SPATIOTEMPORAL CONTROL OF FUNCTIONAL SPECIFICATION … · Nyctotherus ovalis. This may show whether mt wit 1h3 p mor become thae asn-sembled onto pre-existing tubules or are newly

y.Cell Sri. 76, 337-355 (1985) 337Printed in Great Britain © The Company of Biologists Limited 1985

SPATIOTEMPORAL CONTROL OF FUNCTIONALSPECIFICATION AND DISTRIBUTION OF SPINDLEMICROTUBULES WITH 13, 14 AND 15PROTOFILAMENTS DURING MITOSIS IN THECILIATE NYCTOTHERUS

U.EICHENLAUB-RITTER*

Institut fur Zellmorphologie, Ruhr-Universitat Bochum, D-4630 Bochum 1,Postfach 102148, FD.R.

SUMMARY

The formation of microtubules with more than 13 protofilaments in the ciliate Nyctotherus ovalisLeidy seems to be a highly ordered process. Such microtubules are restricted to the nucleoplasmand, moreover, to certain stages of nuclear division. They assemble during anaphase of micro-nuclear mitosis and during the elongation phase of macronuclear division. The number of micro-tubules with more than 13 protofilaments in the micronuclear nucleoplasm increases as anaphaseprogresses. Furthermore, assembly of microtubules with 14 and 15 protofilaments seems to proceedconcomitantly with net disassembly of 13-protofilament microtubules, because the total amount ofpolymerized tubulin in the interpolar spindle region remains approximately constant between midanaphase and late telophase. In addition, evidence for spatial control of the distribution ofmicrotubules with different protofilament numbers in the micronuclear stembody has been found.The percentage of microtubules with 13 protofilaments per stembody cross-section is highest at theends of the stembody, while the percentage of microtubules with either 14 or 15 protofilamentsincreases as the middle of the stembody is approached.

Temporal control of polymerization of microtubules with high protofilament numbers seems tobe exerted independently in the two types of nuclei. For example, when the macronucleus starts toelongate it contains microtubules with more than 13 protofilaments but the metaphase micronucleusstill possesses only microtubules with 13 protofilaments at this stage.

Control of fidelity of protofilament numbers is not lost in the early stages of micronuclear ormacronuclear division when cells are exposed to 2H2O or media containing taxol. Even microtubulesthat reassemble during recovery of metaphase micronuclei from nocodazole-induced microtubuledepolymerization, in either the absence or presence of 2HzO and taxol, possess 13 protofilaments.Similarly, if the introduction of microtubules with 14 and 15 protofilaments is inhibited during earlymicronuclear anaphase and delayed for 60 min by exposure to nocodazole, such microtubules stillassemble during telophase when recovery is permitted.

Microtubules that have been assembled under normal conditions show differential sensitivity tonocodazole. During metaphase, nocodazole induces disassembly of most microtubules. There is anincrease in microtubule stability that coincides with the appearance of microtubules with highprotofilament numbers during early anaphase. However, considerable numbers of 13-protofilamentmicrotubules, as well as microtubules with 14 and 15 protofilaments, exhibit such stability duringanaphase. Observation of some doublet-like microtubules in the micronucleus and elongating

•Present address: MRC Clinical and Population Cytogenetics Unit, Western General Hospital,Crewe Road, Edinburgh EH4 2XU, U.K.

Key words: microtubules, protofilament number, mitosis, spindle elongation, ciliate, nocodazole,taxol.

338 U. Eichenlaub-Ritter

macronucleus, and conspicuous hooks on some microtubules in the micronuclear stembodyprovides evidence for structural and, possibly, also functional similarities between thesemicrotubules and ciliary doublet microtubules.

INTRODUCTION

Most microtubules (mt) included in the cytoskeletal framework of eukaryotic cellsseem to be composed of 13 protofilaments (p) (Ledbetter & Porter, 1964; Tilney etal. 1973; La Fountain & Thomas, 1975; Eichenlaub-Ritter & Ruthmann, 1983). Theonly consistent deviation from this general pattern is observed either, in vivo, in theB and C tubules of doublets and triplets of cilia, basal bodies and centrioles or theirderivatives (Amos & Klug, 1974; Fujiwara & Tilney, 1975; Linck, 1976) or, in vitro,in mt assembled from tubulin under certain conditions (reviewed by Burton, 1981).In those few instances in which mt with other than 13 p occur in vivo, aside from thedoublets and triplets, the unusual microtubular lattice seems to be a cell-specificvariant of the ordinary microtubular architecture (Burton &Hinkley, 1974; Nagano &Suzuki, 1975; Burton, Hinkley & Pierson, 1975; Chalfie & Thomson, 1982; Saito &Hama, 1982; Davis & Gull, 1983). In contrast, the mt with more than 13 p, whichhave been reported for two different ciliate cells, the heterotrich Nyctotherus ovalisLeidy and the hymenostome Paramecium tetraurelia, do not constitute all, or evenmost, of the cells' tubules (Cohen, Beisson & Tucker, 1980; Eichenlaub-Ritter &Tucker, 1984). There have been indications that the formation, as well asdistribution, of these tubules is under stringent spatiotemporal control. For instance,mt with 14, 15 and 16 p occur only in the nucleoplasm of dividing ciliates. No mt withdiameters greater than 24 mn, or more than 13 p, have been observed among the manykinds of cytoplasmic mt in both ciliates, not even in those of arrays that arepolymerized exclusively during cell division (Eichenlaub-Ritter & Tucker, 1984).

One of the aims of the present investigation was to determine the spatialdistribution and exact time of assembly of mt with 13, 14 and 15 p in the ciliateNyctotherus ovalis. This may show whether mt with more than 13 p become as-sembled onto pre-existing tubules or are newly initiated and polymerized. Secondly,it may reveal whether assembly of mt with differing p numbers is an ordered process.It is conceivable that control of mt fidelity is lost in favour of rapid mt assembly. Inciliates, a fast and extensive elongation of the interpolar micronuclear spindle regionseems to be the main mechanism by which mitotic separation of the chromatin isachieved. For instance, the mt bundle comprising the central portion of the micro-nucleus at telophase, the stembody, can extend to lengths of more than 30/im inNyctotherus (Eichenlaub-Ritter, 1982). The p number of mt that are assembled undercertain conditions in vitro, and in the absence of microtubule-organizing centres(MTOCs), depends merely on temperature, pH and ionic environment (Burton &Himes, 1978). If the same applies to rapidly polymerizing mt in the stembody onewould expect that mt with 13,14 and 15 p would be randomly distributed with respectto each other. In contrast, the results of the present study demonstrate that spatial andtemporal order occur during the formation and positioning of mt with different pnumbers.

Fidelity of protofilament numbers in microtubules 339

In this investigation, three approaches have been used to assess which factors mightinfluence fidelity of p number during polymerization of mt. Dividing ciliates were ex-posed to 2HzO-containing media to disturb the cellular polymerization equilibrium andinduce unscheduled assembly. In cells with open mitosis, heavy water enhancespolymerization, although this occurs in a fashion that still seems to be governed byMTOCs (reviewed by Takahashi & Sato, 1982). The plant derivative taxol stabilizesextant mt (Schiff & Horwitz, 1980) and induces assembly of free cytoplasmic tubulininto stembody- or aster-like microtubular aggregates in living cells (Heidemann & Gal-las, 1980; Simoneef al. 1981; Masurovsky, Peterson, Crain & Horwitz, 1982; Schat-ten, Schatten, Bestor & Balczon, 1982). During taxol-induced polymerization, site-specificity may be lost (De Brabanderet al. 1981). In the present study, interphase cells,as well as dividing cells, were exposed to this compound and its effects on the cell'smicrotubular complement were monitored. Ciliates were also exposed to the syntheticantimitotic agent nocodazole, because it readily penetrates the ciliate cell (Eichenlaub-Ritter & Ruthmann, 1983), and its action is reversible. The consequences of especiallyrapid mt assembly following mt depolymerization and release from drug treatment hasbeen examined to ascertain whether fidelity of p number is lost under these conditions.

Since mt with 14 and 15 p are restricted to certain cellular compartments anddefined times of cell development, it seems likely that structural modification goeshand in hand with functional specification (Eichenlaub-Ritter & Ruthmann, 19826).The possibility of correlations between microtubular structure, stability and specificfunctions of individual mt classes has been explored.

MATERIALS AND METHODS

Nyctotherus ovalis Leidy (Ciliata, Heterotricha) is a commensal of the cockroach intestine.Rearing conditions for the host, Blaptica dubia, isolation procedure and optimal conditions fortemporary cultivation of ciliates have been described elsewhere (Eichenlaub-Ritter & Ruthmann,1982a). Since micro- and macronuclear division coincide with cell division in this ciliate cells canbe selected individually for treatment or fixation at a certain stage according to the shape of the celland the position and form of the macronucleus (Eichenlaub-Ritter & Ruthmann, 1982a).

Heavy water treatment was carried out in isolation medium (0-9 % NaCl, 0-01 % NaHCC>3, 0-2 %Casamino acids (DIFCO), pH adjusted to 6-5) made up in 70 %2H2O. Taxol (a kind gift from Dr M.de Brabander) was dissolved in dimethylsulphoxide (DMSO) to give a concentration of 10~2 M. Thisstock solution was diluted with isolation medium (final concentration between 10~s and 10~4 M-taxol).For nocodazole experiments (Janssen Pharm., Beerse) a stock solution of 0-4mg nocodazole/mlDMSO was diluted to a final concentration of 2/ig nocodazole/ml isolation medium. In all experi-ments controls were exposed to corresponding amounts of DMSO alone. Incubation of cells wascarried out at 28 °C, the optimal temperature for culture of these ciliates. Development of cells andprogression of ciliates through division was followed by observation with a dissecting microscope.

For electron microscopy some cells were fixed directly in 2-5 % glutaraldehyde in sodiumcacodylate buffer, osmicated and embedded as previously described (Eichenlaub-Ritter & Ruth-mann, 1982a).

For tannic acid mordanting of the microtubular protofilament lattice the ciliates have to bepermeabilized before fixation. This is necessary because the thick ciliate pellicle and the envelopeof the micro- and macronucleus prevent penetration of tannic acid into the cell (Eichenlaub-Ritter& Tucker, 1984). Ciliates were exposed for 90s to a mt-stabilizing buffer, PM2G medium (Duerr,Pallas & Solomon, 1981) containing 0-1 % NP 40. Fixation was then started by addition of somedrops of a freshly made solution of 8 % tannic acid (Roth, Karlsruhe) in 0-15 M-cacodylate buffer


(pH 7-2) and 2-5 % glutaraldehyde. After several changes cells were left in fixative for 1 h (roomtemperature). After washing (5 times, 12 h at 4 °C) cells were osmicated and embedded (Eichenlaub-Ritter & Tucker, 1984). For sectioning, ciliates can be oriented so that the spindles will be cutapproximately perpendicular to the division axis. By alternately observing thin sections in theelectron microscope and shifting the angle of sectioning slightly, one can adjust the block in such away that cross-sections of most mt, exactly perpendicular to their long axis, can be obtained. It is onlyfor such mts that all protofilaments can be detected clearly and p number determined unequivocally.From such series of sections, the best ones containing high numbers of mt with recognizable p patternwere chosen for the determination of the relative contents of mt with 13, 14 and IS p. Symbols for pnumbers of individual mt were marked on transparent sheets placed over prints of electronmicrographs (final magnification between X 150000 and X 180000). From these the total overallnumber of mt present per stembody cross-section, as well as the percentage of mt for which p numberscould not be detected clearly, and mt with particular p numbers were determined.

RESULTS

Formation and distribution of microtubules with 14 and 15 protofilaments

The present study confirmed earlier observations that indicated that, beforeanaphase, all mt present in the mitotic micronucleus consistently possess 13 p.

W#C*>

Fig. 1. Cross-sections of metaphase micronucleus after tannic acid mordanting, A. Al-though the nucleus is swollen, deterioration of the spindle (sp) is not observed. Theholokinetic composite chromosomes (cch) are still aligned at the equatorial plane, ma,macronucleus; nu, nucleoplasm. xl3 000. B. Tannic acid staining reveals that kinetochoremt are composed of 13 p. x400000.


Although the micronucleus swells during fixation, possibly due to the presence ofdetergent (J. B. Tucker, personal communication), the metaphase spindle remains asa compact body in the centre of the nucleoplasm. No deterioration of mt or displace-ment of composite chromosomes from the spindle equator was detected (Fig. 1A).Kinetochore mt (kmt) are closely spaced, but, as noted in conventionally fixed cells(Eichenlaub-Ritter & Ruthmann, 1982a), do not touch each other (Fig. 1A-B). Thelarge, single macronucleus of dividing ciliates, in which the micronucleus is inprophase to metaphase of mitosis, usually has a flattened, ellipsoidal shape, as ininterphase. Few mt are found in its nucleoplasm; they are apparently distributed atrandom. In all cross-sections obtained from such macronuclear mt, the tubules arecomposed of 13 p. Occasional macronuclei that have started to elongate and attain amore cylindrical shape may, however, be observed in ciliates in which themicronucleus is still in metaphase. In these, the macronucleus contains well-alignedmt bundles running parallel to the long axis of the macronucleus at some distance fromthe nuclear membrane. Most tubules are composed of 13 p but some always possess14 or 15 p. Mt in the metaphase micronuclei of such cells possess only 13 p.

Microtubules with 14 and 15 p are first found in the micronucleus at early anaphase.They are mostly, but not exclusively, observed in the interpolar space of the alreadyslightly elongated micronucleus. Although nuclear swelling has occurred, some mtremain attached to the nuclear membrane. The rest of the interpolar mt do not form

Fig. 2. A, B. Small bundles of mt in the interpolar part of an early anaphase micronucleusinclude mt with 15 p (arrow in A) and mt with hooks (arrowhead in B). X320000 . C. Fourserial sections in which the termination of a mt (arrow in the left panel) in some electron-dense materail can be followed. The mt has completely disappeared in the last section onthe right side. xl60000.


Fig. 3


a compact spindle body but are singly spaced or form groups of cross-bridged mt inthe electron-translucent nucleoplasm (Fig. 2A, B). Some mt possess 13 p but otherswith 14, and IS p are also observed (Fig. 2A-C). Counts of the number of mt withdiffering p numbers in random sections of two mid- to late anaphase micronucleirevealed that most consist of 13 p (Table 1). About 30 % have higher p numbers. Noobvious pattern in the spatial relationship of neighbouring mt relative to each otherwas apparent. The terminal portions of most sheath mt underneath the nuclear mem-brane, and of mt in the more central portion of the micronucleus (Table 1), end incontact with electron-dense material (Fig. 2c). Some have no visible MTOCassociated with their ends. The p number of each individual mt seems to remainconstant. Opening up of 13pmt and incorporation of an additional p was neverobserved. Micronuclear mt with 'hooks' are seldom found at this stage (Fig 2). Themacronucleus of each cell with an anaphase micronucleus is typically cylindrical; itcontains well-aligned mt bundles and most mt possess 13 p, although a few have 14or 15 p.

During telophase the central portion of the dumb-bell-shaped micronucleusbecomes occupied by a stembody composed of many parallel aligned mt (Eichenlaub-Ritter, 1982). Swelling of the nucleus is, at this stage, less pronounced, although theentire stembody may sometimes curve into a corkscrew shape. However, it is stillpossible to obtain cross-sections in which a considerable number of mt reveal their pvalues (Fig. 3). Counts of the relative numbers of mt with 13, 14 and 15 p indicatethat, in the course of anaphase, an alteration takes place in the mt complement. Thestembody is largely composed of 14 p mt. A well-defined relationship in the arrange-ment of mt with differing p numbers is still not demonstrable. For instance, attemptsto show a zonal gradient of mt positioning, in which mt with high p numbers occupythe periphery of the nucleus to a greater extent than those with 13 p or vice versa, wereunsuccessful. However, comparisons of the percentage of mt with differing p num-bers in sequential sections of stembodies showed that, proceeding from the end of astembody towards its centre, there is an increase in the number of mt per stembodycross-section (Eichenlaub-Ritter, 1982), and the proportion of mt with 15 pincreases,while that with 13 p decreases (Table 1). To exclude errors in the determination ofrelative content of mt, closely adjacent sections of the same stembody with variablenumbers of mt revealing their p pattern were compared. The relative percentages ofmt with differing p numbers is nearly identical (Table 1, triangle). Counts made atthe extreme ends and in more central regions of other stembodies confirm this finding.While many mt with 13 p are present at the ends of the stembody, their numberdecreases towards the more central part, while the percentage of mt with 14 p in-creases slightly and that with 15 p increases markedly (Table 1).

The cylindrical macronucleus of ciliates in which the micronucleus has reachedtelophase is stretched and lies parallel to the division axis and in the centre of the cell.

Fig. 3. Cross-section through a telophase stembody; mt are closely spaced, swelling of thenucleus has not occurred; mt with 13,14 and 15 p are lying side-by-side (arrow). Most mtare composed of 14 p. xl40000.


Table 1. Percentage of microtubules with 13, 14 and 15 protofilaments present in theinterpolar space of micronuclei at anaphase and telophase

Mid anaphase

TelophaseEnd of stembody

Near central part ofstembody

Total

230243

730759

A 855A 856

Counted

3696

x 132

122100

145313383201

13(%)

75-064-667-4

51-637-0

19-322-3

9-79-9

14(%)

22-025-024-2

45-955-0

64-164-2

60-062-2

15(%)

2-810-4

8-3

2-48-0

16-613-330-327-9

Data for anaphase micronuclei represent the combined counts of p content of mt in severalsections of two anaphase micronuclei (first two lines). Average percentage of mt with different pnumbers is given in the third line (x). Data for p content of mt in stembodies are based uponindividual sections of two representative nuclei in which the total number of mt present per cross-section could also be determined (first column total). The middle two lines show counts made at theend of a stembody while the four lines below are for the more central portion of a stembody. Thetriangles indicate that data were obtained from two consecutive thin sections. The percentage of mtwith 14 and 15 p increases from anaphase to telophase and from the end towards the middle of thestembody.

As noted already for earlier stages, macronuclear mt are largely composed of 13 p.However, close observation consistently reveals some mt with 14 and 15 p, andoccasionally even 16 p (Eichenlaub-Ritter & Tucker, 1984). Other noteworthyfeatures of macronuclear mt at this stage include the close juxtapositioning of adjacentmt and the frequent occurrence of mt with hooks or doublet-like microtubular con-figurations (Fig. 4A). These do not seem to result from lateral association of the endsof one mt and the tip of another. Analysis of serial sections showed that both membersof a 'doublet' terminate at the same point. As in the micronucleus, no visible MTOC,except some electron-dense material, seems to be associated with the microtubularends. In the micronucleus, similarly, doublets or mt with hooks can be found in thestembody, predominantly in the more central portion of it (Fig. 4B). Such mt con-figurations were never observed in metaphase micronuclei.

Microtubule fidelity after treatment with heavy water and taxol

In order to initiate assembly of additional mt (e.g. see Sato, 1982), and possibly alsosuch mt with high p numbers, cells containing prophase micronuclei were transferredto isolation medium made up in 70 % 2HzO. No change in cell behaviour or structurewas detected at the light-microscope level. Examination of metaphase micronuclei inthe electron microscope also reveals no changes in p numbers. Kinetochoremicrotubules (kmt) consist of 13 p (Fig. 5A). Depolymerization appears undisturbed

Fidelity of protofilament numbers in microtubules m

4A

Fig. 4. Microtubules with hooks and doublet mt (arrows) that are similar to the A and Btubules of cilia found among mt bundles in an elongating macronucleus (A) and in theinterpolar space of a mid-anaphase micronucleus (B). X240000 .

by 2HzO since the kmt of cells that have entered anaphase while subjected tohave shortened. Microtubules with more than 13 p are found only in the interpolarpart of anaphase nuclei. Those mt that, characteristically for this ciliate's mitosis,extend from the polar parts of the spindle through the chromatin of the compositechromosomes into its interpolar region at early anaphase, are composed of 13 p (Fig.5B), as in the controls. Control of p number seems to be retained in all 2HzO-exposedcells.

Taxol may induce the formation of stembody-like mt even in non-dividing cells (DeBrabander et al. 1981). Therefore, interphase ciliates were treated for up to 4h with10~5 M-taxol. No unusual microtubular aggregates were observed in the nuclei or thecytoplasm. In taxol-treated early division stages (60min, 10~5M) the compositechromosomes of metaphase micronuclei become normally aligned at the spindleequator. However, the spindle itself seems to be less compact after the treatment.Most kmt no longer extend to the nuclear membrane at the spindle poles but terminate

U. Eichenlaub-Ritter

:*« JO

Fig. 5. Treatment of dividing micronucleus with 2H2O and taxol. A. Section through thekinetochore of a micronucleus that had entered anaphase while subjected to 2H2O. Allkinetochore mt possess 13 p. x260000. B. A mt with 13 p passing through the chromatininto the interpolar spindle region of the same micronucleus. x260000. c. Taxol-induced'pseudoasters' in a metaphase micronucleus. x67 000. D. Kinetochore mt of a taxol-exposedmetaphase micronucleus with 13 p. x200000. E. Microtubules composed of 14p and,possibly, one with 15 p in a telophase micronucleus after taxol treatment.

in small 'pseudoasters' in the polar nucleoplasm (Fig. 5c). Nevertheless, only mt with13 p were detected (Fig. 5D). Macronuclei of the same cells are still devoid of mt, asin control organisms. Anaphase micronuclei subjected to 10~sM-taxol (60min) ex-hibit no gross alteration in the microtubular skeleton either. However, during lateanaphase spatial alignment of mt may be disturbed by taxol. The elongated interpolarregions of such micronuclei are sometimes unusually curved. Interdigitating groups


of mt run obliquely with respect to the spindle axis. Because of this, only a few mt canbe sectioned perpendicular to their long axis for assessment of p number (Fig 5E) . Forone nucleus, about 70 % of these mt were composed of 13 p, while 30% possessed 14and 15 p. This ratio is comparable to that determined for untreated organisms. Manymt bundles are found in the dividing macronuclei of taxol-treated cells. As is the casein the micronucleus, their alignment sometimes seems perturbed by the treatment,but p numbers appear normal. Microtubules with 13 p predominate but mt with 14and 15 p are also present.

Nocodazole-stability of microtubules with differing protofilament content

Microtubules in the stembody of Nyctotherus are much more stable with respect tolow temperatures and vinblastine than those polymerized earlier, during prophase(Eichenlaub-Ritter & Ruthmann 19826). A similar increase in stability of nuclear mtcan be shown after nocodazole treatment. Exposure to 2/ig nocodazole for 20minleads to the complete disassembly of mt in the metaphase micronucleus. Kinetochoreplates are devoid of mt and the composite chromosomes become dispersed in thenucleoplasm (Fig. 6A). AS after treatment with cold and vinblastine, most mt in thepolar part of anaphase micronuclei also become depolymerized. However, a sheath ofmt underneath the nuclear membrane in the interpolar spindle region is nocodazole-insensitive. Stable mt are composed of 13-15 p (Fig. 6B). The mt that remain in thecentral part of the interpolar spindle also possess from 13 to 15 p. Stembody mt ofnocodazole-exposed telophase micronuclei {2(ig, 60min) are, apparently, allnocodazole-insensitive. Although mt with 14 p dominate, mt with 13 and 15 p are alsopresent (Fig. 6D). NO mt were found in macronuclei at early stages of division afternocodazole treatment. However, mt of elongating, cylindrical macronuclei show thesame stability towards the drug as the stembody mt, irrespective of their p numbers(Fig. 6c).

Although mt that are already present in the stembody are resistant to nocodazole,assembly of new mt is not initiated in the presence of the drug. Ciliates exposed tonocodazole during early division and left in the drug-containing medium for extendedperiods (up to 4ih) do not exhibit the typical changes in cell or macronuclear shape.Electron-microscopic observation reveals that the condensation/decondensationcycle of micronuclear chromatin proceeds. Nucleolus-like spheres, which are typicalfor this ciliate's telophase, are formed as in the control. However, the nucleoplasm ofmacro- and micronuclei is depleted of mt. The chromatin remains undivided in thecentre of the micronucleus. Occasionally, some unordered mt that are preferentiallylocated underneath the nuclear membrane may be observed; they possess 13-15 p.

Nuclear microtubules in ciliates that have recovered from nocodazole treatment in theabsence and presence of2H2O and taxol

In ciliates that have been exposed to nocodazole during early division (60min,2^ig/ml) and then transferred to drug-free isolation medium, a rapid reassembly ofmt takes place in the metaphase micronucleus. After 20 min of recovery the composite


Fig. 6. Nocodazole-exposed ciliates. A. Kinetochore microtubules become depolymerizedand composite chromosomes (cch) are dispersed in the micronuclear nucleoplasm; kineto-chores (arrows) are structurally unchanged. x 17 000. B . Stable mt with 13 p underneath thenuclear membrane (rwi) of a mid anaphase micronucleus, x320000. c. Nocodazole-stablemacronuclear mt. x320000. D. Portion of a nocodazole-insensitive telophase stembody.Most mt possess 14 p, some with 13 p are also evident (arrow). x370 000.


chromosomes are, again, aligned at the spindle equator. The p number of the stillshort kmt is 13, as in untreated controls.

Late metaphase to early anaphase cells that are treated with nocodazole for 60 minalso rapidly resume mitosis after drug release. The previously roundish cells elongateand the macronuclei stretch within 15 min. Many mt are reassembled in thenucleoplasm of the micronucleus, which contains the decondensed, although un-divided, chromatin and nucleolus-like spheres typical of telophase. The micronucleiare irregularly lobed, with mt bundles traversing the nucleoplasm in all directions.Although no bipolar stembody is formed, reassembled mt possess from 13 to 15 p, asnormally expected for this stage. Even when recovery takes place in 70% 2HzO,unusual, free aggregates of mt are not found either in the micro- or macronucleus orin the cytoplasm. Metaphase mt in such cells are composed of 13 p.

The polymerization-inhibiting effect of the nocodazole treatment is not overcomeby the addition of up to 2 X 10~5M-taxol to the medium. However, after 30 min ofrecovery in 10~4M-taxol without nocodazole composite chromosomes become realig-ned in the equatorial plane. Kinetochore microtubules have formed and extendtowards the spindle poles (Fig. 7A). AS observed already in ciliates exposed to taxolalone, the spindle is broader and flatter than usual. Kinetochore microtubules appearless bundled and form unusual lateral associations. Many of them traverse the polarspace, nearly perpendicular to the division axis. They do not terminate at discrete sitesat the inner nuclear membrane but either end free in the nucleoplasm or are attachedto small, aster-like foci, which are located at some distance from the nuclear mem-brane. All mt present in the micronucleus seem to be composed of 13 p (Fig. 7B). Inmicronuclei that have entered anaphase during recovery in nocodazole-free, taxol-containing medium, control of p number also appears undisturbed. A nearly normal-looking stembody may be formed during recovery in taxol. Nucleolus-like spheres arepresent among the decondensing chromatin in the future daughter nuclei of latedivision stages. Stembody mt occupy the nucleoplasm in the interpolar part of thespindle but are not as well aligned with each other as in the controls. The stembodyappears less compact, particularly in its central portion, and mt are not so closelyspaced or cross-bridged compared to untreated controls. Although alignment is dis-turbed, stembody mt possess 13—15 p, as in controls (Fig. 7c).

Macronuclear mt are found in abundance in late stages of macronuclear divisionswhen recovery is taking place in taxol, indicating a polymerization-promoting effectof the drug at this time. Site-specificity of assembly seems to be retained because mostmt bundles still occur close to the nuclear membrane. Microtubules with 13 and 14 pwere found (Fig. 7D).

DISCUSSION

Polymerization of microtubules vrith 14 and 15 protofilaments in anaphasemicronuclei

The present study demonstrates that mt with more than 13 p are assembled duringonly the later stages of micronuclear division in N. ovalis. All metaphase mt seem to


Fig. 7. Nocodazole treatment and recovery in taxol-containing medium, A. Compositechromosomes become realigned at the spindle equator during recovery. Most mt extend-ing towards the poles are found in the vicinity of the kinetochores. They are often laterallyassociated with each other or with some densely staining material (arrowheads), xl 1 800.B. Reassembled metaphase mt still possess 13 p. x330OOO. c. Microtubules with 13 p(arrow) and 14 p (arrowhead) in a nocadozole-recovering, taxol-exposed late anaphasemicronucleus; nra, nuclear membrane. x300000. D. Microtubules repolymerized in themacronucleus during recovery in taxol. x300000.

consist of 13 p. The proportion of mt with higher p numbers increases in themicronucleus from anaphase to telophase. Hence, tubules with 14 and 15 p mustassemble in the nucleoplasm during anaphase. Our first question concerned the originof mt with more than 13 p. Theoretically, it is possible that the 13 p mt open up toincorporate additional subunits. The latter has, to the author's knowledge, never been


observed, although mt with more than 13 p rarely occur in living cells. In vitroassembled mt seem to possess one specific p number (Burton, 1981). Even in thosecases in which the absolute p number changes (for instance, when 'incomplete' Btubules of axonemal doublets initiate outgrowth of 'complete' mt with 13 or 15 p), theangle of binding between neighbouring subunits appears to remain constant all alongthe tubule (Scheele, Bergen & Borisy, 1982). Linear continuity between mt portionswith different p numbers was also never found in Nyctotherus.

What defines and determines the binding and positioning of the first protofilamentsin the newly formed mt with more than 13 p in Nyctotherus}. At the ultrastructurallevel at least, the presence of a specific MTOC for each kind of mt was not detected.Electron-dense material was present at the ends of some mt with 13, 14 and 15 p butmany mt seem to terminate without any recognizable structure attached to their ends.The only structure we found that might influence the appearance of mt with high pnumbers is the nuclear membrane. In the micronucleus, mt with 14 and 15 p are oftenassociated with the nuclear membrane at early anaphase. In the macronucleus mt withhigh p numbers are found only among the mt positioned near the nuclear membranein late stages of division. If the nuclear envelope is involved in initiation of assemblyof mt with 14 and 15 p, this must occur in a fashion much different from that occurringin a typical MTOC. Usually polymerization is induced so that the mt project perpen-dicularly from the nucleating surface of the MTOC.

The relative proportion of mt with high p numbers increases in the micronucleusfrom mid-anaphase to telophase. This raises the question of whether continuousinitiation and polymerization of these tubules occur during anaphase. Provided ourdata on the relative numbers of mt with different p numbers in the interpolar spaceare representative of all anaphase micronuclei, the total number present per cross-section in the interpolar space of anaphase and telophase micronuclei determined in anearlier study (Eichenlaub-Ritter, 1982) can be used to estimate the total number of mtper cross-section with a particular p number present in the interpolar space (Table 2).

For example, approximately 25 % (see Table 1) of the 2500-3000 mt scored percross-section in a similar anaphase stage in the earlier study should have 14 p. Thisyields 600-750 mt with 14 p. In the mid-portion of telophase stembodies between750-1000 mt are present, which means that 450-650 mt (roughly 60-65 %) possess14 p. Surprisingly, the total number of mt with 14 p seems to be almost identical inanaphase and telophase nuclei, although their relative proportions compared withthose of the 13 p mt increase (from less than 30% to 65%). This implies that the

Table 2. Estimated absolute numbers ofmt with 13, 14 and 15ppresent in anaphaseand telophase micronuclei

13pmt 14pmt 15pmt

Anaphase 1600-2000 600-750 200-250Centre of stembody 110-150 450-650 255

For explanation see text.


14 p mt that start to assemble during early anaphase may elongate further during laterstages of mitosis. Neither initiation of new 14pmt nor gross disassembly of thesetubules may be necessary for stembody formation. The same seems to be true for the15pmt. For example, 200-250 mt with 15 p (8% of 2500-3000 mt) are alreadypresent during anaphase. This value corresponds closely with the estimate of 255 for15 p mt (30 % of 850 mt) in the central portion of the telophase stembody. In contrast,the 1600-2000 mt with 13 p (67 % of 2500-3000) present during anaphase decreaseto 110-150 mt (15% of 750-1000) in the centre of the telophase stembody, and evenat its ends only about 115 mt (50 % of 230) with 13 p are found. Does this mean that13 p mt become disassembled during anaphase? It is also possible that mt with 13 pbecome redistributed along the entire stembody while mt with 14 and 15 p elongate.If this is the case, the total polymerized tubulin content of micronuclei should increasefrom anaphase to telophase. This, however, was not found. In mid to late anaphasemicronuclei, the average number of 2500-3000 mt are situated in an interpolar regionwith a length of 4-8 nm. Therefore, 1-2-4 cm of polymerized tubulin is present in theinterpolar space during anaphase. The stembody possesses 2-6 m of polymerizedtubulin (2x5/tfn with 500mt plus 21/im with 1000mt; see Eichenlaub-Ritter,1982). Thus, the total amount of polymerized tubulin present during mid to lateanaphase is nearly the same as that in the telophase stembody, although the relativeproportion of mt with high p numbers increases. This suggests that a preferentialdepolymerization of mt with 13 p occurs during anaphase while polymerization oftubules with 14 and 15 p is favoured.

Control of microtubule fidelity

Since the amount of polymerized tubulin is about the same in the interpolar spaceof anaphase and telophase micronuclei, it would be most uneconomical for the cell tosynthesize new tubulin molecules for the assembly of 14 and 15 p mt while free tubulinfrom the depolymerizing 13 p is readily available. It seems more probable that freetubulin from the 13 p mt may become incorporated into the growing 14 and 15 p mt.If this were the case, identical tubulin would be present in the 13, 14 and 15pmt.Since the actual number of p present in the mt stembody varies from one ciliate speciesto another (Eichenlaub-Ritter & Tucker, 1984), it appears most unlikely that eachciliate possesses different tubulins which induce the formation of mt with its ownspecific p number.

Taxol, which can possibly act like a microtubule-associated protein (MAP) andmay abolish the preferential initiation of mt assembly at MTOCs, has been shown todisturb the interactions between mt in such protozoans as the heliozoa (Hausmann,Linnenbach & Patterson, 1983). It also affects the alignment of mt in themicronuclear spindle oiNyctotherus, in a similar way to that observed in the centriole-free spindles of plant cells (Mole-Bajer & Bajer, 1983) or PTK cells (De Brabanderet al. 1984). Whether the formation of 'pseudoasters' in the nucleoplasm, flatteningof the spindle and induction of some lateral associations of kmt in Nyctotherus iscaused mainly by taxol binding to tubulin, its substitution for MAPs or disturbanceof site-specific assembly cannot be decided from the present observations. However,


it can be shown that temporal control of polymerization of mt with 14 and 15 p isindependent of the alignment of mt and, possibly of the position of initiation sites inthe nucleus. Even in those cells in which the spatial order has been disturbed, as seenin the taxol-treated cells as well as in nocodazole-released late division stages, thecorrect kind of mt becomes polymerized.

The present study shows also that sudden changes in tubulin concentration alonedo not determine mt structure. Neither 2HzO treatment and exposure to taxol norrelease from a nocodazole block cause the precocious polymerization of mt with morethan 13 p. Hence, loss of control of mt fidelity in favour of rapid mt assembly duringstembody formation does not seem to occur.

Function of microtubules with 14 or 15 protofilaments

Microtubules with more than 13 p occur in the nucleoplasm of the micronucleusand the macronucleus of Nyctotherus only at stages during which both nuclei undergoelongation. It is possible, therefore, that mt with 14 and 15 p are functionally specifiedfor essential roles in the elongation procedures. Does the stembody exert a pushingforce and, if so, to what extent is mt polymerization (Bajer, Cypher, Mole-Bajer &Howard, 1982; King & Hyams, 1983) and/or intertubule sliding (Eichenlaub-Ritter,1982) involved? A cone-like protrusion of micronuclear membrane in vinblastine-treated Nyctotherus seems to be produced by the assembly of stembody-like mt in thisregion of the nucleus (Eichenlaub-Ritter & Ruthmann, 19826). Similarly, lobes intelophase micronuclei, which were observed in the present study in cells released froma nocodazole block, always contain mt. The micronuclei of drug-exposed ciliatesdevoid of mt have a more or less spherical shape. Even in anaphase to telophasemicronuclei of nocodazole-released cells an elongation of the spindle is observed,although direct interactions between neighbouring tubules seem to be hindered tosome extent by the wide spacing of adjacent mt. Therefore, the assembly of mt mayparticipate in the formation of the stembody and its extension.

However, in all cases in which mt in the interpolar space appeared badly aligned,and widely and unevenly spaced, a more pronounced curving of the stembody wasobserved. This could mean that only in those regions of the spindle in which mt areable to contact each other and slide does additional elongation occur. Intertubulesliding could also account for the unequal, non-random distribution of mt with 13,14and 15 p along the stembody. If, for instance, 15pmt become preferentiallydislocated they could accumulate in the middle of the spindle. In this context, it ispertinent that the unusual tubular polymorphs that resemble ciliary doublet mt interms of structure (see Fig. 4A, B) and anti-mitotic drug stability are found onlyamong the stembody and macronuclear mt bundles containing mt with 14 and 15 p.If these tubulin polymorphs share other common properties with ciliary doublets theymight possess attachment sites for dynein-like ATPases and provide a basis for inter-tubule sliding similar to that proposed for spindle elongation during anaphase B inorganisms with open mitoses (Sakai et al. 1976; Cande, 1982). At present we aretesting this hypothesis by investigating the influence on micro- and macronucleardivision of agents that specifically inhibit dynein ATPases.

354 U. Eichenlaub-Bitter

The author thanks Dr A. Ruthmann for his support during the study, Dr J. B. Tucker for manyhelpful suggestions and for critical reading of the manuscript, and Mrs M. Seebach for excellenttechnical assistance.

REFERENCESAMOS, L. A. & KLUG, A. (1974). Arrangement of subunits in flagellar microtubules. J. Cell Set.

14, 523-549.BAJER, A. S., CYPHER, C , MOLE-BAJER, J. & HOWARD, H. M. (1982). Taxol-induced anaphase

reversal: evidence that elongating microtubules can exert a pushing force in living cells. Proc.natn.Acad. Sci. U.SA. 79, 6569-6573.

BURTON, P. R. (1981). Polymorphic assemblies of tubulin. In Cell and Muscle Motility, vol. 1 (ed.R. M. Dowben & J. W. Shay), pp. 289-333. New York. London: Plenum.

BURTON, P. R. & HIMES, R. H. (1978). Electron microscope studies on pH effects on assembly oftubulin free of associated proteins: Delineation of substructure by tannic acid staining. J. CellBiol. 77, 120-133.

BURTON, P. R. & HINKLEY, R. E. (1974). Further electron microscopic characterization ofaxoplasmic microtubules of the ventral nerve cord of the crayfish. J. submiscrosc. Cytol. 6,311-326.

BURTON, P. R., HINKLEY, R. E. & PIERSON, G. B. (1975). Tannic acid-stained microtubules with12, 13 and 15 protofilaments. J. Cell Biol. 65, 227-233.

CANDE, W. Z. (1982). Nucleotide requirements for anaphase chromosome movements in per-meabilized mitotic cells: anaphase B but not anaphase A requires ATP. Cell 28, 15-22.

CHALFIE, M. & THOMSON, J. N. (1982). Structural and functional diversity in the neuronalmicrotubules of Caenorhabditis elegans.J. Cell Biol. 93, 15-23.

COHEN, J., BEISSON, J. & TUCKER, J. B. (1980). Abnormal microtubule deployment duringdefective macronuclear division of a Paramecium mutant. J. Cell Set. 44, 153-167.

DAVIS, C. & GULL, K. (1983). Protofilament number in microtubules in cells of two parasiticnematodes. J. Parasit. 69, 1094-1099.

D E BRABANDER, M., D E MEY, J., GEUENS, G., NUYDENS, R. & WILLEBRORDS, R. (1984). The

effects of taxol on microtubule assembly throughout the mitotic cycle in PTK cells. J. submicrosc.Cytol. 16, 1-2.

D E BRABANDER, U., GEUENS, G., NUYDENS, R., WILLEBRORDS, R. & D E MEY, J. (1981). Taxolinduces the assembly of free microtubules in living cells and blocks the organizing capacity of thecentrosomes and kinetochores. Proc. natn. Acad. Sci. U.SA. 78, 5608-5612.

DUERR, A., PALLAS, D. & SOLOMON, F. (1981). Molecular analysis of cytoplasmic microtubulesin situ: identification of both widespread and specific proteins. Cell 24, 203-211.

EICHENLAUB-RITTER, U. (1982). Micronuclear mitosis in the ciliate Nyctotherus ovalis Leidy: anunusual chromosome arrangement and kinetochore structure and its implication for spindlestructure and function. In Microtubules in Microorganisms (ed. P. Cappuccinelli & N. R. Mor-ris), pp. 351-376. New York, Basel: Marcel Dekker.

EICHENLAUB-RITTER, U. & RUTHMANN, A. (1982a). Holokinetic composite chromosomes with"diffuse" kinetochores in the micronuclear mitosis of a heterotrichous ciliate. Chromosoma 84,701-716.

EICHENLAUB-RITTER, U. & RUTHMANN, A. (19826). Evidence for three "classes" of microtubulesin the interpolar space of the mitotic micro-nucleus of a ciliate and the participation of the nuclearenvelope in conferring stability to microtubules. Chromosoma 85, 687-706.

EICHENLAUB-RITTER, U. & RUTHMANN, A. (1983). The oral apparatus of Nyctotherus ovalisLeidy: possible involvement of microtubules in the structural support of a cortical domain andthe translocation of vesicles. Differentiation 24, 97-110.

EICHENLAUB-RITTER, U. & TUCKER, J. B. (1984). Microtubules with more than 13 protofilamentsin dividing nuclei of ciliates. Nature, Land. 307, 60—62.

FUJIWARA, K. & TILNEY, L. G. (1975). Substructural analysis of the microtubule and its polymor-phic forms. Ann. N. Y. Acad. Sci. 253, 27-50.

HAUSMANN, K., LINNENBACH, M. & PATTERSON, D. J. (1983). The effects of taxol onmicrotubular arrays: in vivo effects on heliozoen axonemes. Ultrastruct. Res. 82, 212—220.


HEIDEMANN, S. R. & GALLAS, P. T. (1980). The effect of taxol on living eggs of Xenopus laevis.Devi Biol. 80, 489-494.

KINGS, S. M. & HYAMS, J. S. (1983). Analysis of anaphase B in Saccharvmyces cerevisiae usinga monoclonal antibody against yeast tubulin. Eur.J. Cell Biol. 29, 121-125.

LA FOUNTAIN, J. R. JR & THOMAS, H. R. (1975). The infrastructure of spindle microtubules afterfreeze-etching and negative staining in situ. J. Ultrastruct. Res. 51, 340-347.

LEDBETTER, M. C. & PORTER, K. R. (1964). Morphology of microtubules of plant cells. Science144, 872-874.

LINCK, R. W- (1976). Flagellar doublet microtubules: fractionation of minor components and-tubulin from specific regions of the A-tubule.J. Cell Sci. 20, 405-439.

MASUROVSKY, E. B., PETERSON, E. R., CRAIN, S. M. & HORWITZ, S. B. (1982). Taxol-induced

microtubule formations in fibroblasts of fetal mouse dorsal root ganglion-spinal cord cultures.Biol. Cell 46, 213-216.

MOLE-BAJER, J. & BAJER, A. S. (1983). Action of taxol on mitosis: modification of microtubulearrangements and function of the mitotic spindle in Haemanthus endosperm. J. Cell Biol. 96,527-540.

NAGANO, T. & SUZUKI, F. (1975). Microtubules with 15 subunits in cockroach epidermal cells. J.Cell Biol. 64,242-245.

SAITO, K. & HAMA, K. (1982). Structural diversity of microtubules in the supporting cells of thesensory epithelium of guinea pig organ of Corti. J. Electron Microsc. 31, 278-281.

SAKAI, H., MABUCHI, I., SHIMODA, S., KURIYAMA, R., OGAWA, K. & MOHRI, H. (1976).

Induction of chromosome motion in the glycerol isolated mitotic apparatus: nucleotide specificityand effects of antidynein and myosin sera on the motion. Dev. Growth & Differ. 18, 211-219.

SATO, H. (1982). Role of spindle microtubules for the anaphase chromosome movements in fer-tilized sea urchin eggs. Cell Differ. 11, 345-348.

ScHATTEN, G., SCHATTEN, H., BESTOR, T. H. & BALCZON, R. (1982). Taxol inhibits the nuclearmovements during fertilization and induces asters in unfertilized sea urchin eggs. J. Cell Biol. 94,455-465.

SCHEELE, R. B., BERGEN, L. G. & BORISY, G. G. (1982). Control of the structural fidelity ofmicrotubules by initiation sites. J . molec. Biol. 154, 485-500.

SCHIFF, P. B. & HORWITZ, S. B. (1980). Taxol stabilizes microtubules in mouse fibroblast cells.Proc. natn. Acad. Sci. U.SA. 77, 1561-1565.

SIMONE, L. D., BRENNER, S. L., WIBLE, L. J., TURNER, D. S., BRINKLEY, B. R. (1981). Taxol-induced microtubule initiation and assembly in mammalian cells. J. Cell Biol. 91 (2, Pt 2): 337a(abstr.).

TAKAHASHI, T. C. & SATO, H. (1982). Thermodynamic analysis of the effect of D2O on mitoticspindles in developing sea urchin eggs. Cell Struct. Funct. 7, 349—357.

TILNEY, L. G., BRYAN, J., BUSH, D. J., FUJIWARA, K., MOOSEKER, M. S., MURPHY, D. B. &SNYDER, D. H. (1973). Microtubules: evidence for 13 protofilaments. J. Cell Biol. 59, 267-275.

(Received 24 August 1984 -Accepted, in revised form, 8 February 1985)

Documents

SPATIOTEMPORAL CONTROL OF FUNCTIONAL SPECIFICATION … · Nyctotherus ovalis. This may show whether mt wit 1h3 p mor become thae asn-sembled onto pre-existing tubules or are newly