The intracellular origin of flagellar hairs in the dinoflagellate Woloszynskia micra leadbeater & dodge

The intracellular origin of flagellar hairs has now been demonstrated in a wide range of motile cells. Manton, Rayns, Ettl & Parke (1965) first noted that the fine caducous hairs on the flagella of 2 species of Heteromastix (Prasinophyceae) were produced within vesicles of the Golgi apparatus. More recently it has been shown that the tubular flagellar hairs found on many algal and fungal zoids are produced within vesicles of the endoplasmic reticulum (ER) and dilations of the nuclear envelope (Bouck, 1969; Leedale, Leadbeater & Massalski, 1970; Heath, Greenwood & Griffiths, 1970).

The dinoflagellates differ from all other motile cells except the euglenoid flagellates (Leedale, 1967) in that they possess a single row of long fine hairs on one of their flagella (Pitelka & Schooley, 1955; Leadbeater & Dodge, 1967,a, b; Dodge, 1967; Dodge & Crawford, 1968). An intensive study of the fine structure of Woloszynskia miera Leadbeater & Dodge (Leadbeater & Dodge, 1966; Leadbeater, 1967) revealed the presence of large dilations of the nuclear envelope and ER that contained bundles of fibrils of unknown significance. As will be shown below, the similarity in the position and appearance of these dilations to those containing flagellar hairs in other flagellates (see Leedale et al. 1970 etc.) together with the similarity in dimensions and appearance of the individual fibrils to the flagellar hairs suggests that in the dinoflagellate W. miera these also originate within the cell.

Two unialgal isolates of Woloszynskia miera Leadbeater & Dodge (Plymouth culture collection nos. 207 and 366) were supplied by Dr M. Parke of the Marine Biological Association, Plymouth. Cultures of the isolates were maintained in Erdschreiber seawater at 20 °C under 16 h light/8 h dark cycle.

For light microscopy a small quantity of culture was fixed in a solution of 2 % osmium tetroxide in 01 M acetate veronal buffer at pH 7 0. Cells were viewed with anoptral contrast microscopy.

For electron microscopy of whole mounts, cells were concentrated by gentle centrifugation, fixed in a solution of 2% osmium tetroxide in acetate-veronal buffer at pH 7·0, washed and then dried on Formvar-coated grids. Some specimens were shadowcast with gold/palladium and others were negatively stained with sodium phosphotungstate at pH 7·0.

Material for embedding was centrifuged to form a pellet and then fixed for 1·5 h in ice-cold 5 % glutaraldehyde in a balanced salt solution with acetate-veronal buffer at pH 7·8. This was followed by washing in buffer, postosmication in 2 % osmium tetroxide in acetate-veronal at pH 7·8 (2 h), dehydration in a graded ethanol series, and embedding in Araldite. Sections were cut with a glass knife on a Porter Blum I ultramicrotome, stained with methanolic uranyl acetate followed by Reynold’s lead citrate. Most observations were made on a Zeiss EM 9 microscope at Birkbeck College, London. Supplementary observations were made on an AEI EM 6 microscope in Birmingham.

Cells of Woloszynskia miera are approximately 9–15 μm long and 8–14 μm wide and are divided almost equally into an upper epicone and lower hypocone by the almost transverse girdle or sulcus (Fig. 2). Two flagella emerge from the cell in the mid-ventral region. The longitudinal flagellum (about 10 μm long) extends posteriorly whilst the transverse flagellum (about 30 /μm long) encircles the cell within the girdle.

The longitudinal flagellum of W. miera bears a covering of short fine hairs (Fig. 3). On dried whole mounts the hairs appear to be borne bilaterally but whether this is an artifact caused by the method of preparation is not known. The hairs, approximately 10 nm in width and 0·5 μm in length, show no differentiation at the base or tip.

The sheath surrounding the transverse flagellum is expanded laterally and contains the axoneme and ‘striated strand’ (see Leadbeater & Dodge, 1967 a) separated by packing material. Dried whole-cell mounts show that the axoneme undulates, simulating a sine curve when flattened, whilst the accessory striated strand follows a shorter, almost straight course (Fig. 4). The relationship between the axoneme and the striated strand can be seen best in fixed cells mounted in water and viewed with anoptral contrast microscopy (Fig. 1). The axoneme winds around the shorter striated strand which runs along the centre of the helix. This arrangement results in the helical movement of the flagellum during swimming. The transverse flagellum bears a unilateral row of long (approximately 3–4 μm), fine hairs which appear to be attached to the sheath adjacent to the axoneme (Fig. 4). In shadowcast whole mounts (Fig. 4) the flexible hairs frequently clump together in groups of 2 or 3. When negatively stained the hairs are more rigid in appearance but are of regular width throughout their length (Fig. 5). In both shadowcast and negatively stained preparations the average width is 10 nm.

In sections examined by electron microscopy, most cells can be seen to contain large dilations of either the nuclear envelope or the ER (Figs. 6–9, 11, 12) and the latter may be closely associated with the Golgi apparatus (Fig. 9). These dilated profiles, which may be up to 5 μm in length, usually contain bundles of fibrils (Figs. 7–12) which are now interpretable as developing or mature flagellar hairs. This interpretation is based on the equivalent location and appearance of vesicles known to contain developing flagellar hairs in other algal and fungal zoids (see Bouck, 1969; Leedale et al. 1970; Heath et al. 1970 for list of species) and on the similarity, in appearance and width, between the individual fibrils within the vesicles and the flagellar hairs.

Although the individual fibrils cannot be clearly seen in permanganate-fixed cells (Fig. 6 and inset) the membrane surrounding the vesicles (v¹ and v², Fig. 6) and their connexion with the ER are obvious (Fig. 6, inset). On the other hand in cells postosmicated after glutaraldehyde fixation the fibrillar nature of the contents is distinct (Figs. 7–12). In whatever plane a large dilation is sectioned, fibrils running in more than one direction can usually be found. When observed in side view the fibrils are seen as delicate, more or less parallel lines sometimes marked by irregular transverse striations (Fig. 12) which may be artifacts. In transverse section the extreme thinness of the hairs is more suggestive of a fine granular deposit (Figs. 8, 9, 12). As the fibrils are so delicate no developmental stages have been observed. The approximate width of mature fibrils is 10 nm (Fig. 10). This is substantially less than the width of the comparable fibres in Olisthodiscus luteus Carter (Leadbeater, 1969) and the other heterokont organisms enumerated below but agrees closely with the flagellar hairs of this organism. The length of individual fibrils varies according to the plane of section but in median longitudinal section, fibrils measuring 5 μm have been recorded.

Under normal light conditions (16 h light/8 h dark) only a few cells possess dilations of the nuclear envelope. In cultures that have been maintained in darkness for 3 days prior to fixation the number of cells with perinuclear dilations increases from approximately 2 to 55 % (see Table 1).

A section of a cell fixed after 3 days continuous darkness is illustrated in Fig. 7. The mitochondria are large, very little ER is present and there is an accumulation of fibrillar material within the perinuclear space (arrow).

In cells grown under normal light conditions the dilations of the nuclear envelope appear to detach from the nucleus and come into close association with the Golgi apparatus (Fig. 9). These fibril-containing vesicles (f) also lie close to vesicles containing trichocyst precursors (t) (Fig. 11). The latter are probably of Golgi origin. However, on no occasion has any definite membrane contact been observed between fibril-containing ER vesicles and either the Golgi apparatus or vesicles containing trichocyst precursors. Movement of the mature fibrils to the surface of the cell and their subsequent discharge on to the flagella has not been observed.

The general morphology of Woloszynskia miera has already been well documented (Leadbeater & Dodge, 1966, 1967a, b;Leadbeater, 1967). The very fine hairs on transverse flagella have also been recorded in other dinoflagellate genera, for example Gymnodinium sp. (Pitelka & Schooley, 1955), Aureodinium pigmentosum Dodge (Dodge, 1967) and Amphidinium carteri Hulbert (Dodge & Crawford, 1968). Recent observations on dinoflagellates collected from Norwegian waters show that the presence of long, fine hairs is a standard feature of transverse flagella (B. Leadbeater, unpublished observations). The occurrence of fine hairs on the longitudinal flagellum does not appear to be so common. Pitelka & Schooley (1955) did not observe hairs on the longitudinal flagellum of Gymnodinium sp. and they were not present on the longitudinal flagella of Aureodinium pigmentosum (Dodge, 1967) or Amphidinium carteri (Dodge & Crawford, 1968).

The average width of the hairs on both the transverse and longitudinal flagella is 10 nm irrespective of the method of preparation. This is identical to the average width of fibrils located within dilations of the nuclear envelope and ER. Furthermore, the fine fibrils found within dilations of the nuclear envelope in Amphidinium carteri are also approximately 10 nm in width (measurement taken from fig. 21 in Dodge & Crawford, 1968). The length of the hairs of the transverse flagellum is approximately 5 μm and in exceptional circumstances when exact median longitudinal section of the intracellular fibrils is obtained, a length of 5 μm has been recorded. At present, it is not possible to distinguish between intrinsically short fibrils which might eventually form the short hairs of the longitudinal flagellum and oblique or tangential sections of longer fibrils that will ultimately form the hairs of the transverse flagellum. Whether the hairs destined for a longitudinal flagellum are formed in separate dilations from those destined for a transverse flagellum is also unknown.

The presence in dinoflagellates of vesicles containing fibrils, sometimes arranged in bundles, has been recorded by several authors. Bouck & Sweeney (1966, fig. 8) observed ‘vacuoles with fibrous contents of unknown significance’ in Prorocentrum micans Ehrenberg. Dodge (1967, fig. 6) found a ‘variably shaped fibrous body’ which was frequently associated or adjacent to the nucleus in Aureodinium pigmentosum. In Amphidinium carteri the ‘fibrous body’ was found to develop from a dilation of the nuclear envelope (Dodge & Crawford, 1968). Leadbeater (1967) made a thorough study of fibril-containing dilations in Woloszynskia miera. The fibrils were located within dilations of the nuclear envelope and the ER and their close association with the Golgi apparatus and trichocyst precursors was well established. However, as the significance of these fibrils was unknown the record was incomplete. Following the recent observations on the intracellular origin of flagellar hairs (Bouck, 1969; Leedale et al. 1970; Heath et al. 1970) it became obvious that there was a striking similarity in location and appearance between the dilations of the nuclear envelope and ER containing flagellar hairs in antherozoids of Ascophyllum nodosum (L.) Le Joi and Fucus spp. (Bouck, 1969); Olisthodiscus luteus Carter, Bumilleria sicula Borzi, Heterococcus spp., Tribonema spp. (Leedale et al. 1970); Saprolegnia ferax (Gruithuisen) Thuret, Dictyuchus sterile Coker, Synura caroliniana Whitehand, Cryptomonas sp. (Heath et al. 1970) and those containing fibrils in Woloszynskia miera.

The movement of the fibril-containing dilations in the cell under different light conditions appears to be related to the fate of the ER. During dark treatment the ER disappears from the cell cytoplasm and this probably prevents the fibril-containing dilations of the nuclear envelope from moving into the surrounding cytoplasm. This explains why there is a considerable increase in the number of cells with fibrilcontaining dilations of the nuclear envelope. When the cells are returned to normal light conditions ER is re-formed and the dilations pass into the rest of the cytoplasm.

The close spatial association of the fibril-containing dilations in W. miera with the Golgi apparatus (Fig. 9; also figs. 6, 8 in Leadbeater & Dodge, 1966) and the trichocyst precursors (Fig. 11) does not appear to imply structural continuity. The fibril-containing dilations in Prorocentrum micans are similarly in close association with trichocyst precursors (Bouck, 1969, fig. 7), and the ER vesicles containing flagellar hairs in Tribonema vulgare (Leedale et al. 1970, fig. 23) are in close association with cisternae and vesicles of the Golgi apparatus.

The final migration of the fibril-containing ER dilations to the cell surface and the subsequent discharge of the fibrils on to the flagella has not been observed. This, unfortunately, is also the gap in the reports of Leedale et al. (1970) and Heath et al. (1970). Bouck (1969) only observed the outward migration of the vesicles containing presumptive mastigonemes in Ascophyllum and Fucus but was unable to actually observe the deposition of the hairs on the flagellum.

It now appears that the intracellular origin of flagellar hairs is standard for a wide range of algal and some fungal zoids. However the evidence in Woloszynskia miera is in no case complete. The absence of information on the mode of deposition has already been noted and chemical data are urgently required before equivalence can be claimed with finality.

I would like to express my grateful thanks to Dr J. D. Dodge for supervising this work and reading the script; Dr M. Parke for kindly supplying the cultures of W. miera, and Professor I. Manton, F.R.S., for commenting on the plates and results. Finally I am grateful to S. R. C. for financial support.