The cortex of Tetrahymena contains a regular array of longitudinal microtubular bands (lms) next to basal body rows running from pole to pole. The lm exhibits a predominant unidirectionality in assembly. The direction of regeneration following breakage of the microtubules is from posterior to anterior of the cell. When the lm and the accompanying basal body row are rotated 180 ° (inverted), so that their polarity is opposite to that of the cell, the predominant direction of regeneration exhibited by the inverted lm is from anterior to posterior. This shows that the lm has an inherent direction of regeneration independent of cellular polarity. This implies that the microtubules constituting the lm have an intrinsic property which controls the direction of assembly. This finding is in accord with the in vitro demonstration using Chlamydomonas flagellar fragments. On the basis of this finding and also the possible pattern of arrangement of the microtubules constituting the lm, it is suggested that growth of the lm involves both elongation of pre-existing microtubules constituting the lm and also laying down of new ones.

The cortex of Tetrahymena contains a regular pattern of microtubular bands that can be visualized easily by cytological staining (Figs, 1, 2, 3 (i), p. 112). This system has been exploited in studies concerning the direction of assembly of microtubules in vivo (Ng, 1978 a). When a longitudinal microtubular band (lm) undergoes breakage, the open end on the posterior portion of the band regenerates to a greater extent than that on the anterior portion; hence, the predominant direction of regeneration is from posterior to anterior of the cell. This finding, like those with in vitro systems, suggests that assembly of microtubules is predominantly unidirectional (see reviews by Borisy, Olmsted, Marcum & Allen, 1974; Rosenbaum et al. 1975).

The demonstration that microtubule assembly proceeds in a particular direction in a living cell does not, however, necessarily argue for an intrinsic property of the regenerating microtubules in controlling the direction of assembly. This is simply because cellular structures do not exist in isolation from one another. The predominant unidirectionality demonstrated in vivo could well reflect influences from other structures on the assembly process, rather than any inherent property of the microtubules themselves. Tetrahymena is a structurally polar entity; the oral apparatus (Fig. 1, p. 112) close to the anterior end defines the antero-posterior axis and other structures bear fixed positional relationships to it. The cellular polarity has obvious morphogenetic significance. For example, prior to division the fission furrow always cuts across the middle of the antero-posterior axis just anterior to the newly developed oral apparatus, giving rise to 2 daughters. The question of interest in the present study is: does cellular polarity affect the posterior-to-anterior regeneration of the lm?

Fig. 1.

Protargol staining showing asymmetric distribution of the major cortical microtubular bands (lm, tm and pm) next to basal bodies (see Materials and methods for description), × 1400.

Fig. 1.

Protargol staining showing asymmetric distribution of the major cortical microtubular bands (lm, tm and pm) next to basal bodies (see Materials and methods for description), × 1400.

To show that unidirectional assembly is an intrinsic property of the microtubule in an in vivo situation, the system employed must satisfy 3 criteria, (i) Two complementary or opposite open ends of microtubules (such as those generated by breakage) must be present allowing an assessment of the direction of regeneration, (ii) The end that favours assembly must be distinguishable (based on criteria other than assembly) from the other end disfavouring assembly. This is because unidirectional regeneration does not necessarily imply assembly from a particular end. And (iii) When the positional relationship between the regenerating microtubule and other cellular structures is altered, the microtubule maintains its original direction of regeneration, thus showing that unidirectional assembly is an intrinsic property of the microtubule.

It is clear from the previous study (Ng, 1978 a) that Tetrahymena satisfies the first 2 criteria: when a break is created on the lm the anterior and posterior portions, each bearing a complementary open end, can be identified. Furthermore, to satisfy the third criterion, advantage can be taken of the fact that ciliary meridians, including the 1ms, can be rotated 180° (Ng & Frankel, 1977). The direction of assembly in 1ms that are 180°-misaligned with respect to the cellular polarity can thus be studied. The present study shows that the 180°-rotated microtubular band maintains its original predominant direction of regeneration independent of cellular polarity. This furnishes an in vivo demonstration that microtubules have an intrinsic property governing the direction of assembly.

Stock and culture

The cells studied are of genotype molb/molb (Frankel, Jenkins, Doerder & Nelsen, 1976). They are all descendants of a single isolate bearing in total 19 ciliary rows, 2 of which (nos. 4* and 5) have been rotated 180° (inverted). Details of methods for isolating cell lines with inverted ciliary rows have been reported in Ng & Frankel, 1977.

The cells are maintained in log-phase in 1 % proteose peptone -0·1 % yeast extract at 28 °C. They form heteropolar doublets characteristic of molb/molb phenotype when grown at 39 °C (see Results).

Staining of cortical microtubular bands

Ng & Nelsen (1977) introduced an improved version of the protargol technique which stains the major cortical microtubular bands as well as basal bodies and cilia. This is employed in the present study.

Identification of the inverted ciliary rows and junctures between the inverted and normal rows

The inverted ciliary rows can be identified by the asymmetrical pattern of microtubular bands next to the basal bodies (Figs. 1, 3). Normally, the longitudinal band (lm) runs from pole to pole on. the right* of each somatic basal body row; the transverse band (tm) arises just anterior to the basal body, extends toward the left and stops short of the ciliary row on the left; the small postciliary band (pm) arises close to the right posterior face of the basal body and extends for a short distance posteriorly. When the ciliary row is inverted, these 3 microtubular bands are at 180° opposite to their normal orientations (Fig. 3(ii)).

Fig. 2.

Electron micrograph showing a transverse section across an lm. The lm is made up of a single layer of microtubules between the inner alveolar membrane (im) and epiplasm (e). × 90000.

Fig. 2.

Electron micrograph showing a transverse section across an lm. The lm is made up of a single layer of microtubules between the inner alveolar membrane (im) and epiplasm (e). × 90000.

Fig. 3.

Drawing showing the configuration of the right and left junctures created as a result of inversion. Fig. 3 (i) shows the regular arrangement of 4 ciliary rows (cf. Fig. 1). Fig. 3 (ii) shows inversion of 2 rows in the middle. For a description of the rj and lj see Materials and methods. A and P denote the anterior and posterior ends of the cell and right (R) and left (L) are defined based on footnote *, p. 111.

Fig. 3.

Drawing showing the configuration of the right and left junctures created as a result of inversion. Fig. 3 (i) shows the regular arrangement of 4 ciliary rows (cf. Fig. 1). Fig. 3 (ii) shows inversion of 2 rows in the middle. For a description of the rj and lj see Materials and methods. A and P denote the anterior and posterior ends of the cell and right (R) and left (L) are defined based on footnote *, p. 111.

Two junctures between the inverted and normal rows, one on each side of the inversion, can be identified: the juncture on the right of the inversion (right juncture) is characterized by 2 basal body rows being ‘bracketed’ by 2 1ms and also by lms interdigitating with each other. The left juncture is characterized by 2 rows of basal bodies bracketing 2 1ms and the absence of any lms.

Formation of heteropolar doublets and development of junctures

Breakage and regeneration of longitudinal microtubular bands (lm) occur at a particular site in heteropolar doublets derived from singlets. Hence, the process of doublet formation and the topography of this site are described in some detail.

When the cells are transferred from 28 to 39 °C for 5 h, many of them exhibit an abortive fission characteristic of the molb/molb phenotype (Frankel et al. 1976). The transverse fission zone fails to develop across some of the ciliary rows (generally the more dorsal ones). The prospective anterior and posterior daughters remain attached and bend over each other to form a V-shaped heteropolar doublet (Figs. 4B, 5B). As a result of bending, the anterior and posterior portions of some ciliary rows belonging to the anterior and posterior daughters, respectively, are brought close together in opposite orientation. Consequently, 2 types of junctures between ciliary rows of opposite orientation are being developed, one on each side of the doublet. On the left-lateral* surface, the juncture being developed is characteristic of the right juncture found in singlets bearing inverted ciliary rows (i.e. 2 1ms bracketing 2 basal body rows; see Materials and methods and cf. Figs. 4B, 3 B); hence it is designated as a developing right juncture, or drj. In addition, the drj in heteropolar doublets is abutted on one end (the dorsal side) by the base of a looping ciliary row. The lm associated with this ciliary row faces the drj. Previous studies (Ng, 1978a, b) show that this lm develops a break at the base facing the drj. This produces 2 complementary open ends for assessing the direction of microtubule regeneration (see next section).

Fig. 4.

Development of the right juncture (drj) on the left-lateral surface and breakage and regeneration of the lm facing the drj. Fig. 4 C-E show regeneration from the posterior (P), anterior (A) and both portions of the lm, respectively. The direction of regeneration cannot be assessed in the cases shown by Fig. 4F. Arrow I, the polarity of the cell; arrow 2, the normal orientation of the lm; arrow 3, the predominant direction of regeneration of the lm (posterior-to-anterior).

Figs. 4–6. Development of the right and left juncture in heteropolar doublets. Except in Fig. 6 A, only 2 ciliary rows are shown. The anterior and posterior portions of one, because of bending of the anterior and posterior daughters over each other, are brought close to each other in opposite orientation and thus develop a juncture. The other ciliary row forms a loop abutting the developing juncture. Basal bodies are represented by dots, next to which is the lm (line). The tms and pms are omitted from the drawings; for details of these in the developing junctures, see Fig. 3B. LLS and RLS are, respectively, left-lateral surface and right-lateral surface of the cell (see footnote *, p. 111).

Fig. 4.

Development of the right juncture (drj) on the left-lateral surface and breakage and regeneration of the lm facing the drj. Fig. 4 C-E show regeneration from the posterior (P), anterior (A) and both portions of the lm, respectively. The direction of regeneration cannot be assessed in the cases shown by Fig. 4F. Arrow I, the polarity of the cell; arrow 2, the normal orientation of the lm; arrow 3, the predominant direction of regeneration of the lm (posterior-to-anterior).

Figs. 4–6. Development of the right and left juncture in heteropolar doublets. Except in Fig. 6 A, only 2 ciliary rows are shown. The anterior and posterior portions of one, because of bending of the anterior and posterior daughters over each other, are brought close to each other in opposite orientation and thus develop a juncture. The other ciliary row forms a loop abutting the developing juncture. Basal bodies are represented by dots, next to which is the lm (line). The tms and pms are omitted from the drawings; for details of these in the developing junctures, see Fig. 3B. LLS and RLS are, respectively, left-lateral surface and right-lateral surface of the cell (see footnote *, p. 111).

The situation with the right-lateral* side of the doublet is just the reverse (Fig. 5 B). Here, the juncture being developed normally is characteristic of the left juncture in singlets bearing inverted ciliary rows (i.e. 2 1ms being bracketed by 2 basal body rows; see Materials and methods and cf. Figs. 5B and 3 B); it is thus designated as a developing left juncture, or dlj. Moreover, the lm associating with the looping ciliary row abutting the dlj does not face the dlj. For some reasons this lm is never seen to develop a break at the base of the loop and hence no microtubular regeneration is observed at this site (Ng, 1978b).

Fig. 5.

Development of the left juncture (dlj) on the right-lateral surface. The lm associating with the looping ciliary row abutting the dlj does not face the dlj. No breakage and regeneration has ever been observed.

Fig. 5.

Development of the left juncture (dlj) on the right-lateral surface. The lm associating with the looping ciliary row abutting the dlj does not face the dlj. No breakage and regeneration has ever been observed.

However, the cells employed in the present study bear 2 inverted ciliary rows (nos. 4 and 5) on the right-lateral surface (Fig. 6 A). When they form heteropolar doublets, frequently a right juncture is developed on the right-lateral surface where normally a a left juncture would have developed (cf. Figs. 5 and 6). The anterior and posterior halves of one inverted ciliary row (no. 4) together constitute the developing right juncture (drj); the other inverted ciliary row (no. 5) makes a loop abutting the drj. The lm associated with this looping inverted ciliary row, now facing the dij, develops a break and yields two open ends for assessing the directionality of microtubule assembly (see next section).

Fig. 6.

Development of a right juncture (drj) on the right-lateral surface where normally the left juncture should have developed (cf. Fig. 5). This happens as a result of the prior presence of inverted ciliary rows no. 4 and no. 5 (Fig. 6 A). Fig. 6 C, D show 2 modes of regeneration of the lm observed (see micrographs Figs. 7–9). The predominant direction of regeneration is from anterior (A) to posterior (P) of the cell (Fig. 6 C). Thus, the direction of lm regeneration (arrow 3) follows the polarity of the lm (arrow 2) which is inverted and thus opposite to the cellular polarity (arrow I).

Fig. 6.

Development of a right juncture (drj) on the right-lateral surface where normally the left juncture should have developed (cf. Fig. 5). This happens as a result of the prior presence of inverted ciliary rows no. 4 and no. 5 (Fig. 6 A). Fig. 6 C, D show 2 modes of regeneration of the lm observed (see micrographs Figs. 7–9). The predominant direction of regeneration is from anterior (A) to posterior (P) of the cell (Fig. 6 C). Thus, the direction of lm regeneration (arrow 3) follows the polarity of the lm (arrow 2) which is inverted and thus opposite to the cellular polarity (arrow I).

In summary, each of the heteropolar doublets under study contains 2 developing right junctures (drjs). The one on the left-lateral surface is developed by, and abutted on one end by normal ciliary rows (Fig. 4B). The one on the right-lateral surface is developed due to the prior presence of 2 inverted ciliary rows on the right-lateral surface; the looping ciliary row and its associating lm abutting this drj are also inverted (i.e. of opposite polarity to the cell) (Fig. 6B).

Regeneration of normal and inverted lms in heteropolar doublets

The protargol technique employed gives a good staining only of the upper surface of the cell (Ng & Nelsen, 1977); hence in all of the cases reported below, only one surface of each heteropolar doublet has been examined.

In total, 152 regenerating normal 1ms of the looping ciliary row abutting the drj on the left-lateral surface have been examined (Fig. 4B–F). Of these, 108 show extension from the posterior portion of the lm (Fig. 4c); 3 from the anterior portion (Fig. 4D) and 30 from both portions, showing an X-shaped structure (Fig. 4E). The remaining 11 cases are ambiguous (Fig. 4F), due probably to shifting of the regenerating lm to align itself more in parallel with the drj during development. Thus, the ratio of extension from posterior portion to extension from anterior portion is (108 + 30): (3 + 30), or about 4: 1. That is, about 81 % of the regeneration is from the posterior portion of the lm. This agrees well with the results obtained in the previous study (86 %; Ng, 1978a). The reader is referred to Ng, 1978a for protargol micrographs of the regenerating normal lm.

As to the regenerating inverted lm in the drj on the right-lateral surface (Figs. 6B–D, 7–9), altogether 40 such cases have been examined. Of these, 32 exhibit regeneration from the anterior portion of the inverted lm (Figs. 6 c, 7, 8) and 8 exhibit regeneration from both portions of the inverted lm (Figs. 6D, 9). The ratio of regeneration from anterior portion to that from posterior portion is thus (32 + 8): 8, or 5: 1. That is, about 83 % of the regeneration is from the anterior portion of the inverted lm.

The result clearly shows that (i) the direction of microtubule assembly on the lm is normally from posterior to anterior of the cell, and (ii) when the regenerating lm is oriented opposite to the cellular polarity the direction of assembly is from anterior to posterior of the cell. It is therefore concluded that the direction of microtubular assembly exhibited by the lm is independent of cellular polarity.

Intrinsic unidirectional assembly

The present study shows that the direction of regeneration of the longitudinal microtubular band (lm) is predominantly posterior-to-anterior; when the lm is rotated 180 °, the direction is predominantly anterior-to-posterior. Thus, the assembly of microtubules in the case of the lm proceeds mainly in a fixed direction defined by the lm and is independent of its orientation with reference to cellular polarity. In the present system, microtubule assembly on 2 complementary open ends of the lm generated as a result of breakage is compared and it is shown that it takes place predominantly on a particular end. This indicates that the molecules on the 2 complementary ends exhibit different properties regarding tubulin polymerization. This difference may reflect a conformational difference in the exposed surfaces of the 2 complementary ends. Regardless of the exact molecular mechanism involved, it is clear from the present study that the information controlling assembly in vivo is an intrinsic property of the microtubules. implications of the assembly on the disfavoured microtubular end and the presence of ‘X-shaped’ structures have been previously noted (Ng, 1978 a).

Much in vitro polymerization work has been done with fragments of neurotubules and flagellar axonemes and also with isolated flagellar axonemes and basal bodies (for reviews, Borisy et al. 1974; Rosenbaum et al. 1975). All of these studies demonstrate a predominantly unidirectional growth of microtubules. The best in vitro evidence showing that the microtubules possess intrinsic information for controlling unidirectional growth is based on tubulin assembly on fragments of flagellar axonemes from Chlamydomonas (Allen & Borisy, 1974). The axoneme fragments bear complementary open ends generated by breakage of longer pieces. Furthermore, the polar orientation of the side arms on the outer doublets allows determination of the orientation of the fragment with respect to the cell and thus makes it possible to distinguish one end of the fragment from the other. This work shows that it is the distal end of the fragment which favours microtubule assembly and hence suggests that the control of unidirectional assembly lies in the microtubules.

Work based on isolated flagellar axonemes do not yield a clear-cut conclusion concerning the notion of intrinsic control of assembly by microtubules. It is not clear that the opposite ends of the isolated axonemes are complementary. While the proximal ends of the outer doublets are produced by detachment from the basal body triplets, the nature of the distal end is not clear. Furthermore, recent studies reveal other structures at the distal tips of the flagellar axoneme (Dentier & Rosenbaum, 1977; Sale & Satir, 1977). The central tubules are ‘capped’ distally by a discrete structure and the A-tubules of the outer doublets end in ‘distal filaments’. It is also known that the central tubules end proximally on or near the basal cup (Ringo, 1967). The presence of such structures complicates the comparison of microtubule assembly on the 2 ends of the axonemes. Until the influence of such structures on microtubule assembly is known and quantitated, the unidirectional assembly observed in such systems does not necessarily demonstrate an intrinsic property of the microtubules in controlling assembly. Such considerations may also be applied to the in vivo study of Whitman (1975) on regenerating flagellar stubs after amputation of the flagella. In addition, it is also interesting to note the recent finding in yeast spindles concerning the structure of microtubular ends (Byers, Shriver & Goetsch, 1978). The proximal end on the spindle pole body is structurally different from the distal end. Microtubule assembly proceeds proximal-to-distal, but not in reverse. The unidirectionality in this case most likely reflects a modification of the proximal end by the spindle pole body material.

Propagation of the longitudinal microtubular band (lm)

During fission, the 1ms are divided up and passed to the anterior and posterior daughters. Each band obviously has to lengthen during the inter-fission period of the cell cycle. The unidirectional assembly of the lm and the pattern of arrangement of the microtubules constituting the band have interesting implications on the mode of propagation of the bands.

Each lm is made up of 7-12 microtubules arranged in a single layer (Allen, 1967) (Fig. 2). However, while the lm extends from pole to pole the individual microtubules appear not to. Electron microscopy of isolated pellicular fragments from 2 related ciliates, Glaucoma chattoni and Colpidium campylum, show that the lm is made up of ‘relatively short, overlapping fibrils (microtubules), each one with its anterior terminus on the right edge of the band and its posterior one on the left. The length of an individual fibril (microtubule) here appears to be about 4 times the meridional distance between cilia’ (Pitelka, 1961) (see Fig. 10). Thus, the long axis of each microtubule makes an acute angle with the long axis of the lm and an individual microtubule is about a third to a quarter the length of the cell. Though not yet confirmed at the ultrastructural level, it is almost certainly the case that some microtubules of the lm terminate on the anterior and posterior poles (Fig. 10); a transverse breakage of the lm at the mid-body region during cell fission would be expected to produce such anterior and posterior termini. Regarding Tetrahymena, a low-power electron micrograph of a pellicular fragment (fig. 5, Pitelka, 1961) and also a published electron micrograph of a grazing section of the surface (fig. 17, Allen, 1967) both suggest the same pattern of arrangement of the microtubules.

Fig. 7.

Protargol pictures of the right-lateral surface of heteropolar doublets corresponding to the drawings in Fig. 6. The drj is developed as a result of the prior presence of inverted ciliary rows. Figs. 7 and 8 show regeneration (arrows) from the anterior portion of the lm facing the drj. Fig. 9 shows regeneration (arrows) from both portions of the lm showing an X-shaped structure, × 1400.

Fig. 7.

Protargol pictures of the right-lateral surface of heteropolar doublets corresponding to the drawings in Fig. 6. The drj is developed as a result of the prior presence of inverted ciliary rows. Figs. 7 and 8 show regeneration (arrows) from the anterior portion of the lm facing the drj. Fig. 9 shows regeneration (arrows) from both portions of the lm showing an X-shaped structure, × 1400.

Fig. 8.

Protargol pictures of the right-lateral surface of heteropolar doublets corresponding to the drawings in Fig. 6. The drj is developed as a result of the prior presence of inverted ciliary rows. Figs. 7 and 8 show regeneration (arrows) from the anterior portion of the lm facing the drj. Fig. 9 shows regeneration (arrows) from both portions of the lm showing an X-shaped structure, × 1400.

Fig. 8.

Protargol pictures of the right-lateral surface of heteropolar doublets corresponding to the drawings in Fig. 6. The drj is developed as a result of the prior presence of inverted ciliary rows. Figs. 7 and 8 show regeneration (arrows) from the anterior portion of the lm facing the drj. Fig. 9 shows regeneration (arrows) from both portions of the lm showing an X-shaped structure, × 1400.

Fig. 9.

Protargol pictures of the right-lateral surface of heteropolar doublets corresponding to the drawings in Fig. 6. The drj is developed as a result of the prior presence of inverted ciliary rows. Figs. 7 and 8 show regeneration (arrows) from the anterior portion of the lm facing the drj. Fig. 9 shows regeneration (arrows) from both portions of the lm showing an X-shaped structure, × 1400.

Fig. 9.

Protargol pictures of the right-lateral surface of heteropolar doublets corresponding to the drawings in Fig. 6. The drj is developed as a result of the prior presence of inverted ciliary rows. Figs. 7 and 8 show regeneration (arrows) from the anterior portion of the lm facing the drj. Fig. 9 shows regeneration (arrows) from both portions of the lm showing an X-shaped structure, × 1400.

Fig. 10.

Drawing to show the arrangement of microtubules constituting the lm. A,P, R, L are, respectively, anterior, posterior, right and left. Each microtubule starts at the left edge of the lm, extends anteriorly and terminates at the right edge of the lm. This is the case for all microtubules except those cut off at the anterior and posterior poles of the cell. ably are open ends produced as a result of breakage of the lm at the mid-body region during cell fission. The cell, however, does not grow exclusively at one pole. Hence, the lm probably slides antero-posteriorly between the inner pellicular membrane and the epiplasm so as to make way for assembly at the anterior termini, (iii) The elongating microtubules, because of their angular arrangement in relation to the lm, extend until ultimately reaching or defining the right edge of the band. Clearly, then, elongation of the pre-existing microtubules is not sufficient for propagation of the lm. New microtubules must be laid down. Initiation of a new microtubule probably takes place at the left edge of the lm and, to maintain an invariant pattern, at a fixed distance from the left terminus of the left-most pre-existing microtubule (Fig. 11). Once initiated, the new microtubule elongates anteriorly until the right edge of the lm is reached. The validity of this hypothetical mode of lm growth remains to be tested.

Fig. 10.

Drawing to show the arrangement of microtubules constituting the lm. A,P, R, L are, respectively, anterior, posterior, right and left. Each microtubule starts at the left edge of the lm, extends anteriorly and terminates at the right edge of the lm. This is the case for all microtubules except those cut off at the anterior and posterior poles of the cell. ably are open ends produced as a result of breakage of the lm at the mid-body region during cell fission. The cell, however, does not grow exclusively at one pole. Hence, the lm probably slides antero-posteriorly between the inner pellicular membrane and the epiplasm so as to make way for assembly at the anterior termini, (iii) The elongating microtubules, because of their angular arrangement in relation to the lm, extend until ultimately reaching or defining the right edge of the band. Clearly, then, elongation of the pre-existing microtubules is not sufficient for propagation of the lm. New microtubules must be laid down. Initiation of a new microtubule probably takes place at the left edge of the lm and, to maintain an invariant pattern, at a fixed distance from the left terminus of the left-most pre-existing microtubule (Fig. 11). Once initiated, the new microtubule elongates anteriorly until the right edge of the lm is reached. The validity of this hypothetical mode of lm growth remains to be tested.

Given this possible pattern of the microtubules making up the lm in Tetrahymena and the finding that the lm extends from posterior to anterior, the following mode of growth of the lm is suggested: (i) Those microtubules terminating on the right edge of the band do not elongate beyond the width of the lm. (ii) Elongation of the bn primarily takes place at the anterior termini of the microtubules (Fig. 11); these presum-

Fig. 11.

Hypothetical scheme showing propagation of the lm. Microtubules I and 2 terminate on the right edge (R) of the lm and no longer elongate. Microtubules 3-9 elongate anteriorly until finally reaching or defining the right edge of the lm (e.g. at t, microtubule 3). A new microtubule (10) is initiated at the left edge of the lm at a fixed distance (arrow) from the posterior terminus of microtubule 9.

Fig. 11.

Hypothetical scheme showing propagation of the lm. Microtubules I and 2 terminate on the right edge (R) of the lm and no longer elongate. Microtubules 3-9 elongate anteriorly until finally reaching or defining the right edge of the lm (e.g. at t, microtubule 3). A new microtubule (10) is initiated at the left edge of the lm at a fixed distance (arrow) from the posterior terminus of microtubule 9.

     
  • oa

    oral apparatus

  •  
  • lm

    longitudinal microtubular band

  •  
  • tm

    transverse microtubular band

  •  
  • pm

    postciliary microtubular band

  •  
  • c

    cilium

  •  
  • bb

    basal body

  •  
  • ik

    inverted ciliary row

  •  
  • rj

    right juncture

  •  
  • lj

    left juncture

  •  
  • drj

    developing right juncture in heteropolar doublets

  •  
  • dlj

    developing left juncture in heteropolar doublets

  •  
  • e

    epiplasm

This work is supported by grant 158/359 from the University of Hong Kong. Drs K. Aufderheide, R. V. Dippell, J. Frankel and Mr R. Hinrichsen have offered helpful suggestions on the manuscript.

Allen
,
C.
&
Borisy
,
G. G.
(
1974
).
Structural polarity and directional growth of microtubule of Chlamydomonas flagella
.
J. molec. Biol
.
90
,
381
402
.
Allen
,
R. D.
(
1967
).
Fine structure, reconstruction and possible functions of components of the cortex of Tetrahymena pyriformis
.
J. Protosool
.
14
,
553
565
.
Borisy
,
G. G.
,
Olmsted
,
J. B.
,
Marcum
,
J. M.
&
Allen
,
C.
(
1974
).
Microtubule assembly in vitro
.
Fedn Proc. Fedn Am. Sacs exp. Biol
.
33
,
167
173
.
Byers
,
B.
,
Shriver
,
K.
&
Goetsch
,
L.
(
1978
).
The role of spindle pole bodies and modified microtubule ends in the initiation of microtubule assembly in Saccharomyces cerevisiae
.
J. Cell Sci
.
30
,
331
352
.
Dentler
,
S. L.
&
Rosenbaum
,
J. L.
(
1977
).
Flagellar elongation and shortening in Chlamydomonas. III. Structures attached to the tips of flagellar microtubules and their relationship to the directionality of flagellar microtubule assembly
.
J. Cell Biol
.
74
,
747
759
.
Frankel
,
J.
,
Jenkins
,
L. M.
,
Doerder
,
F. P.
&
Nelsen
,
E. M.
(
1976
).
Mutations affecting cell division in Tetrahymena pyriformis. I. Selection and genetic analysis
.
Genetics, Princeton
83
,
489
506
.
Ng
,
S. F.
(
1978a
).
Directionality of microtubule assembly: an in vivo study with the ciliate, Tetrahymena
.
J. Cell Sci
.
33
,
227
234
.
Ng
,
S. F.
(
1978b
).
Origin and inheritance of an extra band of longitudinal microtubules in Tetrahymena cortex
.
Protistologica (in Press)
.
Ng
,
S. F.
&
Frankel
,
J.
(
1977
).
180°-rotation of ciliary rows and its morphogenetical implications in Tetrahymena pyriformis
.
Proc. natn. Acad. Sci. U.S.A
.
74
,
1115
1119
.
Ng
,
S. F
&
Nelsen
,
E. M.
(
1977
).
The protargol staining technique: an improved version for Tetrahymena pyriformis
.
Trans. Am. microsc. Soc
.
96
,
369
376
.
Pitelka
,
D. R.
(
1961
).
Fine structure of the silverline and fibrillar systems of three tetrahymenid ciliates
.
J. Protozoal
.
8
,
75
89
.
Ringo
,
D. L.
(
1967
).
Flagellar motion and fine structure of the flagellar apparatus in Chlamydomonas
.
J. Cell Biol
.
33
,
543
571
.
Rosenbaum
,
J. L.
,
Binder
,
L. L
,
Granett
,
S.
,
Dentler
,
W. L.
,
Snell
,
W.
,
Sloboda
,
R.
&
Haimo
,
L.
(
1975
).
Directionality and rate of assembly of chick brain tubulin onto pieces of neurotubules, flagellar axonemes and basal bodies
.
Ann. N.Y. Acad. Sci
.
253
,
147
177
.
Sale
,
W. S.
&
Satir
,
P.
(
1977
).
The termination of the central microtubules from the cilia of Tetrahymena pyriformis
.
Cell Biol. Int. Reports
1
,
45
49
.
Whitman
,
G. B.
(
1975
).
The site of in vivo assembly of flagellar microtubules. y4nn
.
N. Y. Acad. Sci
.
253
,
178
191
.
*

Numbering of ciliary rows starts from the mid-ventral row associated with oral morphogenesis and proceeds clockwise around the cell.

*

Throughout this report, ‘right’ and ‘left’ refer to the right and left of an observer, imagining that he stands inside and lines up antero-posteriorly with the cell and turns around his long axis to face the surface of the cell described. This convention gives a consistent description of the positions of neighbouring organelles on the entire surface of the cell. The terms ‘right-lateral’ and ‘left-lateral’ are used to describe the right and left sides of the cell, respectively, with reference to the ventral oral apparatus.