Aphidicolin induces alterations in Golgi Complex and disorganization of microtubules of HeLa cells...

JOURNAL OF CELLULAR PHYSIOLOGY 176:602–611 (1998)

Aphidicolin Induces Alterations inGolgi Complex and Disorganization of

Microtubules of HeLa Cells UponLong-Term Administration

HIROYUKI TANAKA,1 HAJIME TAKENAKA,1 FUMIAKI YAMAO,2

AND TATSUO YAGURA1*1Laboratory of Life Science, Department of Chemistry, Faculty of Science,

Kwansei Gakuin University, Nishinomiya-shi, Hyogo-ken, Japan2National Institute of Genetics, Mishima, Shizuoka-ken, Japan

Treatment of HeLa cells with aphidicolin at 5 or 0.5 mg/ml induced cell cyclearrest at G1/S or G2/M phase, respectively, and was accompanied by unbalancedcell growth. Long-term administration of aphidicolin (more than 48 h) resultedin noticeable loss of reproductive capacity though cells were viable at the time oftreatment. Immunofluorescence with anti-Golgi membrane protein monoclonalantibody (mAbG3A5) showed disfigurement of the characteristic mesh-like con-figuration when cells were treated for more than 48 h. Interestingly, we foundthat the fragmented Golgi complex formed a ring around the nucleus in morethan 20% of the cells. Immunoelectron microscopy using mAbG3A5 antibodydemonstrated that the stack structure of the fragmented Golgi complex in aphidi-colin-arrested cells appeared partially broken up and seemed to have converted toa vesicle-like structure. Analysis using an antibody to tubulin and anticentrosomehuman autoimmune serum showed that alterations in the Golgi complex wereinduced even by the lower 0.5 mg/ml dose. These alterations were accompaniedby both changes in the distribution of microtubules and an increase in the numberof centrosomes. These cells lost their distinct perinuclear microtubule organiz-ing center (MTOC). On the other hand, treatment with aphidicolin at 5 mg/mldid not induce multiplication of the centrosome although the loss of distinctMTOC was still evident. No changes took place in the Golgi complex, microtu-bule, or centrosome of cells treated with 0.5 mg/ml aphidicolin when cyclohexi-mide was added simultaneously to the culture. J. Cell. Physiol. 176:602–611,1998. q 1998 Wiley-Liss, Inc.

Aphidicolin is a tetracyclic diterpene tetraol obtained toxic if the cells have a period of inhibition-free growthfollowing removal of the agent (Kung et al., 1990). Fur-from Cephalosporium aphidicola and certain otherthermore, aphidicolin kills human neuroblastoma cellsfungi (Brundret et al., 1972), which inhibits DNA poly-in vitro in a cell-type specific manner during continuousmerase a and hence inhibits DNA synthesis in eukary-administration at concentrations which completely in-otic cells (Ikegami et al., 1978). Aphidicolin does nothibit cell proliferation (Cinatl et al., 1992). In CHOinhibit synthesis of either protein or RNA (Spadari etcells, one half of the cells that transpired after treat-al., 1984) and arrests cell growth at the G1/S phasement with aphidicolin had associated aberrant mitotic(Costa et al., 1992; Sukhorukov et al., 1994). This char-processes (Kung et al., 1990). Thus, cell killing byacteristic of aphidicolin induces unbalanced growth ofaphidicolin may be partially ascribed to disordered mi-cells in prolonged culture, because selective blocking oftotic control in cells with unbalanced growth. However,DNA synthesis results in a cell population havingnothing as yet addresses the fundamental question ofnearly unchanging DNA content while both protein and

RNA continue to increase (Costa et al., 1992; Cinatlet al., 1992; Goswami et al., 1992; Sukhorukov et al., Contract grant sponsor: Ministry of Posts and Telecommunica-1994). tions (Frontier Research Project); Contract grant sponsor: Sup-

port Center for Advanced Telecommunications Technology Re-The effect of aphidicolin on DNA replication is revers-search.ible (Iliakis et al., 1982; Tang et al., 1992), thus short-*Correspondence to: Tatsuo Yagura, Laboratory of Life Science,term administration of aphidicolin is used to obtain cellDepartment of Chemistry, Faculty of Science, Kwansei Gakuinsynchronization (Spadari et al., 1984). On the otherUniversity, Nishinomiya-shi, Hyogo 662, Japan.hand, agents including aphidicolin, which act in a cell

cycle phase-specific manner, are reported to be cyto- Received 12 January 1998; Accepted 3 March 1998

q 1998 WILEY-LISS, INC.

893B 10313/ 893f$$0313 06-24-98 01:14:06 wlcpa W Liss: JCP

PERTURBATION IN GOLGI BY APHIDICOLIN 603

Fig. 1. DNA histograms of HeLa cells grown in the presence of Cells were analyzed by flow cytometry, which revealed the distribu-aphidicolin or aphidicolin/cycloheximide. Asynchronously dividing tion of nuclei in cell cycle phase after propidium iodide staining. TheHeLa cells were mock treated (A–C) or exposed to 0.5 mg/ml (D–F), sharp peak on the left in A is the G1 peak (2n) and the subsequent5 mg/ml (G–I) aphidicolin or aphidicolin (0.5 mg/ml )/ cycloheximide small peak is the G2/M peak (4n). The results depicted are those of(10 mg/ml) (J–L). A,D,G,J: Culture at t Å 24 h; B,E,H,K: t Å 48 h; one representative experiment.C,F,I,L: t Å 72 h. The drugs were added to the cultures at t Å 0 h.

why mitotic control falls into disorder after treatment using antibodies directed against the Golgi cisternalmembrane glycoprotein p138 (Yamauchi et al., 1992)with aphidicolin or why cell killing during aphidicolin

administration in culture occurs in certain human cell to detect changes in the Golgi complex. The reason weaimed at the Golgi complex is that the localization andstrains (Cinatl et al., 1992). In the area of chemothera-

peutics, many agents which inhibit specific cellular pro- morphology of the Golgi complex in cultured cells arewell documented as being altered by drugs or culturecesses in proliferating cells are used to combat cancer.

Recently, some of the mechanisms of cell killing have conditions (Fujiwara et al., 1988; Matlin et al., 1988;Turner and Tartakoff, 1989; Lucocq, 1991; Jantti andbeen discussed in relation to apoptosis (Ishida et al.,

1992). However, detailed physiological and biochemical Kuismanen, 1993). Antibody mAbG3A5 made possiblethe detection of membrane vesicles derived from theeffects of drugs acting in a cell cycle-specific manner,

such as aphidicolin, are not well understood. Golgi complex during the mitotic phase (Asada and Ya-gura, 1995) as well as Golgi stack cisternae fragmentedThe purpose of this study is to examine how the cellu-

lar physiology of cancer cells changes during drug during taxol treatment (Hoshino et al., 1997). Usingthis antibody, we demonstrated that a prolonged ad-treatment which inhibits a specific cellular process in

a cell cycle-specific manner and the consequences in- ministration of aphidicolin to HeLa cells resulted inalterations to both the localization and morphology ofcurred by these changes. In an attempt to pinpoint

these physiological and biochemical changes, we tested the Golgi complex, accompanied by loss of reproductivecapacity. Evidence is provided to show that these alter-aphidicolin employing an immunocytochemical approach


TANAKA ET AL.604

Fig. 2. Cell viability during aphidicolin treatment and loss of repro-ductive capacity after release from aphidicolin arrest. HeLa cellsgrown in the presence of 0.5 mg/ml (filled circle, filled triangle) and 5mg/ml (open circle, open triangle) aphidicolin. At the times indicated,cultures were washed with fresh medium. Then the cells were de-tached from plates and processed for analysis for viability and cellplating efficiency. The number of viable cells was estimated immedi-ately after detaching the cells by staining with ethidium bromide/acridine orange (filled and open circles). The remaining cells werereplated in regular culture medium for determination of cell platingefficiency (filled and open triangles). Cell plating efficiency (definedas the ratio of total colonies formed to the number of single cellsinitially inoculated per plate) was generally 0.6–0.7 for mock-treatedHeLa cells (100% value in the figure).

Fig. 3. Classification of types of Golgi complex observed in aphidi-colin-arrested HeLa cells. Log phase culture of HeLa cells was treatedwith aphidicolin at 0.5 mg/ml and cultured for 72 h. Immunofluores-cence microscopy using the anti-Golgi membrane glycoprotein,

ations can be ascribed to the disorganization of microtu- mAbG3A5 (A–D), was performed as described under Materials andMethods. E–H: DNA stained with Hoechst 33258 in correspondingbules in cells during unbalanced growth. These resultscells. The Golgi complex was classified into four representative typeselucidate part of the reason why long-term administra- according to fluorescence patterns (see text for the characterizationtion of aphidicolin causes nonproliferation in HeLa cells of each type). A, E: type 1; B, F: type 2; C, G: type 3; D, H: type 4.

even after removal of aphidicolin. Bar, 10 mm.

MATERIALS AND METHODSReagents

mouse IgG antibody was from DAKO A/S (Glostrup,Aphidicolin was kindly supplied by Dr. B. Hesp (Im- Denmark). Rhodamine-conjugated donkey anti-rabbit

perial Chemical Industries., Ltd., London, UK) and IgG antibody and rhodamine-conjugated goat anti-hu-taxol was obtained from Wako Chemical Co., Osaka, man immunoglobulin antibody were from Chemicon In-Japan. Aphidicolin and taxol were dissolved in a vol- ternational Inc. (Temecula, CA) and Cappel (Durham,ume of dimethyl sulfoxide (Merck, Rahway, NJ) to give NC), respectively. Horseradish peroxidase- conjugateda final concentration of dimethyl sulfoxide of less than rabbit anti-mouse antibody was from DAKO A/S. Auto-0.1% of cell culture medium. Cycloheximide (Sigma, St. immune human serum reactive against centrosomeLouis, MO) was dissolved in sterilized distilled water was kindly supplied by Medical & Biological Labora-at 10 mg/ml. tories Co., Ltd., Nagoya, Japan.

Antibodies Cell cultureHeLa cells were routinely maintained in Eagle’s min-The hybridoma cell line (mAbG3A5) which secreted

monoclonal antibody against the p138 Golgi membrane imum essential medium supplemented with 10% fetalcalf serum in glass culture dishes.protein of human origin was prepared as described pre-

viously (Yamauchi et al., 1992) and purified by chroma-Immunofluorescence microscopytography on columns of affi-Gel protein A. Rabbit anti-

body to tubulin was obtained from ICN Biomedicals, Cells were prepared by methods described previously(Asada and Yagura, 1995). Briefly, cells (3 1 104 orInc., Costa Mesa, CA. FITC-conjugated rabbit anti-



fewer) were plated onto glass coverslips and placed in-side 50 1 10 mm Falcon plastic Petri dishes 1 day beforedrug administration. The number of cells plated ontothe dishes was carefully controlled to avoid semicon-fluent to confluent cultures which showed no changesin the Golgi complex during drug administration. Cov-erslip cultures were fixed with 5% paraformaldehyde(pH 7.0) in phosphate-buffered saline and processed forimmunocytochemical analysis as described previously(Asada and Yagura, 1995). Specimens were observedand photographed with an Olympus fluorescence mi-croscope.

Immunoperoxidase electron microscopyImmunoelectron microscopy using mAbG3A5 anti-

body was performed as described previously (Asada andYagura, 1995). Ultrathin sections were examined witha JEOL JEM-100S electron microscope.

Flow cytometry and stainingSamples were stained by addition of 100 ml propid-

ium iodide solution (500 mg/ml) to cells immediatelybefore analysis with a FACScan (Becton Dickinson[BD], Sunnyvale, CA) using an argon ion laser. Dataacquisition and analysis were done with Lysis II soft-ware (BD). Fluorescence signals were recorded andstored in listmode data files; for each assay 40,000 cellswere used.

Analyses of cell viability and cellplating efficiency

Cultured cells exposed to aphidicolin were detachedfrom the Petri dishes using 0.5 mM EDTA in phos-phate-buffered saline (PBS) and resuspended in Eagle’sminimum essential medium supplemented with 10%fetal calf serum. The cells were then plated on 90 1 10mm glass dishes (1,000 or 500 cells per dish) with 20ml of Eagle’s minimum essential medium supple-mented with 10% fetal calf serum. After a 1-week cul-ture, colonies were fixed/stained with 0.5% crystal vio-let in 20% ethyl alcohol. Cell plating efficiency was thendetermined for each culture.

To determine the number of viable cells after expo-sure to aphidicolin, cells were detached with EDTA so-lution and washed with Eagle’s minimum essential me-dium supplemented with 10% fetal calf serum. Cellsresuspended in the medium were mixed with PBS con-taining acridine orange (0.1 mg/ml) and ethidium bro-mide (0.1 mg/ml). The number of viable cells was deter-mined with a fluorescence microscope.

RESULTSAphidicolin arrests HeLa cells at G1/S or

G2/M phase depending on doseIt is known that aphidicolin arrests cells at the G1/S

phase by inhibiting DNA replication. As shown in Fig-

Fig. 4. Changes in the ratio of each type of Golgi complex duringculture in the presence of 0.05 mg/ml (A), 0.5 mg/ml (B), or 5 mg/ml (C)aphidicolin. Aphidicolin was added to the log phase of cultured HeLacells on coverslips at the concentrations indicated. After the timeindicated during aphidicolin arrest, the coverslips were withdrawnand cells were fixed/stained as described under Materials and Meth-ods using the anti-Golgi antibody mAbG3A5. The ratio of each typeof Golgi complex was determined. Classification of each type was ac-cording to that shown in Figure 3. Filled square, type 1; filled circle,type 2; filled triangle, type 3; open circle, type 4.


TANAKA ET AL.606

Fig. 6. Golgi stack cisternae convert to vesicles in aphidicolin-ar-Fig. 5. Ultrastructure of Golgi complex fragmented during culturerested HeLa cells. HeLa cells were treated with 0.5 mg/ml aphidicolinin the presence of aphidicolin. HeLa cells were treated with 0.5 mg/for 72 h and processed for immunoelectron microscopy usingml aphidicolin for 72 h and processed for immunoelectron microscopymAbG3A5 antibody as described under Materials and Methods. Typeusing mAbG3A5 antibody as described in Materials and Methods.3 Golgi complex is shown. A portion of the Golgi stack structure re-Whole images of the dispersed Golgi complex indicate that the imagesmained (arrows) but most others lost their stack structure and con-depicted represent type 3 Golgi complexes. The resolution of organelleverted to vesicles (some vesicles are indicated by arrowheads). Bar,structures is poor in the immunoelectron micrograph. This is due to0.1 mm.unavoidable saponin treatment before fixation, because mAbG3A5

antibody does not stain the Golgi complex if there is no extensivepermeabilization. N, nucleus. Bar, 0.1 mm.

marker using the monoclonal antibody against Golgimembrane glycoprotein p138 (mAbG3A5) (Yamauchiet al., 1992; Asada and Yagura, 1995). After 72 h

ure 1 (G-I), flow cytometry indicates that aphidicolin aphidicolin treatment at 0.5 mg/ml, we found four mainat 5 mg/ml indeed arrested HeLa cells at G1/S phase. types of Golgi complex in the cells. Figure 3 showsBy contrast, over a 48-h period, aphidicolin at 0.5 mg/ these. Type 1 is a normal Golgi complex; in type 2, theml slowed down the cell cycle progression and accumu- Golgi complex disperses as fragments lose their typicallated in cells at G2/M phase (Fig. 1E). At both doses, mesh structure but remains near the original microtu-the number of cells of DNA content differed from the bule organizing center (MTOC) region; in type 3, theusual 2n and 4n increase after 72 h treatment, indicat- Golgi complex moves near the nuclear periphery or dis-ing that progressive distortion in DNA metabolism took perses into the cytoplasm; and in type 4, the frag-place after long-term administration of aphidicolin mented Golgi complex surrounds the entire nucleus.(Fig. 1F, I). Both doses of aphidicolin induced unbal- The peculiar distribution of type 4 Golgi complex wasanced growth and an increase in cell volume. On the quite distinct from that observed in HeLa cells with noother hand, when cycloheximide was administered drug treatment where the Golgi complex localized to awith aphidicolin, neither distorted DNA metabolism restricted perinuclear region.nor did cell cycle arrest take place (Fig. 1J-L). As shown Figure 4 shows the time course of changes of eachlater, no unbalanced growth was observed in these type of Golgi complex during aphidicolin treatment. Atcells. 0.05 mg/ml aphidicolin, the ratio of Golgi complex types

As shown in Figure 2, the proportion of viable cells 2 and 3 increased with time, but type 4 was not ob-remained near control levels throughout the 72-h served even after 72 h culture (Fig. 4A). On the othercourse of aphidicolin treatment at both 5 and 0.5 mg/ hand, the ratio of type 1 cells declined to 0 or a veryml concentrations in spite of obvious distortion of DNA low percentage after 72 h culture in the presence of bothmetabolism in these cell populations. However, a time- 0.5 mg/ml (Fig. 4B) and 5 mg/ml (Fig. 4C) aphidicolin. Independent decrease in cell plating efficiency was ob- contrast, the ratios of types 2 and 3 increased rapidlyserved (Fig. 2). This decrease might not be due to imme- after treatment up to 24 or 48 h at 0.5 or 5 mg/ml aphidi-diate cell death after release from aphidicolin treat- colin concentrations. Type 4 increased markedly afterment since cells exposed to 5 mg/ml aphidicolin for 72 48 h. As shown clearly in Figure 4C, sequential changesh remained viable for at least 48 h by simply changing in each type are evident in cells treated with 5 mg/mlthe culture medium. Thus, the cells were not directly aphidicolin. In mock treatment culture in which corre-killed by exposure to aphidicolin for at least three gen- sponding concentrations of dimethyl sulfoxide wereerations, but rather lost their reproductive capacity added in place of aphidicolin solution, no change wasafter prolonged exposure. observed in the percentage of each type (type 1, ca.80%;

type 2, ca. 20%; types 3 and 4, none) during 72 h ofAlterations in the localization and morphologyculture.of the Golgi complex during culture in the

presence of aphidicolin Ultrastructure of the Golgi complex inaphidicolin-arrested HeLa cellsTo investigate possible physiological changes during

the period of unbalanced growth and the decrease in Figure 5 shows immunoelectron microscopy of 72 hcell culture exposed to 0.5 mg/ml aphidicolin. A portioncell plating efficiency, we chose the Golgi complex as a



Fig. 7. Multipole spindle body of the aphidicolin-arrested mitotic staining (B, D) as described under Materials and Methods. A,B andphase HeLa cells. HeLa cells were treated with 0.5 mg/ml aphidicolin C,D, respectively, show the same fields with different filter sets. Bar,(A,B) or 10 mM taxol (C, D) and processed for immunofluorescence 10 mm.microscopy using antitubulin antibody (A, C) or Hoechst 33258 DNA

of the dispersed Golgi complex type 3 is shown. The Taxol leads to the formation of tubulin bundles, disap-pearance of the representative MTOC, and dispersionGolgi complex was fragmented into many parts, and

although a portion of the fragmented Golgi complex of the Golgi complex (Sandoval et al., 1984; Hoshino etal., 1997). This is accomplished through promotion ofretained its representative stack cisternae structure, it

has disintegrated in many other parts. In Figure 6, four microtubule assembly by lowering the critical concen-tration required for tubulin polymerization (Schiff andcolonies of dispersed portions of Golgi complex illus-

trate the detailed structure of dispersed Golgi cister- Horwitz, 1980). Thus, the alterations in the Golgi com-plex induced by aphidicolin may also be related to thenae. Most of the Golgi cisternae in these colonies were

fragmented into many vesicles whose morphology disorganization of microtubules.likely resembles Golgi membrane vesicles which

Disorganization of the distribution ofemerge in due course of fragmentation at the initialmicrotubules and disappearance of distinctphase of mitosis (Asada and Yagura, 1995).MTOC in HeLa cells exposed to aphidicolin

Multipole spindle body is observed in The association of the Golgi complex and microtu-aphidicolin-arrested mitotic HeLa cells bules has been demonstrated in several systems (Rob-exposed to aphidicolin bins and Gonatas, 1964; Rogalski and Singer, 1984;Ho et al., 1989; Hoshino et al., 1997). To assess theIn cultures exposed to 0.5 mg/ml aphidicolin, cells at

mitotic phase were occasionally observed. Upon obser- relationship between alterations in the distributionand morphology of the Golgi complex in aphidicolin-vation using antitubulin antibody, we found that more

than 70% of the mitotic cells had triple or quadruple arrested cells and microtubules, their distribution wasexamined using antitubulin antibody. As shown in Fig-poles of spindle body (Fig. 7A). As shown in Figure 7B,

the mitotic chromosomes were separated into three to ure 8A, in control cells fine microtubules spread out inall directions from a circumscribed juxtanuclear regionfour parts depending on the number of poles. A similar

phenomenon was observed in HeLa cells treated with where the MTOC localized. However, in cells exposedto aphidicolin at either 0.5 or 5 mg/ml for 72 h, thetaxol, but the multiplicity of poles was larger than that

observed in aphidicolin-arrested HeLa cells (Fig. 7C,D). discrete MTOC was lost and long microtubules ran


TANAKA ET AL.608

throughout the cytoplasm (Fig. 8B,C). Furthermore,microtubules surrounding the entire surface of the nu-cleus were often observed (arrows in Fig. 8B,C).

Exposure to 0.5 mg/ml aphidicolin induceschanges in localization and number

of centrosomesTo determine the mechanism involved in the alter-

ations in the Golgi complex and disorganization of mi-crotubules, we examined the centrosome in aphidicolin-arrested cells using anticentrosome human autoim-mune serum. As shown in Figure 9A, most cells at 72h culture exposed to 0.5 mg/ml aphidicolin (Ç70%) hadseveral centrosomes scattered around the peripheralregion of the nucleus or in the cytoplasm. The size ofthe centrosomes varied slightly in individual cells inaddition to showing an increase in number, suggestingthat the centrosome itself was somewhat abnormal inaphidicolin-arrested cells. By contrast, cells exposed to5 mg/ml aphidicolin for 72 h did not show any multiplic-ity of centrosomes (Fig. 9B). This dose-dependent effectof aphidicolin on centrosomes does not support the ideathat alterations in the Golgi complex and disorganiza-tion of microtubules can be ascribed solely to aberrationof the centrosome. Rather, the increase in the numberof centrosomes observed at 0.5 mg/ml aphidicolin treat-ment might be due to arrest of the cell cycle at G2/Mphase and might not be directly related to the disorga-nization of microtubules.

Protein synthesis is required for theperturbation of the Golgi complex,

microtubule, and centrosomeAt present, the only other effect of aphidicolin on cell

function reported is inhibition of DNA polymerase aactivity. To assess whether the perturbation of theGolgi complex, microtubule, and centrosome wascaused by some unknown direct effect of aphidicolin,the effect of cycloheximide treatment on aphidicolin-induced abnormalities was examined. When cyclohexi-mide was added along with aphidicolin to the culture,neither unbalanced growth nor perturbation of theGolgi complex, microtubule, or centrosome was observed(Fig. 10). Thus, it is suggested that the occurrence ofaphidicolin-induced abnormalities requires the contin-uation of certain cellular processes associated with pro-tein synthesis along with unbalanced growth.

DISCUSSIONWe have occasionally observed giant cells containing

a Golgi complex with morphology distinct from othercells in cultures of established cancer cell lines duringour course of immunocytochemical study. This type ofabnormal cell does not seem to proliferate but to emergefrom the major type of proliferating cells at a certainrate during culture. In this study using a drug whichinhibits specific cellular processes in a cell cycle-specific

Fig. 8. Changes in the distribution of microtubules and loss of manner, we may have reproduced part of the phenome-MTOC in aphidicolin-arrested HeLa cells. After 72 h culture in the non which takes place when cancer cells shift to atypi-absence (A) or presence of 0.5 mg/ml (B) or 5 mg/ml (C) aphidicolin,

cal cells in culture. In such cells, disorganization ofHeLa cells were processed for immunofluorescence microscopy usingantitubulin antibody as described in Materials and Methods. Arrow microtubules was induced along with alterations of thein A indicates microtubules at MTOC and arrows in B and C indicate Golgi complex. The results obtained from experimentscells in which microtubules wrapped around the entire nuclei. Note with cycloheximide/aphidicolin treatment in place ofthat the microtubules in aphidicolin-arrested cells were long and

aphidicolin indicate that the alterations in cellular or-rough in their morphology and that such cells lost their discreteMTOC. Bar, 10 mm. ganization of aphidicolin-arrested HeLa cells does not



Fig. 9. Effect of aphidicolin treatment on centrosome. HeLa cells were treated with 0.5 mg/ml (A) or 5mg/ml (B) aphidicolin for 72 h and processed for immunofluorescence microscopy using anticentrosomehuman autoimmune serum. Arrows indicate a set of centrosomes in a single cell. Bar, 10 mm.

derive directly from the specific biochemical action of stead, the nuclear surface region became the site ofnucleation. Similar changes in the nucleation site ofaphidicolin per se. In this regard, it should be noted

that aphidicolin did not induce alterations in cellular microtubules and localization of the Golgi complex areseen in multinucleate myotubes following a myoblastorganization or marked unbalanced growth in semicon-

fluent to confluent cultures. Conditions permitting free fusion event (Tassin et al., 1985). In myotubes, centri-oles do not associate with Golgi complexes. Similarly,unbalanced growth were required for the induction of

the alterations reported here. Thus, the alterations in in several mammalian cell types, noncentrosomal mi-crotubules have been observed (Chalfie and Thomson,the Golgi complex shown in this study might be due

to changes in the cellular physiology and organization 1979; Bray and Bunge, 1981; Murphy et al., 1986; Mo-gensen et al., 1997). Such noncentrosomal microtubuleswhich take place during long-term unbalanced growth.

In cells in which there was unbalanced growth, singu- may be important for the specialized functions of thesedifferentiated cell types. Thus, the rearrangement oflar alterations of the Golgi complex were seen. First step

modification was the dispersal of the Golgi complex from microtubules induced by aphidicolin may indicate thatHeLa cells undergo similar events, although the pro-the original MTOC region as revealed by immunocyto-

chemical examination. A similar change in localization cess might be incomplete. During aphidicolin treat-ment, a reorganization of the microtubule nucleatingwas also observed in taxol-treated HeLa cells (Hoshino

et al., 1997). However, the final distribution revealed a material free from centrioles on the surface of the nu-cleus might generate noncentrosomal microtubules andstriking difference in the two cases; in taxol-treated cells,

the fragmented Golgi complex was entirely dispersed to then induce fragmentation of the Golgi apparatus.One of the interesting results obtained in this studythe cell periphery along with bundles of stabilized micro-

tubules (Sandoval et al., 1984; Hoshino et al., 1997), was that multiplication of the centrosome continues inHeLa cells arrested by aphidicolin at G2/M phase. Aswhereas the Golgi complex moved to surround the nu-

cleus in cells arrested by aphidicolin. This difference a result, the cells contained several centrosomes afterprolonged culture with aphidicolin. Centrosome dupli-might reflect the distribution of microtubules because

microtubules play roles in both orderly membrane traffic cation irrespective of cell division has been found inembryonic cells (Nagano et al., 1981; Sluder and Lewis,through Golgi elements (Pastan and Willingham, 1981)

and in maintaining the integrity and location of the Golgi 1987; Debec et al., 1996). Centrosome duplication con-tinues in the absence of protein synthesis in Xenopuscomplex in interphase cells (Thyberg and Moskalewsky,

1985; Kreis, 1990). HeLa cells exposed to aphidicolin lost and sea urchin embryos, indicating that the embryoscontain preexisting pools in which to assembletheir characteristic distribution of microtubules ex-

tending in a radial array from the perinuclear MTOC centrosomes (Gard et al., 1990; Sluder et al., 1990). Theresults obtained from the experiments on embryonictoward the periphery of the cell. Interestingly, microtu-

bules which surrounded the nucleus were frequently ob- cells indicate that centrosome cycles and nuclear divi-sion cycles are regulated separately. This may be aserved in aphidicolin-arrested cells but were seldom ob-

served in cells treated with taxol. Thus, we concluded peculiar event limited to embryonic cells since inhibi-tion of DNA synthesis does arrest spindle pole body orthat the disorganization of microtubules accompanied by

the loss of distinct perinuclear MTOC resulted in the centrosome duplication at a stage similar to metaphasein yeast (Byers and Goetsch, 1974) and mammalianalterations of the Golgi complex.

In cells arrested with aphidicolin at 0.5 mg/ml, the cells (Rattner and Phillips, 1973; Kuriyama and Borisy,1981). These early embryonic cell cycles are controlledcentrosome was seemingly intact and localized to the

perinuclear region but the number had increased. How- differently, with some cell cycle checkpoints lacking ormodified from those of somatic cell cycles (Hartwell andever, it seems unlikely that the centrosome harboring

region could nucleate the microtubule assembly. In- Weinert, 1989). In experiments using HeLa cells, we


TANAKA ET AL.610

is clearly separated from the regulation of nuclear divi-sion cycles in HeLa cells treated with drugs which act ina cell cycle-specific manner. In addition to the obviousdistortion of DNA metabolism during aphidicolin treat-ment, the progression of centrosome cycles and the per-turbation of microtubule organization in the absence ofnuclear division might result in either aberrant mitoticprocesses following removal of aphidicolin as shown bythe study of Kung et al. (1990) or in cell killing (Cinatlet al., 1992).

LITERATURE CITEDAsada, S., and Yagura, T. (1995) Golgi membrane vesicles in HeLa

mitotic cells are identified with monoclonal antibody made againstGolgi cisternal membrane protein P138. Cell Struct. Func., 20:445–453.

Bray, D., and Bunge, M.B. (1981) Serial analysis of microtubules incultured rat sensory axons. J. Neurocytol., 10:589–605.

Brundret, K.M., Dalziel, W., Hesp., B., Jarvis, J.A.J, and Neidle, S.(1972) X-raycrystallographic determination of the structure of theantibiotic aphidicolin: Atetracycline diterpenoid containing a newring system. J. Chem. Soc. Chem. Comm., pp 1027–1028.

Byers, B., and Goetsch, L. (1974) Duplication of spindle plaques andintegration of the yeast cell cycle. Cold Spring Harbor Symp. Quant.Biol., 38:123–131.

Chalfie, M., and Thomson, J.N. (1979) Organization of neuronal micro-tubules in the nematode Caenorhabditis elegans. J.Cell Biol.,82:278–289.

Cinatl, J., Cinatl, J., Mainke, M., Weißflog, A., Steigmann, G., Rabe-nau, H., Doerr, H.-W., and Kornhuber, B. (1992) Aphidicolin selec-tively kills neuroblastoma cells in vitro. Cancer Lett., 67:199–206.

Costa, J.J., Keffer, J.M., Goff, J.P., and Metcalfe, D.D. (1992) Aphidi-colin-induced proliferative arrest of murine mast cells: Morphologi-cal and biochemical changes are not accompanied by alterations incytokine gene induction. Immunology, 76:413–421.

Debec, A., Kalpin, R.F., Daily, D.R., McCallum, P.D., Rothwell, W.F.,and Sullivan, W. (1996) Live analysis of free centrosomes in normaland aphidicolin-treated Drosophila embryos. J. Cell Biol., 134:103–115.

Fujiwara, T., Oda, K., Yokota, S., Taakatsuki, A., and Ikehara, Y.(1988) Brefeldin A causes disassembly of the Golgi complex andaccumulation of secretory proteins in the endoplasmic reticulum.J. Biol. Chem., 263:18545– 18552.

Gard, D.L., Hafezi, S., Zhang, T., and Doxsey, S. (1990) Centrosomeduplication continues in cycloheximide-treated Xenopus blastulaein the absence of a detectable cell cycle. J. Cell Biol., 110:2033–2042.

Goswami, P.C., Higashikubo, R., Tolmach, L.J., and Roti Roti, J.L.(1992) G1 shortening following unbalanced growth: A specific v.nonspecific effect. Cell Prolif., 25:251–260.

Hartwell, L.H., and Weinert, T.A. (1989) Checkpoints: Controls thatensure the order of cell cycle events. Science, 246:629–634.

Ho, W.C., Allan, V.J., van Meer, G., Berger, E.G., and Kreis, T.E.(1989) Reclustering of scattered Golgi elements occurs along micro-tubules. Eur. J. Cell Biol., 48:250–263.

Hoshino, H., Tamaki, A., and Yagura, T. (1997) Process of dispersionand fragmentation of Golgi complex by microtubule bundles formedin taxol treated cells. Cell Struct. Funct., 22:325–334.

Ikegami, S., Taguchi, T., Ohashi, M., Oguro, M., Nagano, H., andFig. 10. Cycloheximide antagonizes the effect of aphidicolin. Cyclo-Mano, Y. (1978) Aphidicolin prevents cell division by interferingheximide at 10 mg/ml was added along with aphidicolin (0.5 mg/ml) towith the activity of DNA polymerase a. Nature, 275:458–460.coverslips of log phase cultured HeLa cells. After 72 h culture, cells

Iliakis, G., Nusse, M., and Bryant, P. (1982) Effects of aphidicolin onwere processed for immunofluorescence microscopy using mAbG3A5cell proliferation, repair of potentially lethal damage and repair ofantibody (A), antitubulin antibody (B) or anticentrosome human auto-DNA strand breaks in Ehrlich ascites tumor cells exposed to X-immune serum (C). Arrows in C indicate some of the centrosomes.rays. Int. J. Radiat. Biol., 42:417–434.Bar, 10 mm.

Ishida, Y., Furukawa, Y., DeCaprio, J.A., Saito, M., and Griffin, J.D.(1992) Treatment of myeloid leukemia cells with the phosphataseinhibitor okadaic acid induces cell cycle arrest at either G1/S or G2/M depending on dose. J. Cell. Physiol., 150:484–492.found progression of centrosome cycles irrespective of

Jantti, J., and Kuismanen, E. (1993) Effect of caffeine and reducedthe arrest of cell division. We have obtained evidence temperature (207C) on the organization of the pre-Golgi and thein another study that ICRF 193, an inhibitor of topo- Golgi stack membranes. J.Cell Biol., 120:1321–1335.

Kreis, T.E. (1990) Role of microtubules in the organization of the Golgiisomerase II (Tanabe et al., 1991), has similar effectsapparatus. Cell Motil. Cytoskeleton, 15:67–70.to aphidicolin at 0.5 mg/ml on HeLa cells (unpublished

Kung, A.L., Zetterberg, A., Sherwood, S.W., and Schimke, R.T. (1990)results). ICRF 193 has been shown to arrest cells at Cytotoxic effects of cell cycle phase specific agents: Results of cellG2/M phase and to induce unbalanced growth (Perez cycle perturbation. Cancer Res., 50:7307–7317.

Kuriyama, R., and Borisy, G.G. (1981) Centriole cycle in Chinese ham-et al., 1997). Thus, the regulation of centrosome cycles



ster ovary cells as determined by whole-mount electron microscopy. Sandoval, I.V., Bonifacino, J.S., Klausner, R.D., Henkart, M., andWehland, J. (1984) Role of microtubules in the organization andJ. Cell Biol., 91:814–821.

Lucocq, J. (1991) Okadaic acid induces Golgi apparatus fragmentation localization of the Golgi apparatus. J. Cell Biol., 99:113s–118s.Schiff, P.B., and Horwitz, S.B. (1980) Taxol stabilizes microtubulesand arrest of intracellular transport. J. Cell Sci., 100:753–759.

Matlin, KS, Skins, J., and McNeil, P.L. (1988) Reduced extracellular in mouse fibroblast cells. Proc. Natl. Acad. Sci. USA, 77:1561–1565.Sluder, G., and Lewis, K. (1987) Relationship between nuclear DNApH reversibly inhibits oligomerization, intracellular transport, and

processing of the influenza hemagglutinin in infected madin-darby synthesis and centrosome reproduction in sea urchin eggs. J. Exp.Zool., 244:89–100.canine kidney cells. J. Biol. Chem., 263:11478–11485.

Mogensen, M.M., Mackie, J.B., Doxsey, S.J., Stearns, T., and Tucker, Sluder, G., Miller, F.J., Cole, R., and Rieder, C.L. (1990) Protein syn-thesis and the cell cycle: Centrosome reproduction in sea urchinJ.B. (1997) Centrosomal deployment of g-tubulin and pericentrin:

Evidence for a microtubule nucleating domain and a minus-end eggs is not under translational control. J.Cell Biol., 110:2025–2032.Spadari, S., Sala, F., and Pedrali-Noy, G. (1984) Aphidicolin and eu-docking domain in certain mouse epithelial cells. Cell Motil. Cy-

toskeleton, 36:276–290. karyotic DNA synthesis. Adv. Exp. Med. Biol., 179:169–181.Sukhorukov, V.L., Djuzenova, C.S., Arnold, W.M., and Zimmermann,Murphy, D.B., Grasser, W.A., and Wallis, K.T. (1986) Immunofluo-

rescence examination of beta tubulin expression and marginal band U. (1994) DNA, protein, and plasma-membrane incorporation byarrested mammalian cells. J. Membr. Biol., 142:77–92.formation in development of chicken erythroblasts. J. Cell Biol.,

102:625–638. Tanabe, K., Ikegami, Y., Ishida, R., and Andoh, T. (1991) Inhibitionof topoisomerase II by antitumour agent Bis(2,6-dioxopiperazine)Nagano, H., Hira, S., Okana, K., and Ikegami, S. (1981) A chromo-

somal cleavage of fertilized starfish eggs in the presence of aphidi- derivatives. Cancer Res., 51:4903–4908.Tang, S., Poulin, L., and Levy, J.A. (1992) Lack of human immunode-colin. Dev. Biol., 85:409–415.

Pastan, I.H., and Willingham, M.C. (1981) Journey to the center of ficiency virus type 1 (HIV-1) replication and accumulation of viralDNA in HIV-1-infected T cells blocked in cell replication. J. Gen.the cell: Role of the receptosome. Science, 214:504–509.

Perez, C., Vilaboa, N.E., Garcia-Bermejo, L., de Blas, E., Creighton, Virol., 73:933–939.Tassin, A.N., Paintrand, M., Berger, E.C., and Bornens, M. (1985) TheA.M., and Aller, P. (1997) Differentiation of U-937 promonocytic

cells by etoposide and ICRF-193, two antitumour DNA topoisomer- Golgi apparatus remains associated with microtubule organizingcenters during myogenesis. J. Cell Biol., 101:630–638.ase II inhibitors with different mechanisms of action. J. Cell Sci.,

110:337–343. Turner, J.R., and Tartakoff, A.M. (1989) The response of the Golgicomplex to microtubule alterations: The roles of metabolic energyRattner, J.B., and Phillips, S.G. (1973) Independence of centriole for-

mation and DNA synthesis. J. Cell Biol., 57:359–372. and membrane traffic in Golgi complex organization. J.Cell Biol.,109:2081–2088.Robbins, E., and Gonatas, N.K. (1964) Histochemical and structural

studies on HeLa cell cultures exposed to spindle inhibitors with Thyberg, J., and Moskalewsky, S. (1985) Microtubules and the organi-zation of the Golgi complex. Exp. Cell Res., 159:1–16.special reference to the interphase cell. J. Histochem. Cytochem.,

12:704–711. Yamauchi, T., Higashiura, M., and Yagura, T. (1992) Novel mono-clonal antibody mAb G3A5 recognizes 138-kDa glycoprotein local-Rogalski, A.A., and Singer, S.J. (1984) Association of elements of the

Golgi apparatus with microtubules. J. Cell Biol., 99:1092–1100. ized on the Golgi membrane. Cell Struct. Funct., 17:213–222.


Aphidicolin induces alterations in Golgi Complex and disorganization of microtubules of HeLa cells...

Documents

Transcript of Aphidicolin induces alterations in Golgi Complex and disorganization of microtubules of HeLa cells...