neighbouring double rows is consistent with the hypothesis. Two different structures have a nine-fold symmetry in the sperm tails of the two species: the doublets, thus also the bridges and the worps of the honeycomb, and the double rows of glycogen granules. Both structures are separated by a distance comopylephorus.

ULTRASTRUCTURAL

ANALYSIS

OF SPERM

545

FLAGELLA

parable to that found between the double rows of particles in the fractures. However, no bridge and no teeth have been observed between glycogen granules and flagellar membrane, whereas the periodicity of the teeth is in good agreement with the one of the double rows of membrane particles. Our double rows are interrupted and show a certain constancy of length both within and between the two species. The particles in each double row are in register, i.e. they are not disposed as in a zipper like in the other models. A unique structure present in Monopylephorus sperm tail is the muff. We hypothesize that the same particles forming the double rows are arranged perpendicularly to the main axis of the flagellum to form the muffs: it is highly significant that the parallel rows of particles forming each muff are always in even number. Furthermore we often have observed images in which a double row terminates abruptly in a muff (Fig. 7). Longitudinal and cross sections of the muffs of PAFG fixed material reveal a characteristic increasing of the sperm tail diameter (Fig. 16): in these areas the small teeth underlying plasma membrane are evident. It is impossible to observe the teeth underlying plasma membrane in tannic acid fixed material, since, by this technique, the cortical web becomes extremely electron dense and its thickness is the same as that of teeth. If our model is correct, the appearance of the teeth in the cross sections of the flagella at the level of the muffs should be the same as the one of longitudinal sections cutting the double rows. In fact we have observed teeth underlying the double rows of particles only in some of the best sections. This can be explained by comparing the thickness of the sections with the diameter of the teeth. Conclusions

Despite the differences in the ultrastructural morphology of the axonemal machinery, the periaxonemal region of the two oligochaete species considered in this paper is remarkably similar. The system formed by bridges, teeth and zipper lines in the two oligochaete species, considered here, seems to perform a mechanical function, perhaps as a consequence of the relevant distance between axonemal doublets and membrane. Between

neighbouring bridges, the plasma membrane is evidently swollen, but not at the level of the muffs. A somewhat similar geometry of the zipper lines is present in the cephalopod Lo&o. It may be relevant to note that in this animal the distance between doublets and membrane is about the same as in oligochaetes. In the squid, however, accessory fibers are interposed between doublets and plasma membrane in correspondence with zipper lines and no bridges are visible. Besides the mechanical functions it may be possible that, following the model proposed by Dentler et al. (1980), the dynein bridges starting from the doublets interact with a membrane tubulin (Stephens et al., 1987). This last could be localized, in our model, within the teeth embedded in the cortical web and connected to zipper line particles. Thus a second dynein-tubulin system would support the dynein-tubulin interaction responsible for the doublet sliding. A striking difference between Tubifex and Monopylephorus sperm tails resides in the presence of the muffs surrounding the axonemes in Monopylephorus. This fact could be perhaps related to the different arrangement of the spermatozoa in the two species: in Tubifex, in fact, the spermatozoa are confined parallel in the axial cylinder of the spermatozeugmata, packed by the cortex. The cortex with its movements bring the spermatozoa towards the egg (Ferraguti et al., 1988). On the contrary, in Monopylephorus no spermatozeugma is present and the sperm is contained in infoldings of the spermathecal walls. The muffs, often found at the same level in neighbouring tails could have some function in holding the spermatozoa together (Fig. 17). Acknowledgements

The authors wish to thank Roman0 Dallai for the freeze fracture replicas of Tubifex, Andreina Paoletti for the identification of the animals, Marina Camatini for the use of freeze-fracture apparatus, Giovanni Bernardini and Giulio Lanzavecchia for their helpful suggestions during the research and comments on the manuscript, Daniela Ruprecht for her technical help and Marco Picco for the drawing. This research has been supported by an M.P.I. 40% grant (Interazioni Cellulari).

546

FERRAGUTI

ET AL.

References Afzelius, B. A. 1988. Microtubules in spermatids of stick insects. J. Ultrasrruct. Mol. Srruct. Res., 98,94-102. Anderson, W. A. and Personne, P. 1970. The localization of glycogen in the spermatozoa of various invertebrates and vertebrate species. J. Cell Rim’., 44,29-51. Baker, H. R. and Brinkhurst, R. 0. 1981. A revision of the genus Monopilephoms and redefinition of the subfamily Rhyacodrilinae and Branchiurinae (Tubificidae, Oligochaeta). Can. J. Zool., 59, 939-%5. Braidotti, P. and Ferraguti, M. 1982. Two sperm types in spermatozeugmata of Tubifex tubifex (Ann&da, Oligochaeta). J. Morpho[. , 171, 123-136. Crouau, Y. 1980. Comparison of a new structure associated with the membrane of 9+0 cilia of chordotonal sensilla with the central structure of motile cilia and flagella. Eiol. Cell., 39, 349352. Daddow, L. Y. M. 1982. A double lead stain method for enhancing contrast of ultrathin sections in electron microscopy: a modified multiple staining technique. J. Microsc., 129, 147-153. Dallai, A. and Afzelius, B. A. 1982. On zipper-lines of particle arrays within the plasma membrane of hemipteran spermatozoa (Heteroptera, Insecta). J. Ultmtruct. Res., 80, 197-205. Dentler, W. L. 1981. Microtubule-membrane interactions in ctenophore swimming plate cilia. Tissue Cell, 13, 197208. Dentler, W. L. 1990. Linkages between microtubules and membranes in cilia and flagella. In Ciliary and Flagellar Membranes (eds. R. A. Bloodgood), pp. 31-64. Plenum Press, New York. Dentler, W. L., Pratt, M. M. and Stephens, R. E. 1980. Microtubule-membrane interactions in cilia. II. Photochemical cross-linking of bridge structures and the identification of a membrane-associated dynein-like ATPase. J. Cell Biol., 84,381-403. Dustin, P. 1984. Microtubules. Springer Verlag, Berlin. Ermak, T. H. and Eakin, R. M. 1976. Fine structure of the cerebral and pygidial ocelli in Chone ecaudata (Polychaeta: Sabellidae). J. Ulfrasfmcr. Res., 54, 24f260. Erseus, C. 1990. Cladistic analysis of the subfamilies within the Tubiticidae (Oligochaeta). Zool. Scripta, 19, 57-63. Erstus, C. and Paoletti, A. 1986. An Italian record of the aquatic ohgochaete Monophylephorus limoslls (Tubificidae), previously known only from Japan and China. 2001. Scripta, 15, 45-51. Ferraguti, M. 1984. The comparative ultrastructure of sperm flagella central sheath reveals a new autapomorphy of the group. 2001. Scripta, 13, 201-207. Ferraguti, M., Grassi, G. L. and Em&is, C. 1989. Different models of tubiticid spermatozeugmata. Hydrobiologia, 180, 7>82. Ferraguti, M., Bernardini, G., Melone, G. and Dallai, R. 1988. Structure and function of the metachronal wave in Tubifex tubifex spermatozeugmata (Annelida, Oligochaeta). 1. Ultrastruct. Mol. Struct. Res., 99,79-95. Friend, D. S. and Fawcett, D. W. 1974. Membrane differentiations in freeze-fractured mammalian sperm. J. Cell Biol., 63, 641464.

Henley, C. 1973. Ultrastructure

of the spermatozoon

of the earthworm as revealed by negative stain. J. Morphol.,

140, 197-213.

Millard de Montrion, C. 1986. Zipper lines in the flagellar plasma membranes of the squid (Loligo) spermatozoon. Tissue Cell, l&251-265. Olson, G. E. and Link, R. W. 1980. Membrane differentiations in spermatozoa of the squid, Loligo pealeii. Gamete Res., 3, 329-342.

Olson, G. E., Lifsics, M., Fawcett, D. W. and Hamilton, D. W. 1977. Structural specializations in the flagellar plasma membrane of opossum spermatozoa. J. Ulfrmtruct. Res., 59, 207-221. Selmi, M. G., Giusti, F. and Manganelli, G. 1988. The fine structure of the mature spermatozoon of Onchidella celtica (Cuvier) (Mollusca: Gastropoda) and its phylogenetic implications. J. Ultra.struct. Mol. Strucf. Res., 99,107-123. Stackpole, C. W. and Devorkin, D. 1974. Membrane organization in mouse spermatozoa revealed by freeze-etching. J. Ultrastruct. Res., 49, 167-187.

Stephens, R. E., Oleszko-Szuts, S. and Good, M. .I. 1987. Evidence that tubulin forms an integral membrane skeleton in molluscan gill cilia. J. Cell Sci., 88,S27-S36. Suzuki, F. and Nagano, T. 1980. Morphological relationship between the plasma membrane and microtubules in the end piece of the boar spermatozoon. J. Electron Microsc., 29, 190-192. Thiery, J. P. 1967. Mise en evidence des polysaccharides sur coupes fines en microscopic Clectronique. J. Microscopic, 6, 987.

Todd, P. R. 1976. Ultrastructure of the spermatozoa and spermiogenesis in New Zealand freshwater eels (Anguillidae). Cell Tiss. Res., 171, 221-232.