Effects of indomethacin on the divisional morphogenesis and cytoskeleton-dependent processes of Tetrahymena

cell biochemistry and function

Cell Biochem Funct 2003; 21: 169–175.

Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/cbf.999

Effects of indomethacin on the divisional morphogenesis andcytoskeleton-dependent processes of Tetrahymena

Peter Kovacs* and Eva Pallinger

Department of Genetics, Cell- and Immunobiology, Semmelweis University, Budapest, Hungary

Indomethacin (0.1 mM) causes significantly altered phospholipid synthesis in Tetrahymena and is able to influence theinositol phospholipid signalling system.9 In the present study the effects of indomethacin on the course of cell division,cyclin expression, the cortical microtubular system and on cytoskeleton-dependent processes (motility, phagocytosis) wereinvestigated. As expected from its interference with the synthesis of phospholipids, indomethacin affected Tetrahymena in anumber of ways: the structure of the cortical microtubular system became irregular; in many cells the stomatogenesis(development of new oral apparatus) and the development of the fission furrow was not accompanied by elongation ofthe macronucleus, which is a typical phenomenon of the normal course of mitosis: apparently indomethacin uncouplesthese phenomena. After indomethacin treatment, the expression of both cyclin A and cyclin B1 were reduced signifi-cantly. The cell growth rate, motility and phagocytotic activity were all considerably reduced. There are probably additionalmechanisms responsible for the effect of indomethacin on the systems that control divisional morphogenesis, formicrotubule-dependent processes and for the connection between nuclear and cortical alterations during the cell cycle.Effects on protein phosphorylation/dephosphorylation, on cyclin expression and on microtubular functions are probablyinvolved. These possibilities are discussed. Copyright # 2003 John Wiley & Sons, Ltd.

key words—Tetrahymena; cell cycle; cytoskeleton; motility; phagocytosis; signalling

INTRODUCTION

The course of divisional morphogenesis in Tetrahy-mena may be divided into two clearly visible majorevents: (a) alterations in the cytoskeleton, accompany-ing the development of a new oral apparatus (stoma-togenesis) and the division furrow; and (b) theelongation and division of the macronucleus. The cellcycle in Tetrahymena may consist of separate clustersof events, with a varying degree of coupling betweenthe clusters.

In Tetrahymena, the moment of the change of pat-tern in cortical components was timed to the meta-phase/anaphase transition of the mitotic cycle of themicronucleus.1 The macronuclear elongation occursearly in cytokinesis at the first sign of a fission furrow.Usually, these phenomena occur in a coordinated

fashion; common developmental signals may coordi-nate all these changes. In Tetrahymena, the macronuc-lear DNA synthesis can be readily dissociated fromthe cortical development, i.e. it is not initiated at anyspecific phase of the macronuclear DNA replicationcycle.2

Macronuclear division in Tetrahymena is an active,internally-directed process, which is normally coordi-nated with cell division.3 When colchicine is given toTetrahymena populations just before macronuclearelongation begins, the intranuclear microtubules dis-appear, yet most macronuclei elongate and divideendogenously at the normal time.3 Some of thesemacronuclei, however, fail to elongate normally andappear to be sliced in two by the division furrow atthe end of cytokinesis.4 These observations suggestthat cytokinesis can take place even when the endo-genous macronuclear division is prevented.

In higher animals, mitosis involves proteinphosphorylation/dephosphorylation events, includ-ing phosphoproteins present in the microtubular-organizing centre.5 The interference of some drugs

Received 25 March 2002Copyright # 2003 John Wiley & Sons, Ltd. Accepted 23 May 2002

* Correspondence to: Peter Kovacs, 1445 Budapest, Nagyvarad ter4, POB 370, Hungary. Tel: (36-1) 210-2930. Fax: (36-1) 303-6968.E-mail: [email protected]

with nuclear elongation in Tetrahymena is likely to bedue to a change in the dynamic state of microtubulesand their impaired interaction with microtubule-associated proteins. An additional possibility is thedisturbance of the normal function of the cell-cycleregulatory proteins such as cyclins and cyclin-dependent kinases (CDKs). The eukaryotic cell cycleis regulated by the sequential activation of differentCDK/cyclin complexes. Although cell cycle regu-lation in ciliates differs in some respects fromthat of other eukaryotes, the cyclin motifs haveclearly been conserved during evolution. Parameciumcyclin homologues Cyc1 and Cyc2 display 42–51%sequence identity to other eukaryotic mitotic cyclinswithin the cyclin box region.6

Indomethacin is an inhibitor of prostaglandin (PG)synthesis by inhibiting the cyclooxygenase pathwayfor arachidonate metabolism. In addition, indometha-cin interferes with the metabolism of inositol phos-pholipids. The alteration in the synthesis of theselipid-derived second messengers with subsequentactivation/inactivation of protein phosphorylation cas-cades is an important link between the signalling andthe dynamics of the cytoskleton. Under in vitro condi-tions, the other lipid-derived second messengers, thePGs, are important in the organization and stabilityof the actin cytoskeleton.7 Thus indomethacin is pre-sumably able to infuence the function of the cytoske-letal system in several ways. Another effect ofindomethacin on the systems responsible for signal-ling, is the inhibition of cyclin expression and retino-blastoma protein phosphorylation.8 These effects arepotentially important in the regulation of the cellcycle, and in the normal course of mitosis.

In our previous experiments, indomethacin in Tet-rahymena totally inhibited the incorporation of 32Pinto phospholipids at a dose of 2.0 or 5.0 mM, andthese treatments caused some signs of cellular damage(e.g. rounding off), while treatment with 0.1 mM indo-methacin gave rise to significantly altered phospholi-pid synthesis—without spectacular morphologicalalterations;9 for this reason, in the present experimentsmainly this latter concentration was used. Wehypothesized that these effects of indomethacin onthe signalling systems and the cytoskeleton are alsoable to influence the normal course of divisional mor-phogenesis and cytoskeleton-dependent processes inTetrahymena. Thus, in the present study, the effectof indomethacin on the course of mitosis, on the cor-tical microtubular system, on cyclin expression, ongrowth, and on skeleton-dependent processes (motility,phagocytosis) were investigated. By studying these

parameters we hoped to learn more about the signal-ling system of unicellular Tetrahymena.

MATERIALS AND METHODS

Chemicals

Indomethacin, monoclonal (mouse ascites) anti-acetylated tubulin antibody, anti-cyclin A (mouse)and anti-cyclin B1 (rabbit) antibodies, FITC-labelledanti-mouse and rabbit goat IgG, TRITC-labelledConA (concanavalin A) and daunorubicin wereobtained from Sigma (St Louis, MO, USA). All otherchemicals used were of analytical grade availablefrom commercial sources.

Tetrahymena cultures

In the experiments, a-micronucleated T. pyriformis GLstrain was tested in the logarithmic phase of growth.The cells were cultivated at 28�C in 0.1% yeast extractcontaining 1% tryptone medium. Before the experi-ments, the asynchronous cell populations were washedwith fresh culture medium and were resuspended at aconcentration of 5� 104 cells ml�1.

Confocal scanning laser microscopic (CSLM)analysis of Tetrahymena cells labelled withmonoclonal anti-acetylated tubulin

To localize tubulin-containing structures, indometha-cin (0.1 mM; 1 h)—treated and untreated (control)cells were fixed in 4% paraformaldehyde dissolvedin PBS, pH 7.2. After washing with wash buffer(WB; 0.1% BSA in 20 mM Tris-HCl, 0.9% NaCl,0.05% Tween 20, pH 8.2) the cells were incubatedwith monoclonal anti-acetylated tubulin antibodydiluted 1 : 500 with antibody buffer (AB; 1% BSAin 20 mM Tris-HCl, 0.9% NaCl, 0.05% Tween 20,pH 8.2) for 45 min at room temperature. After threewashings with WB, the anti-tubulin antibody-treatedcells were incubated with FITC-labelled anti-mousegoat IgG (diluted to 1 : 500 with AB) for 45 min atroom temperature. Then, the cells were washed twicewith WB, and treated with 0.01 mg ml�1 daunorubicinfor 10 min in order to localize the macronucleus. Afterthe daunorubicin treatments, the cells were washedfour times and mounted onto microscope slides. Thedouble labelled and mounted cells were analysed ina Bio-Rad MRC 1024 confocal scanning laser micro-scope (CSLM) equipped with a krypton/argon mixedgas laser as a light source. Excitation was carried out

170 p. kovacs and e. pallinger

Copyright # 2003 John Wiley & Sons, Ltd. Cell Biochem Funct 2003; 21: 169–175.

with the 480 nm line (for FITC) and 530 nm (for dau-norubicin) from the laser.

Flow-cytometric (FACS) analysis of the bindingof cyclin A and cyclin B1 to Tetrahymena

Tetrahymena populations were treated with 0.1 or0.25 mM indomethacin for 60 min at room tempera-ture. Untreated cells served as controls. After treat-ments, the cells were fixed in 4% paraformaldehydedissolved in PBS, pH 7.2. After washing with WB,the cells were incubated with monoclonal anti-cyclinA or cyclin B1 antibody diluted 1 : 300 with AB for45 min at room temperature. After three washingswith WB, the cells were incubated with FITC-labelledanti-mouse IgG (for cyclin A) or FITC-labelled anti-rabbit IgG (for cyclin B1) for 45 min at room tempera-ture. After these incubations, the cells were washedfour times with WB. To determine the nonspecificbinding of second (labelled) antibodies the primaryantibodies were omitted as absolute controls. Themeasurement of fluorescence intensity was carriedout in a FACS Calibur flow cytometer (BecktonDickinson, San Jose, USA), using 104 cells for eachmeasurement. For the measurement and analysis aCellQuest 3.1 program was used. During the evalua-tion, cell populations had been separated on the basisof size defined by ‘gating’. In the identical cell popu-lations, the FITC-labelled second antibody contentinside the cells was measured. The evaluation of theresults was done by comparison of percentagechanges of geometric mean channel values to the con-trol groups. Each experiment was repeated threetimes.

Analysis of phagocytotic activity

Tetrahymena pupulations were treated with 0.1 mM

indomethacin in the presence of TRITC-labelledCon A (0.01 mg ml�1, dissolved in fresh culture med-ium). Cells were taken after 1, 5, 15 and 30 min, fixedin 4% paraformaldehyde solution (in pH 7.2 PBS),washed in two changes of PBS, and the number offluorescent food vacuoles was quantified by using aZeiss Fluoval fluorescent microscope.

Effect of indomethacin on the growth rateof Tetrahymena populations

Tetrahymena cultures were washed and resuspendedin fresh culture medium to a final cell density of about2500 cells ml�1. This population was divided into two

experimental groups (in 20 ml): (a) control group; (b)0.1 mM indomethacin—treated group. The cells werecounted in fixed samples (0.5 ml) using a Fuchs–Rosenthal counting chamber at 0, 9, 24, 33 and 48 hof cultivation.

Recording of cell motility

Cell motility was recorded—according to Dryl’s ori-ginal method10—during a 1-s exposure at low micro-scopic power (250�magnification) at 0, 15, 30, 45and 60 min of treatment. After photographic enlarge-ment, cell traces were measured. In each sample, 100cells were measured.

Statistical treatment of the data

Student’s t-test was used for the evaluation of all data,with p< 0.05 accepted as the level of statistical signif-icance.

RESULTS

After 1 h indomethacin (0.1 mM) treatment of theTetrahymena strain used by us, the array of thelongitudinal and transverse microtubule bundlesbecame irregular in some parts of the cytoskeleton;these alterations were clearly visible on the CSLMpictures after immunocytochemical staining of tubulin-containing structures (Figure 1). These microtubularirregularities we found in about 15–20% of the treatedcells; in controls, no such alterations were observed.

According to the CSLM pictures of indomethacin-treated Tetrahymena, the appearance of the new oralapparatus and the development of the fission furrowwas not accompanied by the elongation of the macro-nucleus which is typical of normal mitosis. After thedevelopment of a new oral apparatus and fission fur-row, the macronucleus remained spherical—similarto the interphase nucleus (Figure 2).

Treatment with 0.1 mM indomethacin significantlyhindered the growth rate of Tetrahymena populations(Figure 3). These treatments also reduced the phago-cytotic activity of the cells (Figure 4); in the treatedcells we found food vacuoles in only about 40% ofthe cells, and the number of the food vacuoles in thesecells was reduced, when compared to the controls.The proportion of cells capable of forming foodvacuoles during the time of the experiment remainedalmost constant. In addition to the reduction in thenumber of food vacuoles, the size of these vacuolesalso became smaller in the treated cells.

effects of indomethacin on tetrahymena 171


The motility of cells decreased in culture mediumcontaining 0.1 mM indomethacin; with time the trea-ted cells swam more slowly, and an increasing numberof cells became immobile (Figure 5).

Treatment of Tetrahymena cells with 0.1 and0.25 mM indomethacin for 60 min reduced the expres-sion of cyclin A and cyclin B1 significantly in a con-centration-dependent manner, measured by FACS ofthe binding of anti-cyclin A and cyclin B1 antibodiesto Tetrahymena (Figure 6). In the case of 0.25 mM

indomethacin, the intensity of fluorescence was closeto the absolute control, where no anti-cyclin antibo-dies (primary antibodies) were used.

DISCUSSION

As expected from its interference with the synthesis ofphospholipids,9 indomethacin (0.1 mM) affected Tet-rahymena in a number of ways.

The structure of the microtubular system (longitu-dinal and transverse microtubule bands) in cellsexposed to 0.1 mM indomethacin became irregular,something which was not seen in control cells. In aconsiderable number of cells, stomatogenesis andthe development of a division furrow were seen, butthe macronucleus did not elongate in these cells. Fromthese observations we concluded that indomethacinimpaired the signalling system which is able possiblyto connect the events of divisional morphogenesis.Apparently, indomethacin impaired the coordinationbetween the cytoskeletal and nuclear processes.

The first visible event is the duplication of ciliarybasal bodies necessary for the formation of newmouth ciliature. The basal bodies which begin to mul-tiply form an extensive ‘anarchic field’ and the basalbodies become grouped to form three membranellesand the undulating membrane of the new oral appara-tus. The oral primordia appear at or near theirexpected positions, in which the surface cytoskeletalarray is highly disorganized, so that there are oftenunrecognizable distinct ciliary rows.11 Indomethacintreatment produces in a large proportion of cells, dis-organized areas similar to the ‘anarchic’ field. It isvery likely that from this field the new oral primordiaevolve, but in these cells nuclear elongation does notoccur.

Indomethacin significantly inhibits cyclin expres-sion in endothelial cells and reduces basal andbFGF-stimulated proliferation.8 Mitotic cyclin homo-logues can be found also in ciliates, such as Parame-cium, Tetrahymena, Colpoda and Blepharisma. Thesecyclins exhibit histone H1 kinase activity and may reg-ulate cytokinesis.6 According to our FACS results,indomethacin inhibits cyclin expression in Tetrahy-mena, and this phenomenon indicates that divisionprocesses (macronuclear elongation/division, cytokin-esis) are presumably coordinated (at least in part) bycyclins.

Another possibility for the cell cycle arrest is thatindomethacin affects assembly and/or the functionof some microtubule-based stuctures/processes in Tet-rahymena. In a temperature-sensitive mutant Tetrahy-mena, cilia regeneration is blocked at the restrictivetemperature. At this temperature, these cells arearrested at a specific point in the cell cycle, i.e. aftermacronuclear S phase.12 In accordance with ourpresent findings, it is possible that indomethacintreatment resulted in similar effects on the microtu-bular system, and impaired microtubular functions.Tubulins are known to undergo a wide variety ofpost-translational modifications e.g. glycosylation,acetylation.13 These modifications are required for

Figure 1. Binding of anti-acetylated tubulin antibody to theTetrahymena. Confocal micrographs of control (a) and 0.1 mM

indomethacin-treated cells (b, c and d). *Oral apparatus;bar¼ 5 mm. In the control cells the longitudinal microtubule bundlesrun in parallel; in treated cells, the irregularities of longitudinal andtransverse microtubule bundles are clearly visible



the function rather than the assembly of microtubules.Thus functions affected by decrease of polyglycosyla-tion include ciliary motility, growth, and the locationas well as execution of cytokinesis.14 Similar phenom-ena were found in our indomethacin-treated Tetrahy-

mena populations, and so we cannot rule out aneffect of indomethacin on these post-transcriptionalmodifications.

It is difficult to determine whether indomethacinaffects the organization of cortical microtubules in a

Figure 2. Confocal micrographs of control (a–d) and 0.1 mM indomethacin-treated (e–j) Tetrahymena. Binding of anti-acetylated tubulin tothe microtubular structures, and daunorubicin to the macronuclei (*). Arrow heads indicate the oral apparatus. In the control (untreated)cells the skeletal and nuclear events run in parallel during the cell cycle. In the indomethacin-treated cells these phenomena aredisconnected: the skeletal alterations are not accompanied by nuclear elongation. e and f; g and h; i and j are pictures of the same cells. e, gand i are optical sections from the surface the cells; f, h and j are optical sections from the middle of the cells. a–d¼ �1300; e–j¼ �1100magnification



more direct way which leads to the different effectsobserved by us, or whether microtubular defects andthe arrested/disturbed nuclear division (impairedcell cycle) are secondary effects of an impairedphospholipid signalling system, of altered proteinphosphorylation/dephosphorylation, or a consequenceof reduced food uptake. The reduced motility inindomethacin-treated Tetrahymena may indicate thecytoskeleton-disturbing effect of indomethacin, oran inhibitory effect of indomethacin on food uptake,and thus general metabolism. Therefore, an additional

explanation for the disadvantageous effect of indo-methacin is the reduced phagocytotic activity: reducedfood uptake is able to affect many other cellularprocesses. The undernourished cells are oftenincapable of crossing the G1 phase check point inthe direction of the S phase.

Sodium orthovanadate treatment affected some cellfunctions in Tetrahymena similarly to indomethacin:in vanadate-treated cell populations the motility,growth, phagocytosis and macronuclear elongationwere inhibited. This effect of vanadate is most likely

Figure 3. The effect of indomethacin treatment on the growth rateof Tetrahymena populations. The experiments were carried out intriplicate with a representative experiment shown

Figure 4. The effect of indomethacin treatments on food vacuoleformation of Tetrahymena, measured by the binding of TRITC-labelled ConA to the food vacuoles. The data represent the mean(� SD) derived from three separate experiments

Figure 5. Effect of indomethacin on the swimming speed(motility) of Tetrahymena expressed as a percentage of the control.Means of three independent experiments

Figure 6. FACS analysis of effect of indomethacin on theexpression of cyclin A and cyclin B1 of an asynchronizedTetrahymena population *p< 0.01 relative to the controls. The datarepresent the mean (�SD) derived from three separate experiments.Abs. Control shows the results obtained in the absence of theprimary (anti-cyclin) antibodies



related to its known interference with phosphorylationevents.15 In our previous experiments we found thatvanadate is able to stimulate phospholipase D activityin Tetrahymena; presumably this effect is based on itseffect on the tyrosine kinase activity.16 Vanadate treat-ment, similar to indomethacin, may influence thefunction of a phospholipid-derived signalling system,the outcome of which is revealed by the microtubularsystem and its dependent events. Interference withphosphorylation events may also be responsible forthe impaired cyclin/CDK expression and functions.

Several conclusions can be drawn from our data: (a)there are many structures and events affected by indo-methacin. Thus it is very difficult to determine the pri-mary effect; (b) the disturbance of the signallingsystem (e.g. inositol phospholipids) is manifest onthe cytoskeleton and skeleton-dependent functions;(c) indomethacin treatment significantly reducescyclin expression, and thereby growth rate and thenormal course of the cell cycle; (d) these effects areable to disconnect cytoskeletal and nuclear events dur-ing mitosis.

ACKNOWLEDGEMENT

This work was supported by the Scientific ResearchCouncil, Ministry of Health, Hungary, grant number:T-II. 213/2000.

REFERENCES

1. Frankel J, Nelsen EM, Martel E. Development of the ciliature ofTetrahymena thermophila. II. Spatial subdivision prior to cyto-kinesis. Dev Biol 1981; 88: 39–54.

2. Antipa GA. A temporal analysis of cell cycle and morphoge-netic events in Tetrahymena pyriformis. Acta Protozool 1980;19: 1–14.

3. Jaeckel-Williams R. Nuclear divisions with reduced numbers ofmicrotubules in Tetrahymena. J Cell Sci 1978; 34: 303–319.

4. Tamura S, Tsuruhara T, Watanabe Y. Function of nuclearmicrotubules in macronuclear division of Tetrahymena pyrifor-mis. Exp Cell Res 1969; 55: 351–358.

5. Vandre DD, Borisy GG. Anaphase onset and dephosphorylationof mitotic phosphoproteins occur concomitantly. J Cell Sci1989; 94: 245–258.

6. Zhang H, Adl SM, Berger JD. Two distint classes of mitoticcyclin homologues, Cyc1 and Cyc2, are involved in cell cycleregulation in the ciliate. Paramecium tetraurelia. J Euk Micro-biol 1999; 46: 585–596.

7. Coutant KD, Corvaia N, Ryder NS. Bradykinin induces actinreorganization and enhances cell motility in HaCaT keratino-cytes. Biochem Biophys Res Commun 1997; 237: 257–261.

8. Pai R, Szabo IL, Kawanaka H, Soreghan BA, Jones MK,Tarnawski AS. Indomethacin inhibits endothelial cell prolifera-tion by suppressing cell cycle proteins and PRB phosphoryla-tion: a key to its antiangiogenic action? Mol Cell Biol ResCommun 2000; 4: 111–116.

9. Kovacs P, Csaba G. Indomethacin alters phsopholipid andarachidonate metabolism in Tetrahymena pyriformis. CompBiochem Physiol 1997; 117c: 311–315.

10. Dryl S. Photographic registration of movement of protozoa.Bull Ac Pol Sci 1958; 6: 429–432.

11. Jerka-Dziadosz M, Jenkins LM, Nelsen EM, Williams NE.Cellular polarity in ciliates: persistence of global polarity ina disorganized mutant of Tetrahymena thermophila that disruptscytoskeletal organization. Dev Biol 1995; 169: 644–661.

12. Pennock DG, Thatcher T, Gorovsky MA. A temperature-sensitive mutation affecting cilia regeneration, nuclear devel-opment, and the cell cycle of Tetrahymena termophila is rescuedby cytoplasmic exchange. Mol Cell Biol 1988; 8: 2681–2689.

13. Macrae TH. Tubulin post-translational modifications—enzymes and their mechanisms of action. Eur J Biochem1997; 244: 265–278.

14. Xia L, Hai B, Gao Y, et al. Polyglycylation of tubulin is essentialand affects cell motility and division in Tetrahymena thermo-phila. J Cell Biol 2000; 149: 1097–1106.

15. Nilsson JR. Vanadate affects nuclear division and inducesaberrantly-shaped cells during subsequent cytokinesis in Tetra-hymena. J Euk Microbiol 1999; 46: 24–33.

16. Kovacs P, Csaba G, Nakashima S, Nozawa Y. Phospholipase Dactivity in the Tetrahymena pyriformis GL. Cell Biochem Funct1997; 15: 53–60.



Documents

Effects of indomethacin on the divisional morphogenesis and cytoskeleton-dependent processes of Tetrahymena