Regulation of Ciliary Pattern in Dileptus (Ciliata). I. Sensory Cilia and Their Conversion Into Locomotor Cilia

/. Embryol exp. Morph. Vol. 68, pp. 99-114, 1982 99 Printed in Great Britain © Company of Biologists Limited 1982

Regulation of ciliary pattern in Dileptus (Ciliata). I. Sensory cilia and their conversion into locomotor cilia

By KRYSTYNA GOLINSKA1 From the Nencki Institute of Experimental Biology, Department of Cell Biology, Warsaw

SUMMARY The surface of Dileptus contains three different regions: locomotor, oral and sensory. Each region has cilia with a specific structure and arranged in a characteristic pattern. In the morphogenetic situation when a sensory region transforms into a locomotor one, sensory cilia undergo structural changes converting them into locomotor cilia. The evidence for this is that cilia are found in the transforming region with an inner microtubular pattern intermediate between that of sensory and locomotor cilia. There are also changes in distribution of sensory units leading to a pattern characteristic of locomotor cilia. The conversion of sensory cilia into locomotor ones is also confirmed by a complete lack of evidence for resorption of sensory units within the transforming region, although the resorption is usually very easily observed with the transmission electron microscope. Transformation lasts about 5 h after the operation; afterwards locomotor cilia of normal appearance occupy the transformed region. This way of regulation of ciliary pattern has not been previously described. Its most surprising feature is the regulation of inner structure in an already differentiated ciliary unit. Some aspects of mechanisms which could control this kind of pattern regulation,! are discussed.

INTRODUCTION The surface of a ciliate cell is usually subdivided into several cortical territories where ciliary structures have an organization characteristic of the region. During post-traumatic regeneration the lost regions form anew, while the intact regions (in morphallactically regenerating ciliates) diminish in size, and the resulting organism is a perfectly proportioned miniature (Jerka-Dziadosz, 1976; Golinska & Kink, 1977; Golinska, 1979; Bakowska & Jerka-Dziadosz, 1980). Ciliary pattern on regenerating fragments is adjusted through precisely localized proliferation and resorption of ciliary units (Jerka-Dziadosz & Golinska 1977). The ciliature of Dileptus, according to morphological criteria, falls into three categories: oral, locomotor, and sensory cilia. Cilia of each kind occupy a separate region on the cell, and are organized in a specific pattern. During 1 Author's address: Nencki Institute of Experimental Biology, Department of Cell Biology, Warsaw 02-093, Poland. 100 K. GOLINSKA post-traumatic regeneration ciliature of each kind seems to regulate in its specific way. The oral ciliature in Dileptus is regulated through proliferations and resorptions of ciliary units that occur only within special areas of proliferation and resorption localized at the both ends of an elongated oral structure (Golinska & Kink, 1976, 1977; Kink, 1976). Locomotor ciliature is regulated through proliferations and resorptions which are not restricted to any localized areas, but rather reflect the changes of the shape of cell. Resorptions are found mostly on the narrowing regions of the cell, whilst proliferations are numerous on the widening or elongating parts (Golinska & Kink, 1977). In most cases, oral or locomotor cilia undergo resorption when present on those parts of cortex that are transformed into another cortical region. The pattern of sensory cilia can probably also be regulated through resorption and/or proliferation of its ciliary units. According to the present observations, however, sensory cilia when present on a territory changing into a region of locomotor cilia undergo internal restructuring leading to transformation of sensory cilia into locomotor ones. This represents the case of transdetermination occurring within organelles. Usually, in ciliates as well as in metazoans, the switch in differentiation is possible only before the structure is fully developed. A well-known example is the formation of a complete mouth out of a fragment of early oral primordia in Stentor (Tartar 1957). In the case described here a differentiated organellum (sensory cilium) undergoes an additional formation of some structural elements and dedifferentiation of the others, becoming the unit of another ciliary pattern. Some implications of this mode of pattern regulation upon the nature of control mechanisms are discussed.

MATERIALS AND METHODS Cultures of Dileptus anser were maintained as described elsewhere (Golinska & Jerka-Dziadosz, 1973). All observations were performed on anterior cellular processes, the so-called probosces. The probosces were isolated from the trunks of cells by transections made by hand, using a microknife. Level of transection was slightly above the base of proboscis, because in such fragments rows of sensory cilia reach the posterior pole of the cell, and this was essential for interpretation of some of the results. To obtain a quantity of viable fragments, some portion of endoplasm from trunk was pushed into probosces before transection. This method was successfully utilized in previous study (Golinska, 1979). For observations, samples of 20-50 fragments were prepared, and then fixed at definite times after the end of the operations. Light-microscopic observations were done on the material stained with protargol, using slightly modified method of Dragesco (1962). Cells were fixed with freshly prepared mixture of 1 part OsO4 (4 %), 1 part of glutaraldehyde (6 %), 1 part of cacodylate buffer 0-1 M, pH 6-5 (HC1), during 15 min. Fixed cells Sensory cilia and their conversion into locomotor cilia 101 were then transferred onto microscope slides and covered with albumin. The preparations were hardened in 1:1 mixture of HC1 20 % and ethanol 95 °C, for about 1 h, then washed in 95 °C ethanol and allowed to dry. Further procedures, from potassium permanganate treatment onwards, followed the procedure of Dragesco (1962). Electron-microscope observations were performed using standard preparative methods, except for fixation. The fixative used was a mixture similar to that described for light-microscope preparations, but cacodylate buffer (0-1 M) was pH 7-2, and time of fixation 30 min. Thin sections were then stained with uranyl acetate followed by lead citrate. The sections were examined in a JEM 100B transmission electron microscope.

RESULTS Ciliary pattern of Dileptus The cell of Dileptus is elongated, at the anterior end of its trunk there are cytostomal structures and a slender process the so-called proboscis (Fig. 1). The body is covered with cilia of three kinds: oral, locomotor and sensory ones. The oral ciliature of Dileptus is situated around the cytostome and on the ventral side of proboscis. Locomotor cilia occupy the whole surface of the trunk and the left and right sides of the basal portion of probsocis. Sensory cilia are localized along the whole length of the dorsal side of proboscis (Fig, 1). Studies on the fine structure of oral ciliature revealed its complexity (Grain & Golinska, 1969). The most characteristic element of oral ciliature of Dileptus is a non-ciliated kinetosome with nematodesma (Fig. 2) - a kind of ciliary rootlet which accompany only the oral ciliature. Locomotor cilia have ciliary shafts of typical structure. Accessory fibres at the proximal level of kinetosome, namely transverse, postciliary, and kinetodesmal fibre, are also similar to those found in other ciliates (Golinska & Kink, 1976). The root fibre is of a kind found in dileptuses only: it is built up of several, usually five, microtubules directed toward the anterior pole of the cell (Fig. 3). The microtubules form a loose bundle, and near to kinetosomes; the short and striated microfilamentous component of this fibre can also be seen. Locomotor cilia are arranged in a similar pattern on the trunk and on the basal portion of proboscis. There are longitudinal rows of single ciliary Units (Fig. 5) uniformly scattered within a row. The number of rows and the number of ciliary units within a row vary considerably. Sensory cilia of Dileptus are situated in prolongation to three, four or five rows of dorsal locomotor cilia. The sensory cilia are grouped in pairs and arranged in longitudinal rows (Fig. 4). The distance in between the sensory pairs lengthens toward the top of proboscis. The sensory cilium differs from locomotor cilium in the fine structure of ciliary shaft, and of the root fibre. Kinetosomes of both organelles are Very 102 K. GOLINSKA

• Oral cilia

Proboscis \

- Sensory cilia

Level of Cytostome transection "

-Locomotor cilia Trunk <

Fig. 1. Side view of ciliary regions in Dileptus.

Fig. 2. Non-ciliated oral kinetosomes bearing nematodesmata («). x 33000. Bar = 0-5 /im. Fig. 3. Kinetosome and rootlet of locomotor cilium. (m.t.), microtubular, and (m./.) microfilamentous component of the root fibre (often of a striated appearance), x 50500. Bar = 0-5 fim. Fig. 4. Pattern of sensory cilia. There are three rows of double-kinetosomal sensory units on the dorsal side of proboscis (left side of the photograph). On the right side of the photograph infraciliary oral structures are visible. Protargol preparation, x 4500. Bar = 5 /an. Fig. 5. Pattern of locomotor cilia. Kinetosomes indicated by arrows, black spheres are nuclei. Protargol preparation, x 5500. Bar = 5 /im. Fig. 6. Sensory (s.c.) and oral (o.c.) cilia on the distal part of proboscis. Differences in length between the cilia of two kinds are clearly visible. Protargol preparation, x 4500. Bar = 5 /an. Sensory cilia and their conversion into locomotor cilia 103 104 K. GOLINSKA

Fig. 7. Sensory cilium. 8, 9, 10: levels corresponding to the sections presented on Figs. 8, 9,10. (v) smooth vesicles which accompany microtubules (m.t.) of the root fibre, x 34000. Bar = 0-5 /im. Fig. 8. Transverse section of sensory shaft. Single outer microtubules (arrows) are more loosely packed than in the basal part of cilium (comp. to Fig. 9), and than in locomotor shaft, x 48000. Bar = 0-5 /tin. Fig. 9. Transverse section of sensory cilium through its basal part. Outer microtubular doublets, axial granule (a.g.), and intertubular links do not differ from those in locomotor cilium. x 52000. Bar = 0-5 /im. Fig. 10. Transverse section of kinetosomal pair of sensory unit. Anterior kinetosome (A) is deprived of accessory fibres, posterior kinetosome (P) has postciliary {p.c.\ transverse it), and kinetodesmal (k) fibres. Arrow indicates section of root fibre belonging to the anterior kinetosome of neighbouring sensory unit, x 49500. Bar = 0-5 fim. Sensory cilia and their conversion into locomotor cilia 105 much alike. The ciliary shaft of a sensory cilium is short (less than a half of the length of a locomotor one) and thick (Fig. 6). Axonemal microtubules are uniformly dispersed within the swollen cilium, and presumably are not inter- connected (Fig. 7). On longitudinal sections the tubules are often of wavy appearance. Outer doublets of the sensory axoneme are doublets only in the very short proximal part (Fig. 9); above the axial granule the subfibre B is no longer accompanying the subfibre A, and only microtubules A are present (Fig. 8). Central pair of sensory shaft is of normal appearance, but does not reach the tip of cilium. On 100 randomly chosen transverse sections of sensory cilia the pattern: nine single microtubules + central pair, was represented on 61 sections, nine single microtubules were present on 19 sections. On 13 sections nine single tubules and dense fibrous cluster (interpreted as the rest of central pair) were observed. Of the remaining sections, 5 showed 11 single microtubules, and this probably represents splitting of the tubules of central pair. Less than nine single tubules were found on 2 sections only, and this indicates that outer A tubules most frequently extend the full length of sensory cilia. The space within the ciliary membrane contains, in addition to microtubules, a very delicate filamentous network, especially dense on the circumference of microtubules (Fig. 8). This is not a special feature of sensory cilia, because the network can be found also within the locomotor cilia, and it fills the space between cytoplasitiic organelles; however, due to the presence of this electron-dense material clinging to microtubules, it is impossible to say whether the A tubules of sensory cilia are equipped with dynein arms, and whether some material resembling spokes or intertubular links is present within sensory shaft. Kinetosomes of sensory organelles have accessory fibres similar to those found in vicinity of locomotor kinetosome. There are transverse, post-ciliary, and kinetodesmal fibres (Fig. 10). The fibres are very much alike in both organelles, but in the case of sensory structures the fibres accompany only the posterior kinetosome out of each pair (Fig. 10), while every locomotor kinetosome possesses the whole set of fibres. The root fibres of locomotor and sensory organelles differ in their composition and arrangement. Root fibres of sensory organelles are identical for both kinetosomes of the pair, and are built up of 8-12 radially dispersing microtubules which enter deeply into endoplasm (Figs. 7, 10). Along the microtubules there are numerous smooth vesicles, spherical or elongated, 43-75 nm in diameter. Most of the vesicles contain some homogenous matter, darker than the surrounding cytoplasm. The root fibres are not preferentially directed toward the anterior or posterior pole of the cell. Several features of sensory cilia resemble growing locomotor cilium» A growing cilium, when it is still very short sometimes looks swollen, and within its shaft single A tubules can be observed (Fig. 12). However, both structures can easily be distinguished, because in the growing cilium all single A microtubules can never be observed in the same transverse section; there are either single and double outer tubules, or a very few single ones. 106 K. GOLINSKA

14 Sensory cilia and their conversion into locomotor cilia 107

Regulation of ciliary pattern on isolated probosees The isolated proboscis contains cilia of all kinds: locomotor, oral and sensory. During regeneration the posterior portion of isolated probsocis is forming the trunk, i.e. the region having cilia of three categories is transforming into region equipped with locomotor cilia only. Oral ciliature on the forming trunk undergoes resorption (Golinska, 1978). Locomotor cilia proliferate. This was observed during this study on numerous electron micrographs and on protargol-stained preparations (Fig. 15). Sensory cilia, according to data presented below, are restructured into locomotor cilia. The supposition that sensory cilia are in situ restructured into locomotor cilia is based upon two premises: First, there is complete lack of evidence for resorption of sensory cilia. Secondly, within the transformingregion there are found numerous peculiar cilia, having internal organization intermediate between that of sensory and of locomotor cilia. Although the lack of resorption might seem to be rather speculative evidence, I think that in this case it can be taken into consideration. Resorbed structures are readily recognized in sections examined by transmission electron mjcro- scopy, in fact, even easier than proliferating structures. This is because of the considerable length of the ciliary shaft, which in Dileptus was observed to be withdrawn under the cell surface in cases when the resorption accompany some morphogenetic process (Kink, 1978; Fig. 11 in this paper). Even if a kinetosome remains in the place, the ciliary shaft loses its filament-mediated contact with ciliary membrane in the necklace region, and in transverse sections microtubules arranged into ciliary pattern can easily be found under the cell surface. This is also true for sensory cilia, i.e. during conjugation (Fig. 13). It has to be stressed that resorption of sensory units has never been observed in the territory transforming into a new trunk of the cell. Images obtained with the light microscope also favour the supposition that

Fig. 11. Resorption of locomotor cilium, fragment 10 to 45 min. after the operation, x 35000. Bar = 0-5/«n. Fig. 12. Growing locomotor cilia, fragment 1-2 h after the operation. Single outer A tubules indicated by arrows, x 45 000. Bar = 0-5 /*m. Fig. 13. Resorption of sensory cilia, during conjugation of Dileptus. Two withdrawn shafts are indicated by arrows, x 56500. Bar = 0-5 /tm. Fig. 14. Dorsal side of isolated proboscis 5-30 min. after the operation. Three double-kinetosomal rows of sensory units reach the posterior pole of the fragment. Heavy black lines on the right are locomotor cilia. Protargol preparation, x 6000. Bar = 5 /mi. Fig. 15. Transforming region of isolated proboscis 2 to 3 h after the operation. (t.c.) region of transforming cilia, (s.c.) unchanged sensory cilia, (I.e.) region of locomotor cilia with numerous proliferating ciliary units (arrows), (p.c.) or&l cilia. Protargol preparation, x 4000. Bar = 5/xm. 108 K. GOLINSKA

19 Sensory cilia and their conversion into locomotor cilia 109 sensory cilia are not resorbed but rather restructured into locomotor cilia. In preparations made 30-60 min after the end of operations, rows of sensory cilia reach the posterior pole of fragment (Fig. 14). Preparations made 2-3 h after the end of operations show rows of single, uniformly spaced kinetosomes in the transforming region, prolongating into rows of sensory units (Fig. 15). There would be some non-ciliated region instead of rows of single cilia, if there had been resorption. At this stage there is resorption of oral ciliature on the ventral side of transforming region and proliferation of locomotor cilia, especially intense on the left side of this region. In preparations made 1-2 h after the operation many fragments show that on the dorsal side of transforming region the cilia are shorter and thicker than locomotor ones, although single and uniformly spaced with rows. Electton microscope observations on fragments fixed 1-3 h after the operation revealed numerous images of cilia which had structures intermediate between that of sensory and of locomotor cilia. These organelles were frequently found in the vicinity of typical sensory units. The images of these peculiar cilia can be arranged into a file of organelles showing the gradual change of axonemal and root fibre ultrastructure, presumably representing the consecutive stages of conversion of sensory into locomotor cilia. All these intermediate organelles are single organelles, not grouped in pairs. It is possible that the very first step in transformation is the separation of kinetosomal pairs and spacing them uniformly within the rows. In all the cases observed, kinetosomes of intermediate organelles were equipped with all the fibrous derivatives: transverse, post-ciliary, kinetodesmal, and root fibres. This indicates that, beside spacing, another early event in transformation is the

Fig. 16. Transforming cilium, 1-5-2-5 h after the operation. Locomotor-type shaft within short and swollen cilium. Root fibre is of the locomotor type (/./\), although some other fibre (arrow) is also present, x 33 500. Bar = 0-5 /Am Fig. 17. Transverse section of transforming cilium, 2 to 3 h after the operation. Outer microtubules form a ring comparable in size to that of locomotor shaft. The first B subunit is formed (arrow). 49500. Bar = 0-5 /em. Fig. 18. Transverse section of transforming cilium 2 to 3 h after the operation. Only one outer microtubule is deprived of the B subunit (arrow), x 45000. Bar = 0-5 fim. Fig. 19. Transforming cilium, 2-3 h after the operation Rootlet is composed of fibres of two kinds: nematodesma («), and locomotor-like fibre (l.r.) x 35000. Bar = 0-5 fim. Fig. 20. Transforming cilium, 2 to 3 h after the operation. Swollen cilium containing locomotor-type axoneme is accompanied by nematodesma-like rootlet («). x 30500. Bar = 0-5 fim. Fig. 21. Transforming cilium 1-5-2-5 h after the operation, having an locomotor like shaft, and nematodesma-like root fibre («), surrounded by smooth vesicles (v). x 33 500. Bar = 0 110 K. GOLINSKA formation of transverse, postciliary, and kinetodesmal fibre for the anterior kinetosome of each disjuncted sensory pair. The changes within the ciliary shaft begin with the formation of a locomotor axoneme inside of the short and swollen sensory cilium (Fig. 16). Distances in between microtubules are usually much larger in sensory than in locomotor axoneme; in the intermediate organelles the distances between axonemal microtubules are almost the same as in the locomotor axoneme. Within the thick cilium a relatively small circle of single A microtubules is visible around the central pair (Fig. 17). The next step is the formation of a B subfibre of outer doublets. In transverse sections the elements containing a B subfibre can be found in different numbers of outer tubules, from 1 to 9 per cilium (Figs. 17, 18). The number of doublets can be different in consecutive cilia within the same row. In sections through axonemes containing almost all the outer doublets the dynein arms are also visible (Fig. 18), and presumably all the other intertubular links are also forming. This stage results in 'locomotor' axoneme within a 'sensory' cilium (Figs. 16, 20). The last stage of formation of locomotor axoneme has to be an elongation of the shaft and narrowing of the whole cilium (Fig. 21), although this is very difficult to detect with the transmission electron microscope. Substantial restructuring also occurs within the root fibres of transforming organelles. The first observed change is the disappearance of smooth vesicles, normally situated in between the microtubules of sensory rootlets. The lack of vesicles probably causes the microtubules to take a position parallel to each other, forming a nematodesma-like root (Figs. 19-21), which in Dileptus usually can be found exclusively with non-ciliated oral kinetosomes (Fig. 2). Some of the smooth vesicles persist at the circumference of the nematodesma- like rootlet. The next stage is perhaps represented by the organelles equipped with both nematodesmata and obliquely oriented root fibre typical for locomotor cilium (Fig. 19). Another possibility is that the oblique fibre is not formed anew, but simply constructed out of a portion of sensory root fibre during formation of the nematodesma. The organelles possessing root fibres of both kinds usually have a locomotor axoneme within the swollen cilium (Fig. 20). Sometimes, however, the cilia completely 'locomotor' in appearance are equipped with root fibres of the two kinds (Fig. 21). The nematodesma-like structure is probably resorbed in later stages of regeneration, because 5 and more hours after the operation in the region where the sensory organelles had been, only the locomotor cilia with their characteristic rootlets were observed.

DISCUSSION Non-locomotor cilia of unknown function were found in many ciliates. The short and stiff cilia localized on the anterior-dorsal side of ciliates belonging to order Haptorida (taxonomy after Corliss, 1979) were first described and named Sensory cilia and their conversion into locomotor cilia \\\ as sensory cilia by von Gelei (1934, 1950, 1954). The supposition concerning the sensory role of these organelles has been neither confirmed nor excludedf by other investigators. At present these organelles are often referred to as 'clavate cilia' (Wessenberg & Antipa, 1968; Holt, Lynn & Corliss, 1973; RodrigueS de Santa Rosa & Didier, 1975), or 'une brosse' by some French authors (Bohatier, .Iftode, Didier & Fryd-Versavel, 1978; Fryd-Versavel, Iftode & Dragesco, 1975). I prefer the term 'sensory cilia' because it implies that an axonemal pattern of ciliary microtubules is modified compared with the pattern of locomotor cilium, as was frequently reported for cilia of chemo- and mechanoreceptors of higher organisms (e.g. Thornhill, 1967, reviewed by Gaffal & Bassemir, 1974). The images obtained in this study strongly suggest that the sensory cilia can be remodelled into locomotor ones when the territory they occupy undergoes conversion into a trunk of the cell. To my knowledge this represents the first description of differentiated organelles undergoing transdetermination. This, however, may not be an exceptional case among ciliates. For instance, Nelsen & Frankel (1979) in a study on a new kinety insertion in Tetrahymena described how kinetosomal couplets of oral origin form additional rows of locomotor cilia. The couplets leave the immature oral apparatus and probably are not fully differentiated; nonetheless they continue being couplets during a considerable period of time, and this indicates that there is at least some interkinetosomal linkage present in oral couplets before their transformation into single locomotor cilia. The formation of locomotor cilia would require also formation of new accessory fibre (i.e. kinetodesmal fibre) and destruction of others (i.e. linkage fibre), thus representing transdetermination of ciliary units. Evidence of another possible case of transdetermination is the fact that locomotor cilia of T6tra- hymena can be incorporated in forming oral parts (Frankel, 1969). This would also require substantial restructuring of kinetosomal fibrous equipment. Unfortunately, no ultrastructural data are available on this redifferentiation and dedifferentiation of ciliary units of Tetrahymena. A common feature of these transdeterminations in Tetrahymena and Dileptus is the obvious influence of a kind of cortical territory upon the kind of cilia which happen to be there. The oral cilia which enter the locomotor territory of Tetrahymena change into locomotor ones; the locomotor cilia within territory determined to be the oral region of dividing Tetrahymena are transformed into oral cilia; the sensory cilia of Dileptus when present on the forming locomotor territory undergo transformation into locomotor cilia. This way of regulation of cortical pattern, through formation and size determination of separate cortical territories (each territory containing ciliary units of its own kind) up to now has been postulated to exist in hypotricrious ciliates only (Tuffrau, 1977). It needs, however, some more experimental data for further confirmation that the regulation of cortical pattern may concern primarily the cortical territories, which in turn influence ciliary structure. Any- way, in the case of Dileptus, the long-range control of patterning (Frankel, 1974) 112 K. GOLINSKA can influence intra-organellar organization of fibrous elements of cortical pattern. The intra-organellar control of patterning of ciliary shafts and rootlets is supposedly exerted by MTOCs - a microtubule-organizing centres (reviewed by Tucker, 1979). MTOCs of ciliary shaft can be formed or activated when some injury to the ciliary unit brings about regeneration of the shaft. The formation and/or regeneration of the ciliary shaft may represent a continuation of activity of the MTOC responsible for the patterning of kinetosome, or there is a separate MTOC for the formation of the ciliary shaft (reviewed by Raff, 1979). In the case of transformation of sensory into locomotor cilium in Dileptus, the question arises whether there are different MTOCs for sensory and locomotor shaft, i.e. whether the transformation starts with the MTOC itself. Another possibility is that the MTOC of sensory and locomotor shaft is of the same kind, and in the case of the sensory cilium there is some restriction to its potential development. The resemblance between sensory cilia and growing locomotor ones (Bohatier, 1979) implies that sensory cilia could be locomotor ones blocked in some stage of their development. The blockage of development of cilium at the stage of formation of B subunit and intertubular links, could be exerted either at the level of de novo protein synthesis, or at the level of their transport to cortical organelles. Anyway, if in the case of sensory shaft there is only some half-way inhibitory effect, then the final accomplishment of the MTOC would be the formation of locomotor cilium. This is supported by observation that the transformation of cilium proceeds without dedifferentiation, through com- pletion of the shaft by addition of new microtubular and filamentous components. Some light could be thrown upon the problem by studies on the way in which formation of sensory cilia is accomplished during another morphogenetic process, when the conversion of trunk into proboscis occurs. The question is, whether locomotor ciliary shaft can be converted into the sensory one, i.e. whether the transformation can proceed in the direction opposite to that described above. In the case of reversibility of the process, supposition about the transformation of the MTOCs themselves during regulation of ciliary pattern would be strongly supported. Patterning of the ciliary rootlet in several respects differs from that of the ciliary shaft. First of all, root fibres do not form as a direct continuation of kinetosomal microtubules, like the axonemal fibres do. Besides, the dedifferentiation was observed to occur within root structures: both smooth vesicles and nemadesma-like rods disappear during the transformation. The formation of the locomotor-type rootlet proceeds simultaneously with dedifferentiation of at least a portion of the sensory-type rootlet. It seems important that the new root microtubules are of different orientation in comparison with the previously existing root microtubules. Instead of microtubules roughly perpendicular to the surface of the cell, there are anteriorly directed microtubules almost Sensory cilia and their conversion into locomotor cilia 113 parallel to the surface. This indicates that the MTOC of root fibre undergoes substantial changes during the transformation, if it is not formed anew.

This investigation was supported by the Polish Academy of Science, research grant no. II MR, PAN. I wish to thank Dr Maria Jerka-Dziadosz and Dr Julita Bakowska for helpful comments on the manuscript, and Mrs Lidia Wiernicka for expert technical assistance*

REFERENCES BAKOWSKA, J. & JERKA-DZIADOSZ, M. (1980). Ultrastructural aspect of size dependent regulation of surface pattern of complex ciliary organelle in a protozoan ciliate. /. Embryol. exp. Morph. 59, 355-375. BOHATIER, J. (1979). Morphogenese de regeneration chez le cilie Condylostoma magnum (Spiegel): Etude ultrastructurale. /. Protozool. 26, 404-414. BOHATIER, J., IFTODE, F., DIDIER, P. & FRYD-VERSAVEL, G. (1978). Sur l'ultrastructure des genres Spathidium et Bryophyllum, Cillies Kinetophragmophora (de Puytorac et al. 1,974). Protistologica 14, 189-200. CORLISS, J. O. (1979). The Ciliated Protozoa. Pergamon Press. DRAGESCO, J. (1962). L'orientation actuelle de la systematique des Cilies et la technique d'impregnation au proteinate d'argent. Bull. Microsc. appl. 11, 49-58. FRANKEL, J. (1969). Participation of the undulating membrane in the formation of the oral replacement primordium of Tetrahymena pyriformis. J. Protozool. 16, 26-35. FRANKEL, J. (1974). Positional information in unicellular organisms. J. Theor. Biol. 47, 439-481. FRYD-VERSAVEL, G., IFTODE, F. & DRAGESCO, J. (1975). Contribution a la connaissance de quelques Cilies Gynmostomes. II. Prostomiens, Pleurostomiens, morphologie, stofnato- genese. Protistologica 11, 509-530. GAFFAL, K. P. & BASSEMIR, V. (1974). Vergleichende Untersuchung modifizierter Cilien- strukturen in den Dendriten mechano- und chemosensitiver Reseptorzellen der Baum- wollwanze Dysdercus und der Libelle Agrion. Protoplasma 82, 177-202. GELEI, J. VON (1934). Die sensorichen Organellen der Ciliaten. Allat. Kozlein. 31, 115-138. GELEI, J. VON (1950). Die Morphogenese der Einzeller mit Riicksicht auf die morphogene- tischen Prinzipen von Sewertzoff. Acta biol. Acad. sci. hung. 1, 69-134. GELEI, J. VON (1954). Ober die Lebensgemeinscheft einiger temporarer Tiimpel auf einer Bergwiese im Borzsonygebirge (Oberungarn). III. Ciliaten. Acta biol. Acad. sci. hung. 5, 259-349. GOLINSKA, K. (1978). The course of in situ remodelling of injured mouthparts in Dileptus (Ciliata, Gymnostomata). Acts Protozool. 17, 47-67. GOLINSKA, K. (1979). Assessment of cell proportions during regeneration of DileptuS anser {Ciliata). Wilhelm Roux's Arch. devl. Biol. 187, 307-321. GOLINSKA, K. & JERKA-DZIADOSZ, M. (1973). The relationship between cell size and capacity for division in Dileptus anser and Urostyla cristata. Acta Protozool. 12, 1-21. GOLINSKA, K. & KINK, J. (1976). The regrowth of oral structures in Dileptus cygnuf after partial excision. Acta Protozool. 15, 143-163. GOLINSKA, K. & KINK, J. (1977). Proportional regulation of body form and cortical organelle pattern in the ciliate Dileptus. J. Cell Sci. 24, 11-29. GRAIN, J. & GOLINSKA, K. (1969). Structure et ultrastructure de Dileptus cygnus Claparede et Lachmann, 1859, Cilie Holotriche Gymnostome. Protistologica 5, 269-291. HOLT, P. A., LYNN, D. H. & CORLISS, J. O. (1973). An ultrastructural study of the tentacle- bearing ciliate Actinobolina smalli n.sp. and its systematics and phylogenetic implications. Protistologica 9, 521-544. JERKA-DZIADOSZ, M. (1976). The proportional regulation of cortical structures in a hypo- trich ciliate Paraurostyla weissei. J. exp. Zool. 195, 1-14. JERKA-DZIADOSZ, M. & GOLINSKA, K. (1977). Regulation of ciliary pattern in ciliates. /. Protozool. 16, 612-637. 114 K. GOLINSKA KINK, J. (1976). A localized region of basal body proliferation in growing cells of Dileptus visscheri (Ciliata, Gymnostomata). J. Cell Sci. 20, 115-133. KINK, J. (1978). Dedifferentiation of fibrillar structures during encystment of Dileptus visscheri (Gymnostomatida). Acta Protozool. 17, 31-45. NELSEN, E. M. & FRANKEL, J. (1979). Regulation of corticotype through kinety insertion in Tetrahymena. J. exp. Zool. 210, 277-288. RAFF, E. C (1979). The control of microtubule assembly in vivo. Int. Rev. Cytol. 59, 1-96. RODRIGUES DE SANTA ROSA, M. & DIDIER, P. (1975). Remarques sur l'ultrastructure du Cilie Gymnostome Monodinium balbiani (Fabre Domergue, 1888). Protistologica 11, 469-479. TARTAR, V. (1957), Deletion experiments on the oral primordium of Stentor coeruleus. J. exp. Zool. 136, 53-74. THORNHILL, R. A. (1967). The ultrastructure of the olfactory epithelium of the lamprey Lampetra fluviatilis. J. Cell Sci. 2, 591-602. TUCKER, J. B. (1979). Spatial organization of microtubules. In: Microtubules (ed. K. Roberts & J. S. Hyams), Academic Press. TUFFRAU, M. (1977). Variations intraclonales du phenotype ventral chez Euplotes vannus, forme mutabilis, Cilie Hypotriche. Protistologica 13, 435-439. WESSENBERG, H. & ANTIPA, G. (1968). Studies on Didinium nasutum. I. Structure et ultrastructure. Protistologica 4, 427-448.

(Received 21 May 1981, revised 4 September 1981)