Cell Biology International 28 (2004) 387e396 www.elsevier.com/locate/cellbi

Analysis of the contraction of an organelle using its birefringency: the R-fibre of the Ceratium (Dinoflagellate) flagellum* ) Hidemi Satoa, Claude Greuetb, Monique Cachonb, Jacky Cossonb,

aNagano University, Shimonogo, Ueda-shi, Nagano 386-12, Japan bU.M.R. 7009 du CNRS, Biologie du De´veloppement, Observatoire Oce´anologique de Villefranche-sur-Mer, 06230 Villefranche-sur-mer, France

Received 23 September 2003; revised 19 January 2004; accepted 15 March 2004

Abstract

Some organelles responsible for contraction consist of bundles of 2e4 nm filaments called nanofilaments. Such organelles are present in the longitudinal flagellum of Ceratium (Dinoflagellate): the R-fibre is the motor system for contraction and parallels the axoneme, which is responsible for wave generation. We used a highly sensitive polarization microscope developed by one of the authors to measure the birefringence of these nanofilament bundles during contraction in vivo. Our results show that the R-fibre gives a highly birefringent signal, retarding the polarization to much the same extent irrespective of the direction of polarization. By rotating the axis of the microscope compensator we confirmed that the birefringence is positive, suggesting that the bundles run parallel to the longitudinal axis of the flagellum. Conversely, when the compensator was rotated contrary to the direction of retardation, the bundle appeared dark (except when the organelle was in a fully contracted state). Experiments performed on detergent-treated and ATP-reactivated flagella show that a portion of the flagella regained activity with the addition of ATP in the presence of low Ca2C concentrations. This demonstrates the ability to reactivate flagellar motility after permeabilization and that axonemal microtubules were not responsible for the strong flagellar birefringence. Combined with complementary data from DIC microscopy of demembranated flagella and electron microscopy, these findings have led to the development of a model of the R-fibre and a comparison with other types of birefringent nanofilament bundles, such as the myoneme of Acantharia. Ó 2004 International Federation for Cell Biology. Published by Elsevier Ltd. All rights reserved.

Keywords: Polarizing microscopy; Nanofilaments; Myoneme; Flagellar motility

1. Introduction (Cachon et al., 1983) and myonemes (Febvre, 1971) of Acantharia. Some protists are able, by virtue of special organelles, All these organelles have 2e4 nm nanofilament to contract parts of their cell bodies. Examples are bundles in their structure. Optical anisotropy has been found in ciliates: the stalk of Vorticella (Amos, 1971), studied by polarizing microscopy in only a few such the appendix of Tontonia (Greuet et al., 1986), the organelles (the Acantharian myonemes, Febvre et al., flagellar rootlets of Chlamydomonas (Salisbury and 1990): they are birefringent and the sign of birefringence Floyd, 1978), and in Dinoflagellates the peduncule can be either positive or negative. Therefore polarization of Noctiluca (Soyer, 1968, 1970), Erythropsidinium microscopy appears as a technique of choice for in vivo (Greuet, 1967, 1981) and Leptophyllus (Cachon and quantitative measurements of the degree of compaction Cachon, 1964). Other examples are the flagellar rootlets of the flagellar organelles made of arrays of very small filaments (2e4 nm) which parallel the axoneme, in the longitudinal flagellum of Dinoflagellates of the Ceratium * This publication is presented in honor of Prof. Hidemi Sato on the occasion of his 77th birthday. group. The aim of polarization microscopy is mostly to ) Corresponding author. quantify low levels of local birefringence, so-called E-mail address: [email protected] (J. Cosson). ‘retardance’ of light in a specimen observed at very

1065-6995/$ - see front matter Ó 2004 International Federation for Cell Biology. Published by Elsevier Ltd. All rights reserved. doi:10.1016/j.cellbi.2004.03.007 388 H. Sato et al. / Cell Biology International 28 (2004) 387e396 high magnification. Retardance is evaluated in radians a Panasonic CCD camera and a Super VHS (NTSC) by rotation of the polarization filter (compensator) in- video tape recorder. The nature and sign of birefrin- cluded in the optical path. As these measurements need gence in flagella were determined by altering the re- also highly sensitive detection (video camera or film) we fractive indices of the surrounding media (Sato et al., used a highly sensitive polarizing microscope to obtain 1975). Images were obtained either from video monitor information about the contraction state of the bundles photographs, or by digitalisation of still frames recorded of nanofilaments. So far, most studies of birefringent on S-VHS and enhanced using MacIntosh software. organelles were devoted to chromosomal organisation (Cachon et al., 1989) or mitotic spindles (Sato et al., 2.3. Electron microscopy 1975). In our present study on the Ceratium flagellum, the latter always appeared to be birefringent, either Fixation was carried out according to Maruyama while it contracts or during its contractionerelaxation (1981,1982). The unicells were fixed in 0.1 M phosphate cycles (Cachon et al., 1991, 1992). In the present paper buffer containing 5% glutaraldehyde, 0.8e1.0 M glu- we describe additional results obtained after demem- cose, pH 7.4e7.8 at room temperature for 1 h, washed branation and other treatments of the flagellum, as well in 0.3 M phosphate and 0.8 M glucose, post-fixed in 2% as ultrastructural features by electron microscopy. OsO4 in phosphateeglucose buffer, progressively dehy- drated, then finally embedded in Spurr’s low viscosity medium (Spurr, 1969). Sections were cut with a MT2 2. Materials and methods Porter Blum microtome, stained with 9% uranyl acetate in methanol followed by lead citrate, then examined 2.1. Biological materials with a Hitachi H 603 electron microscope.

Two species of Ceratium, C. furca (Ehrenberg) and C. 2.4. Detachment of flagella, permeabilization limulus (Gourret), were collected by filtration with thin and reactivation nylon nets (5e20 mm mesh) in Villefranche Bay (France) and Calvi Bay (STARESO, Corsica), and during the 1991 We intended to solubilize the microtubules of the autumn in Toba Bay (Marine Biological Laboratory of axoneme to see if the bundle of nanofilaments was still Sugashima, Nagoya University, Japan). Individual speci- positively birefringent. In some experiments flagella mens were isolated with a Pasteur pipette under a binocular were detached from the cells by an increase of the Ca2C microscope and maintained at room temperature for several concentration up to 20 mM in seawater containing hours in Petri dishes containing Millipore-filtered seawater. 20 mM TriseCl pH 8.2 to prevent any pH shift. A medium, commonly used for dynein coupled activity 2.2. Polarizing microscopy and DIC microscopy (Cosson et al., 1988) was used to permeabilize in situ or detached flagella. This medium is composed of: 0.1 M The specimens were suspended in seawater on a glass potassium glutamate, 0.5 M glucose, 20 mM TriseCl, slide with a coverslip and a thin layer of spacer (Valap, 2 mM EGTA, 4e10 mM Mg acetate, 1 mM dithiothrei- a mixture of Vaseline, Lanoline and Paraffin 2:2:1). A tol, 0.007% Triton X-100, pH adjusted to 7.8 with highly sensitive polarizing microscope, designed by Sato KOH (chemicals were from Sigma, St Quentin, 38297 et al. (1991), specially constructed by Nikon Engineering France). Most flagella remained contracted inside the Co. and located in the Sugashima Marine Biological flagellar pocket, but in a few cases isolated flagella were Laboratory, was used for all observations. All optical observed. These were reactivated by addition of 1 mM lenses including Apochromatic 60! objective lenses ATP, as previously observed for Oxyrrhis flagella (K.K. Nikon, Tokyo and Olympus Optical K.K. Tokyo) (Cosson et al., 1988). For assays of microtubule sliding, and a rectifier (Nikon, K.K.) were selected for high optical the axonemes were briefly digested in the same buffer as quality. The system was set on a vertical optical bench and above with trypsin (0.1 mg/ml) at room temperature. images were collected with a Hamamatsu C.1 100 type SIT camera (Hamamatsu Photonics, Hamamatsu), which can be switched to a photomicrographic recording system 3. Results by 90( rotation of the supporting device (Sato et al., 1991). In addition, this microscope arrangement can 3.1. Behaviour of the flagellum of native easily be converted to DIC (Differential Interference cells observed with DIC optics (Fig. 1) Contrast) by inserting two Wollaston prisms into the optical path. All retardation measurements were obtained The longitudinal flagellum of Ceratium issues from with a BraceeKo¨hler type compensator. a long basal body located at the bottom of a deep The flagellar movement patterns and associated cylindrical flagellar pocket about 6 mm in diameter and changes of birefringence were routinely recorded with 16 mm in length (Cachon et al., 1992). This flagellum H. Sato et al. / Cell Biology International 28 (2004) 387e396 389

Fig. 1. Photographs obtained with Differential Interference Contrast (DIC in aec) and Dark Field (DF in dei) microscopy of three different states of contraction of the longitudinal flagellum of C. furca in aec where only the flagellated half of the cell is visible and of C. limulus in dei only the flagellum is visible, arrow head at origin. In d and in gei, multistrobe illumination at 150 Hz flash frequency. a and d Z relaxed state, with sine waves; b and e Z beginning of contraction (20 ms); c and f Z fully contracted at 40 ms (arrow). gei Z decontraction (1, 2 & 3 s after contraction, respectively). Bar scale in lower part of i panel Z 100 mm. Cell body located on top left of each photograph. reaches 200e300 mm length in some individuals and it state of the elements present in the flagellum, the beats in pseudo-sinusoidal waves, which are quasi-planar axoneme itself, which is paralleled by fibres (see details and circular arc shaped (Brokaw and Wright, 1963). Up below). The elementary filaments of the fibres being in to 10 half-wavelengths ( five full sine waves) can be the size range of 3e4 nm makes polarization microscopy observed along the flagellum when intact. The wave an excellent technique combining a high optical re- amplitude is limited inside the flagellar pocket but its beat solution (but needing high numerical aperture lenses) frequency is the same as outside the pocket (40e50 Hz). with the measurement of compaction of such arrays of From time to time, contact with various agents ( particles, filaments. The main goal is to obtain the measure- surfaces or other unicells) or in the absence of any obvious ment of I, the irradiance of the specimen, which is signal causes arrest of wave generation and sudden defined as retraction of the flagellum into the flagellar pocket, where C it appears multifolded; up to 30 half-waves can be I ¼ Ipðsin 2ðR=2Þ ð1=EFÞÞ e e observed in this coiled state (Fig. 1b c and e f); the where R is the retardance in radians and EF is the distal extremity often remains outside the pocket and extinction factor. The extinction factor EF is defined eventually uncoils a few seconds later. It then extends as the ratio I /I where I is the measurement of the e p c p again (Fig. 1g i), but much more slowly (over a time- irradiance of the field when the polarizer and the ana- course of seconds) than it retracts (Cachon et al., 1992). lyser are parallel and Ic is the irradiance when the retardance R ¼ 0(Swann and Mitchison, 1950). 3.2. Behaviour of the flagella of native cells observed The longitudinal axis of the flagellum was oriented at with the polarizing microscope (Figs. 2 and 3) 45( to the vibrating plane of polarizing light in order to measure the polarization retardation. The sign of the The choice of using polarization microscopy is birefringence was determined in this orientation by mainly because we wanted to quantify the contraction carefully rotating the compensator in the additive or the 390 H. Sato et al. / Cell Biology International 28 (2004) 387e396

Fig. 2. Micrographs of different states of the longitudinal flagellum of C. limulus through its relaxationecontraction cycle on a birefringence level obtained with the light polarization microscope and their corresponding schematic representations in each vertical column. (Bar size in upper left panel Z 100 mm). The cell body is brightly birefringent in the lower portion of each photograph (upper row of panels) and the flagellum (basal part indicated by arrow head) is always positively birefringent when parallel to the axis of the compensator (columns a, b and d). In b, the orientation of the flagellum relative to the axis of light polarization changes along distance on the flagellum (upper panel). If perpendicular to this axis, it is negatively birefringent except when highly contracted (column c). subtractive direction. Retardation was determined ac- known value of retardation, compensates the retarda- cording to the formula: tion of the specimen; D is the specimen retardation; and Dc is the specific retardation constant of the compen- sin 2q ! sinðD=2Þ¼sin 2qt ! sinðDc=2Þ sator. D and Dc are expressed in degrees (Hartshorne where q is the azimuth orientation of the specimen (here and Stuart, 1960). maintained at 45(G5(); qt is the azimuth angle at which Whatever be the state of contractionerelaxation of the BraceeKo¨hler type compensator, which possesses a the flagellum, it always appeared positively birefringent, H. Sato et al. / Cell Biology International 28 (2004) 387e396 391

The total diameter of the flagellum is estimated by electron microscopy as about 200 nm. Thus, the coefficient of birefringence (ne no), which is expressed as retardation divided by thickness (diameter), is calculated as close to 1!102, which strongly suggests that the flagellum contains a closely packed filamentous bundle (nanofilaments).

3.3. Observations on permeabilized flagella (Fig. 4)

After detergent treatment and addition of ATP, some flagella (10e30%) were observed to reactivate motility, either in situ (see Fig. 4) out of their flagellar pocket or in less cases (10e15%), detached from the cell body, but only in conditions where Ca2C ions were at low con- centration (lower than 108 M free Ca), i.e. when the reactivation medium contained EGTA as described in Section 2. In contrast when EGTA was omitted ( free estimated Ca concentration of 1 mM), no wave move- ment reactivation occurred and in addition, flagellar axonemes were difficult to observe because of their extreme coiling, due to the high contraction of the paraaxonemal structure (R-fibre). Fig. 4B shows a per- meabilized flagellum observed while beating in situ: remnants of membrane are still present (arrow heads), indicating that the detergent treatment was mild. The permeabilized/reactivated flagella develop large ampli- tude waves with beat frequency 45G4 Hz in the presence of 1 mM ATP. Fig. 4CeE show successive positions (numbered 1e9) of the flagellum with a progression of waves from the cell body to the tip of the flagellum. In Fig. 3. Micrographs of some contracted or relaxed states and their corresponding schematic representations. Upper row: only a portion of some specimens with exceptionally long flagella, we the flagellum (arrow head at base) is contracted ( positively birefringent observed up to 16 half-wavelength curvatures. in its relaxed part and also in its contracted portion); lower row: the When the demembranated flagella were submitted to flagellum is folded at around 90( angle, half negatively, half positively mild proteolysis (see Section 2) and addition of ATP, birefringent, the contracted portion remaining always positive. (Bar sliding of microtubules was observed in many (O40%) size as in Fig. 2). flagella and lead to axonemal desintegration. The resid- ual structure (R-fibre) devoid of microtubules, was still present and exhibited a birefringence signal, but we with the same value of retardation. If the compensator could observe contraction response only in few cases axis was set parallel with the flagellum axis, the bi- C (5%) when changing the Ca2 concentration of the solu- refringence contrast was white when the flagellum tion (at least 103 M Ca added) as previously observed by undulated slowly, when it was highly contracted, or when Maruyama (1985b) on non-detergent permeabilized it was contracting or relaxing. Conversely, when the models of Ceratium flagella. compensator axis was set perpendicular to the flagellum axis, the birefringence contrast became dark except when the flagellum was in the contracted state (Fig. 2). 3.4. Curvature analysis We examined some particular cases: (1) some seg- ments of the flagellum remain contracted while others Details of wave characteristics were analysed accord- appear stretched (Fig. 3); (2) the flagellum is bent so that ing to Brokaw (1990), using the software described in the distal part is perpendicular to the proximal part. In this publication and traces of flagella from video images these cases, the birefringence was observed as positive or similar to those presented in Fig. 1. Local bending negative in the first part and negative or positive in the (shear angle) was plotted against distance from the cell second part, respectively. The magnitude of the bi- body (Fig. 4F). Three different situations are presented: refringence varied from 1.3 nm to 1.9 nm, the average (1) normal sine waves, (2) contraction step, and (3) value being 1.6 nm. unfolding. The sine wave pattern (1) is very similar to 392 H. Sato et al. / Cell Biology International 28 (2004) 387e396

Fig. 4. Dark field optical microscopy of beating of C. furca longitudinal flagellum, permeabilized and ATP-reactivated, and corresponding shear angle measurements. (A) Cell body brightly reflecting light; its flagellum (at bottom) partly contracted (arrow head). (B) Higher magnification (single flash stroboscopy) of the permeabilized longitudinal flagellum (cell body on top left) with brighter portions corresponding to membrane remnants. (C, D & E): In each panel, three successive images (every 20 ms) of the same permeabilized flagellum (multistrobe illumination at 150 Hz flash frequency) showing wave propagation from 1 to 9. Scale bar Z 100 mm in panel A and 50 mm in panel B (same in C, D and E). Bottom panels (F and G) show, respectively, a plot of the shear angle vs length along the flagellum (F) and the midline of the flagellum extracted from video images of three situations of the same flagellum (G): in 1 Z contraction, in 2 Z sine waves and in 3 Z relaxation. that classically observed in sea urchin sperm flagella and More extreme shear is observed when contraction be- was documented in detail by Brokaw and Wright (1963). comes maximal, as in Fig. 1c. Such unusual shear is We focused more on the contraction step (2), where the attributed to the contraction of the R-fibre that parallels local shear angle reaches much higher values than in (1). the axoneme. The bent portions alternate with short but H. Sato et al. / Cell Biology International 28 (2004) 387e396 393 almost straight segments devoid of shearing. The unfold- elementary microstrands. Its structure and assembly ing process (3) leads to lower local bending, intercalated vary according to the state of contraction and relaxation with less straight segments of greater length. This of the nanofilaments, this variation dependingdamong process may last up to 3e4 s, while contraction is very other parametersdon the fixative composition, which brief (10 ms). During unfolding sine wave pattern acts as a triggering agent for contraction. When exposed progressively reinitiates. to increased Ca2C concentration, or when the fixative contains Ca2C at the concentration of seawater, the flagellum contracts and the filaments of the R-fibre no 3.5. Ultrastructure (Fig. 5) longer show a periodic structure, probably because they are highly coiled (Fig. 5c). They form large tubes (about As previously observed (Maruyama, 1982; Cachon 12 nm in diameter). In the presence of Ca2C-free et al., 1991), the longitudinal flagellum of Ceratium seawater containing fixative, the flagellum is fixed in contains, besides the axoneme, three structures in which a more relaxed state and the bundle of nanofilaments nanofilaments (2e4 nm) are the main components: the shows no striation because they are stretched. In con- paraflagellar rod, which is present in all dinoflagellates trast, when the flagellum is contracted or relaxed (when (Cosson et al. , 1988; Cachon et al., 1988); a thin striated the coils are being formed or on the point of disappear- fibre; and a large bundle called the R-fibre, which is only ing) (Fig. 5a), the R-fibre shows a characteristic periodic found in the longitudinal flagellum of Ceratium. Rapid structure. retraction of the flagellum is mainly due to the large R-fibre. By analysing high resolution electron micro- 4. Interpretation and discussion scopic images, we have demonstrated (Cachon et al., 1992) that this fibre is made of filaments twisted into The positively birefringent structure observed in flagella could be due either to the axoneme itself (the microtubular assembly) or to the nanofilamentous struc- ture, the R-fibre. The oriented nanofilaments in the R- fibre are probably mainly responsible for the observed birefringence, because as shown by Inoue (1986) for other biological structures, incubation in solutions with various refractive indices clearly indicates alteration of the polarization retardation (results not shown). Sup- port for this assertion is provided by data from the cortical skeletons of Noctilucidae such as Leptodiscus medusoides and Leptophyllus dasypus (Cachon and Cachon, 1964). This cortical skeleton constitutes irreg- ular meshes made from parallel layers of nanofilaments. In this case, no microtubule layer is present; however, polarizing microscope observations indicate that the cortical layer is positively birefringent. Therefore, the nanofilaments must be responsible for this birefringence. There is a growing number of examples of the de- scription of arrays of filaments (so-called nanofilaments meaning neither actin nor tubulin polymers) organised in Ca2C contractile organelles and present in many cell types (Cachon and Cachon, 1981; Salisbury, 1989; Salisbury and Floyd, 1978). In the present study on Ceratium, when the sample axis and the compensator axis of the polarizing micro- scope are parallel, the two retardations summate and the intensity of light reaching the analyser is increased. On the other hand, if the axis of the sample and that of the Fig. 5. Electron micrographs of the R-fibre of C. furca partly relaxed compensator are perpendicular, the two effects oppose with a periodic structure in (a). The nanofilaments are coiled but still or overcompensate each other. The background is extin- parallel to each other in (b). They are highly contracted without any guished and the sample appears dark (Fig. 2; diagrams). striation in (c). The nanofilaments are highly coiled in (d). Rf: R-fibre; Ax: axoneme; P: paraflagellar rod; m = membrane. Scale bar (in b) Therefore, in order to exhibit birefringence, the long corresponds to 1 mm for panels a and b; in d, the bar corresponds to axes of the constitutive elements of a filamentous system 0.5 mm for panels c and d. must parallel the axis of the whole structure (Inoue, 394 H. Sato et al. / Cell Biology International 28 (2004) 387e396

1986; Taylor, 1976). In other words, these elements have such that we observe brightness in the additive direc- to be parallel to each other and also to the axis of the tion and darkness in the subtractive direction. The structure itself. bundle is made of nanofilaments lying parallel to the A major piece of evidence identifying the structure flagellum axis. All elements hold about the same order responsible for the observed birefringence comes from of magnitude of fine structural orientation with respect demembranated Ceratium flagella submitted to micro- to the plane of the polarized light and the bire- tubule sliding. After microtubule extrusion, the residual fringence appears homogeneous. No periodicity is structure, resistant to mild proteolysis and mainly com- observed in these conditions. posed of the nanofilament bundles, still presents a bi- refringence signal similar to that observed in vivo. 4.2. Contracted flagellum (Figs. 2c and 6c) However, we could trigger contraction in these condi- tions in only a limited number of examples, so we cannot In this case the bundle is also highly and positively draw any clear-cut conclusions about the relationship birefringent whatever the position of the compensator between contraction and birefringence signal intensity. axis. Two explanations may be considered. First: the nanofilaments are highly coiled (minor coiling) and appear thicker but remain parallel to each other. 4.1. Relaxed flagellum (Figs. 2a and 6a) Though the whole bundle forms a spiral when con- tracted it is still positively birefringent because each When the flagellum is relaxed, the bundle is positively nanofilament, like the whole bundle, looks like a chro- birefringent with a high retardation value; the contrast is mosome with secondary coiling (major coiling). Second: the trajectory of the R-fibre has to be interpreted as a succession of segments, the orientation of each segment being different from that of the plane of the polarized light. Since only the projection of each segment relative to the polarized light need to be considered, and since the folds overlap one another tightly, their projection is amplified two fold or so. The overall birefringence therefore appears highly positive. This is confirmed when the compensator axis is perpendicular to the R- fibre, because the projections of the folds are also important. No periodicity is then observed (Fig. 5c).

4.3. Intermediate states (Figs. 2b and 6b)

In intermediate states, the velocity of contraction or relaxation is much too rapid to allow either observation or video recording. In this state, the periodicity observed ultrastructurally is not conspicuous under light micros- copy because the distance between two consecutive dark bands is always too small (30e60 nm). Nevertheless the contribution of projection of the bundle is important when the two axes are parallel to each other, so that the birefringence is positive (see Fig. 6b corresponding to the contraction while in progress or Fig. 6d correspond- Fig. 6. Theoretical diagrams explaining the birefringence of the R-fibre ing to ongoing relaxation). In conclusion, the filaments which is considered devoid of flanking axoneme in these drawings. (a) appear always positively birefringent whatever the orien- Relaxed state: the filaments are parallel to each other and to the tation of the compensator axis A relative to the R-fibre. orientation of the compensator axis A as well; therefore the birefrin- As described in Fig. 6c, the projections of the filaments gence is positive. In contrast, when the filaments are perpendicular to the compensator axis, the birefringence appears as positive. (b) In the of the R-fibre on the A axis are giving rise to bi- process of contracting, the waves are loose. The projections of the refringence coming in addition to the basic birefringence R-fibre on A axis give rise to a positive birefringence because these signal and finally leading to double the intensity. projections are annihilating each other. (c) Contracted state. The fila- The following comparative discussion between two ments are parallel to the compensator axis A and their projections on contractile organelle present in evolutionary distant this axis are cumulated (s segments) with basic positive birefringence observed in a and b. The resulting birefringence is then brightly species is to evaluate their degree of homology. Because positive. (d) In upper rowdthe process of relaxing leading to the same of the numerous structural and functional character- results as in b. istics shared by both systems, the R-fibre of the H. Sato et al. / Cell Biology International 28 (2004) 387e396 395

Table 1 Some comparative properties of Ceratium flagellum and Acantharian myoneme Contraction movnt Contracted state Relaxation movnt Relaxation state Duration (ms) Birefr. Duration (s) Birefr. Duration (s) Birefr. Duration (s) Birefr. Ceratium flagellum 15 G 1e5 C 0.2e1 G 1e10 Gsine w Acantharian myoneme 15 Too fast to observe 12e20 G or H 1e10 Gundul Values of birefringence are: C Z bright signal, G Z faint signal, H Z very faint signal. sine w Z sine waves developed by the longitudinal flagellum of Ceratium; undul Z slow movement of undulation. longitudinal flagellum of Ceratium and the acantharian in low Ca2C concentrations, but when more Ca2C ions myonemes, we are led to conclude to a high degree of enter the flagellum through the Ca2C channels located similarity. The myonemes of Acantharians are long, on the flagellar membrane, the Ca2C concentration rises ribbon-shaped organelles anchored on one side to the inside the flagellum and the flagellum contracts. Our cell membrane (cortex) and on the other to the apex of results obtained by studies with the polarizing micro- the spicule. These organelles are capable of three types scope strengthen Maruyama’s suggestion. of movements: rapid contraction (10e15 ms), followed by a relaxation (10e12 s), and a slow undulating move- ment. In this latter state, strong negative birefringence is Acknowledgements detected; but the birefringence is positive when the myoneme is stretched between the two anchorage sites This work was supported by the CNRS in France, by (contracted). A periodic structure appears with alter- the JSPS, and by grants in aid of Scientific Research nating dark and bright bands when the undulating provided by the Ministry of Education, Science and movement of myoneme propagates, which means it is Culture of Japan. We warmly thank the staff of going to relax (Febvre et al. 1990). Table 1 summarizes STARESO (Corsica) and all the members of the the similarities between the R-fibre of Ceratium and the Sugashima Marine Biological Laboratory, and particu- Acantharian axopodia, in two cases belonging to larly Prof. Hiroshi Hayashi, for the kind assistance evolutionary very distant species (Baldauf, 2003; provided during our stay. Lecointre and Le Guyader, 2001), Ceratium species belonging to the Dinophytes group ( protophytes) and acantharians belonging to the Actinopodes group References ( protozoa) (Lecointre and Le Guyader, 2001). These organelles are positively birefringent in both Amos WK. Reversible mechanochemical cycle in the contraction of Vorticella. Nature 1971;229:127e8. species during the relaxation state, indicating a non- Baldauf SL. The deep roots of eukaryotes. Science 2003;300:1703e6. periodic structure. There are also striking differences: Brokaw JC, Wright L. Bending waves of the posterior flagellum of while the contracted state gives positive birefringence Ceratium. Science 1963;142:1169e70. in the R-fibre, it gives negative birefringence in the Brokaw JC. Computerized analysis of flagellar motility by digitization myoneme. The latter structure is located between two and fitting of film images with straight segments of equal length. Cell Mot Cytoskel 1990;17:309e16. quite distant anchorage sites, and it shows some per- Cachon J, Cachon M. Leptospathium navicula nov. gen. nov. sp. et iodicity when contracted, while the R-fibre does not. Leptophyllus dasypus nov. gen. nov. sp, Pe´ridiniens Noctilucidae This observation could mean that the myoneme is not Kent du plancton ne´ritique de Villefranche-sur-mer. Bull Inst fully contracted. Therefore, the birefringence varies Oce´anogr Monaco 1964;62(1292):1e12. through the contractionerelaxation cycle: it is positive Cachon J, Cachon M. Movement by non-actin filament mechanisms. Biosystems 1981;14:313e26. when the elements are arrayed in bundles, but negative Cachon J, Cachon M, Boillot A. Flagellar rootlets as myonemal when they are disorganized. There is also an analogy elements for pusule contractility in Dinoflagellates. Cell Mot 1983; between the birefringence of the Ceratium flagellum and 3:61e77. the Acantharian myoneme in respect of their responses Cachon M, Cosson J, Cosson M-P, Huitorel P, Cachon J. Ultra- to Ca2C. In both cases, coiling and uncoiling of the structure of the flagellar apparatus of Oxyrrhis marina. Biol Cell 2C 1988;63:159e68. nanofilaments appear to be controlled by cellular Ca Cachon M, Cachon J, Cosson J, Greuet C, Huitorel P. Dinoflagellates influx and efflux. In Ceratium this seems to be the case in adopt various conformations in response to different needs. Biol native cells as well as in demembranated flagella; our in Cell 1991;71:175e82. vitro observations confirm that at high Ca2C concen- Cachon M, Greuet C, Cosson J, Huitorel P. Analysis of the mechanism tration, the flagellum is highly coiled due to the con- of Dinoflagellate flagella contractionerelaxation cycle. Biol Cell 1992;76:33e42. traction of the R-fibre, which in this case prevents any Cachon J, Sato H, Cachon M, Sato Y. Analysis by polarizing axonemal activity. Moreover, Maruyama (1985a) microscopy of chromosomal structure among dinoflagellates and showed that the flagellum described axonemal waves its phylogenetic involvement. Biol Cell 1989;65:51e60. 396 H. Sato et al. / Cell Biology International 28 (2004) 387e396

Cosson J, Cachon M, Cachon J, Cosson M-P. Swimming behaviour of Maruyama T. Ionic control of the longitudinal flagellum in Ceratium the unicellular biflagellate Oxyrrhis marina: in vivo and in vitro tripos (Dinoflagellida). J Protozool 1985a;32:106e10. movement of the two flagella. Biol Cell 1988;63:117e26. Maruyama T. Extraction model of the longitudinal flagellum of Febvre J. Le myone` me d’Acanthaire: essai d’interpre´tation ultra- Ceratium tripos (Dinoflagellida): reactivation of flagellar retraction. structurale et cine´tique. Protistologica 1971;VII:379e91. J Cell Sci 1985b;75:313e28. Febvre J, Febvre C, Sato H. Polarizing microscope study of a Lecointre G, Le Guyader H. Classification phylogenetique du vivant. contractile nanofilament system: the Acantharian myoneme. Biol Paris: Belin Ed; 2001. Cell 1990;69:41e51. Salisbury JL. Centrin and the algal flagellar apparatus. J Phycol 1989; Greuet C. Organisation ultrastructurale du tentacule d’ Erythropsis 25:201e6. pavillardi Kofoı¨d et Swezy, Pe´ridinien Warnowiidae Lindemann. Salisbury JL, Floyd GL. Calcium induced contraction of the rhizoplast Protistologica 1967;III(3):335e45. of a quadriflagellate green alga. Science 1978;2021:975e7. Greuet C. An unusual system of motility in a Dinoflagellate Sato H, Ellis EW, Inoue´S. Microtubule origin of mitotic spindle form Erythropsidinium pavillardi. Microtubules in microorganisms. birefringence: demonstration of the applicability of Wiener’s Cappucinelli: EMBO Ed; 1981. p. 21. equation. J Cell Biol 1975;67:501e17. Greuet C, Gayol P, Salvano P, Laval-Peuto M. Preliminary report on Sato H, Takenaka H, Niboski T. An attempt to construct super- the ultrastructural organisation of the contractile appendix of sensitive polarizing microscope with video-enhanced system. Acta Tontonia appendiculariformis (Ciliophora, Oligotrichina). Cell Mot Histochem Cytochem 1991;24(3):343e7. Cytoskel 1986;6:217e24. Soyer M-O. Pre´sence de formations fibrillaires complexes chez Hartshorne NH, Stuart A. Crystals and the polarizing microscope. Noctiluca miliaris Suriray et discussion de leur roˆ le dans la motilite´ A handbook for chemists and others. 3rd ed. London: Edward de ce Dinoflagelle´. CR Acad Sci Paris 1968;266:2428e30. Arnold Publishers Ltd.; 1960. p. 76e80. Soyer M-O. Les ultrastructures lie´es aux fonctions de relation chez Inoue S. An introduction to biological polarization microscopy. Noctiluca miliaris Suriray (Dinoflagellata). Zellforschung 1970;104: Video microscopy. New York, London: Plenum Press; 1986. 29e55. p. 477e510. Spurr AR. A low viscosity resin embedding medium for electron Maruyama T. Motion of the longitudinal flagellum in Ceratium tripos microscopy. J Ultrastructure Res 1969;26:31e43. (Dinoflagellida): a retractile flagellar motion. J Protozool 1981; Swann MM, Mitchison JM. Refinments in polarized light microscopy. 28(3):328e36. J Exp Biol 1950;27:226e37. Maruyama T. Fine structure of the longitudinal flagellum in Ceratium Taylor DL. Quantitative studies on the polarization optical properties tripos, a marine dinoflagellate. J Cell Sci 1982;58:109e23. of striated muscle. J Cell Biol 1976;68:497e511.