Reorganization of Keratin Intermediate Filaments by the Drug-Induced Disruption of Microfilaments in Cultured Human Keratinocytes

YaSl10 Kitajima, M.D., Ph.D., Shl1nichiro Ino ue, M .D. , and Hideo Yaoita, M .D. Department of Dermatology, Ji chi Mcdica l School. Tochigi-kcn, Japan

It has been shown to date that a combined trea tment w ith a network of connecting star-like knots or foci . These foci nllcrotubule and microfilament inhibitors alters the cy to­ coincid ed with actin aggregates that were formed by de­ skeletal organiza tion of keratin intermediate fi laments in polymeri za tion of actin fil aments as studied by double im­ cultured HeLa, fetal mouse epidermal, and epithelial PtK2 munofluorescence using antiactin and antikeratin antibod­ cells, although neither of these inhibitors alone is able to ies. Under these conditions, no change in microtubule do so. In the present study, we fo und that di sruption of arrangement was observed. Our observations suggest that microfilaments with cytochalasin B induced a remarkable the stability and architecture of keratin fil ament organi­ reorganiza tion of keratin filaments in cultured human ke­ za tion may be supported with the microfilament rather ratinocytes, w hile disruption of microtubules with colchi­ than the microtubule cy toskeleton in cultured human ke­ cine did not affect keratin fil aments. Keratin fil ament or­ ratinocytes. J Invest D ermatol 87: 565-569, 1986 ganization in the presence of cytochalasi n B demonstrated

eratin filaments arc the intermediate-sized fi l.a mcnts and microfi laments, res pectively. Recentl y, a possibility that of epithelial cells, and fo rm one of the 3 classes of acrylamide disrupts reversibl y the keratin network in PtKI cell s cytoskeletons, along with microfilaments and mi­ has been reported (1 6). Microinj ection of anti-a-keratin antibod­ crotubules, in these cell s (reviewed in [1 ,2]) . Re­ ies into epitheli al cell s has also been used for the sa m e purpose, cently, the biochemica l and immunologic character­ showing that the selective destabilization of the fil aillentous net­ KIstI CS of keratin fil am ents have been reviewed [3,4]. However, works of keratin occurs without changing ce ll shape in PtKI, the function of keratin filaments as the intermediatc fil amcnts in PtK2, and SV -40 transformed human keratinocytes (1 7). A com­ epithelial cells has not yet been clea rly defined, although thcy m ay bined treatment with microtubule and microfilam ent inhibito rs function to support mcchanically the various structures of the has also been used successfull y to induce the reorganiza ti on of cytoplasmic space [1] . the cytoskcletal array of keratin filaments in cultured HeLa and In order to characterize the dynamic functions of kcratin fila­ feta l mouse epithelial cells [1 8, 19). although disruption of micro­ ments, studies of the rearrangement of these filam ents during filaments w ith cytochalas ins [1 8,19) or microtubules with col­ mitosis have been performed [5-15). These studies have indica ted chicine derivatives [1 9,20) did not significa ntly alter the inherent that keratin intermediate fuaments are systemati ca ll y and dynam­ organi zation of keratin fi laments. However, it is of grea t interes t ically rearranged during mitosis and, although not all epithelial to note that both cytochalas in B alone and demecolcine alone cell types undergo this phenomenon, they may play some im­ altered keratin fil am ent o rga nization in epitheli al m onkey kidney portant roles in mitosis. TC7 cell s [21]. B esides these studies, some experiments have been conducted In the present study, it is shown that the disruption of micro­ to establish m ethods to disrupt the structure and fun ctions of fil aments with cytochalas in B alone induces a remarkable reor­ keratin intermediate filaments. There are no biochemical agents ganiza tion of keratin fil ament networks in cultured human ke­ that are known to specifica ll y destabilize the structure and/or ratinocytes, whil e the disruption of microtubules with colchi cine interfere with the fun ctions of keratin and other intermediate does not. filaments in a manner analogous to that of the cytostati c drugs such as colchicine and cytochalas in B, w hich act on microtubules MATERIALS AND M ETHODS Antibodies Polyclonal antibodies against keratin filam ent pro­ tein s of human epidermis, w hich were obtained by immunizing Manuscript received December 23, 1985; accep ted fo r publi ca tion April a rabbit w ith a 50 kD subunit of human epidermal keratin (1 3), 29, 1986. were used. Monoclonal antibodies (mouse IgG) against a-tubulin Supported by Grants-in-Aid for Scientific Research, nos. A-58440047 or actin were purchased from Amersham International (England). and B-59480246 from th e Ministry of Edu ca ti on, Science and Culture, As secondary antibodies, flu orescein isothiocyanate (FITC)- or Japan, and the Lydia O'Leary Memorial Foundation , Japan. Reprint requests to: Yasuo Kitajim3 , M.D., Department of Derma­ rhodamine-lissamide sulfonyl-chloride-labeled goat IgG anti­ tology, Jichi Medical School, Minamikawachi-machi, Kawachi-gun, To­ mouse IgG, and FITC-Iabeled goat IgG antirabbit IgG, w hich ch igi-ken, Japan 329-04. were purchased fr0111 Cappel Laboratori es (Westchester, Penn­ Abbrev iations: sylvani a) , were used in the present study. DMSO: dimethylsulfoxid e FITC: fluorescei n iso thiocyanate Cell Culture After removing 111 0st of the subcutaneous tissue PBS: phos phate-buffered sa line with surgica l scissors , skin biopsies from human foreskin o r other PMSF: ph cnylm cth ancslllfonyl flu orid e sites were cut into small pieces (2 X 5 ml11) and soaked in a TPCK: L-l-tosylamide-2-phcnylcth yl chl oromcth yl ketone solution of1oo0 U / ml dispase (Godo Shusei Co., Matsudo, C hiba,

0022-202X/86/S03 .50 Copy ri ght © 1986 by The Society for Investigative Dermatology, Inc.

565 566 K ITAJIMA. INOUE. AND Y AOITA TH E JO URN A L OF IN VESTIGAT IVE DERMATOLOGY

Japan) in phosphatc-buffercd sa line (PBS) for 15 min at 3re. were soaked in a solution of 0.5% Triton X-IOO, 2 mM phen­ Aftcr thi s in cubation, thc cpidermis could bc cas il y pceled off th e ylmcthancsulfonyl flu oride (PMSF) and 2 mM L-l-tosylamide-2- ' dermis with forceps. Thc epidermis was thcll in cubated in 0.25% phenylethyl chloromethyl ketone (TPCK) for 10 min at room trypsin and 0.05% EDT A for 10-15 min at 37°C to obtain di s­ temperature. perscd keratinocytcs. Approximately 105 keratinocytcs were in­ The cell s were incubated with the first antibody for 45 min at oculated on glass coversli ps in a 35-mm plastic di sh. Thc culturcs room temperature, fo ll owed by several washes in PBS. In some wcrc fcd with Dulbccco's modificd Eagle's medium supple­ specimens, double antibody labelin g was ca rried out as fo ll ows: mcnted with 20% feta l ca lf serum, O.4/kg/ml hyd rocortisonc, 10 first the ce ll s were in cubated with rabbit antibody against keratin ng/ml epidermal growth f.Ktor, 84 ng/mJ cholera toxin , 100 /kg/1ll1 for 45 min fo ll owcd by exten sive w as hing, then with the mono­ streptom ycin , and 100 U/ ml peni cillin [22 1. Thc medium was cl onal antibody against a -tubulin or actin for 45 min. After in­ changed at 2-day intervals. C ulturcs werc grown in a humidificd cubation w ith the first antibodics and w ashing, the ce ll s were atmosphere undcr 5% CO2 at 37°e. C ulturcd cell s wcrc subj ected in cubated with thc specifi c second antibodies, rinsed w ith PBS, to experiments 4-6 days after plating. air-dried, and mOllnted in glycerol. Controls for specifi city of the antibodies were routinely ca rried o ut. Photomicrographs were Treatments with Colchicine and Cytochalasin B Cytochal­ taken with Nikon flu oresccnce microscope VFD-R (Nikon, as in B and colchicin e (Sigma C hemi ca l Co., St. Louis, Missouri) werc di ssolved in OMSO and in PBS, respectivcly, and added in Tokyo, Japan). the medium to o btain the fin al conce ntrati on of 20 /kM for each. RESULTS The cells were incubated in thesc medi a for 30, 60, an d 120 min The cytoskeletal arrangement of keratin fil aments in cultured hu­ and then fixed with cold methanol (- 20°C). For control exper­ man keratinocytes, whi ch was visualized by immunofluorescence iments, the ce ll s werc in cubated in the medium containing OM SO microscopy usin g antikeratin antibodies, showed a dense mesh­ without cytochalasin B and colchicin e for the sa me periods. Cell work around the nucleus and a radiating pattern of f!lament bun­ viabili ty during experimental peri ods was ovcr 90% as judged by dl es to the ce ll periphery (F igs l a, 2a). Microtubules also showed the trypan blue exclusion tcst. an array similar to that of keratin, but they appeared to form Immunofluorescence Microscopy The cell s grown on the much thicker filam ent bundles (Fig Ib) . The flu orescence pattern coverslips were rinsed with PBS at room temperaturc, and dippcd of actin fil aments by usin g antiac tin antibody is shown Fi g 2b, for 7 min in methanol at - 20°e. T hc fi xcd cell s, on the coverslips, whi ch showed diffuse distribution of ac tin fil aments througho ut

Figure 2. Double-labeled il11munofluroescence distribution of keratin ftl­ Figure 1. Double-labeled immunoflu orescencc localization of keratin fil­ amcnts (a) and microfilaments (b) in cultured human keratinocytcs is aments (a) and microtubules (b) in cultured human keratinocytes is dem­ shown by usin g antikeratin and antiactin antibodics. These are untreated onstrated by usin g antikeratin and antitubulin antibodies. These are un­ control cell s. Keratin shows a fibrill ar network (a) but actin shows a denser trcated control cells. Both cytoskcletons show a radia l pattern fro m staining pattern in juxtanuclca r and cell periphery regions (b) . Actin fil­ juxtanuclea r fil amcnt aggregati ons. amcnts arc not clear by immunoflu oresccnce. VOL. 87. N O.5 N OVEMBER 1986 REORGANIZATION OF KERATIN CYTOSKELETON 567 the cell , although fil am ent structures were not resolved by this method. When the cell s werc treated with cytochalas in E , the diffuse distribution of actin fil am ents was disrupted and condenscd into small aggregates in thc cell s as shown in Fi g 3b, and keratin filaments were also rea rranged into a unique pattern, which was a " lattice" structure consistin g of many foc i of small as teroid aggregations of fil ament bundles (Fig 3a). These fo ci of keratin­ filament assembli es appcared to coin cidc with the small aggregates of actin as indicated by arrows in Fig 3a and b. Observations at hig her magnification rcvcalcd that thc centers of som c o f thcsc k nots wcrc no t staincd with alltikcratin antibody, but they were solidly staincd with antiactin antibody (Fi g 3a and b in sets). This rearrangemcnt o fkcratin fil am cnts did no t induce the dctachmcnt of desm osom al cell-to-cell contact. Keratin filamcnts form ed aster-likc foci also at thc ccll contact region (large arrow, Fi g 3a). Although thc keratin filaments wcre rcarranged into the lattice structurc, no significa nt rcarrangcment of microtubuJcs was dc­ tected by thc doublc-immunofluorcscence mcthod (Fig 4a and b) . T h ese alterations of the kcratin fil amcnt and actin organization were observcd in 60-90% ofthc ccll s ill culture dishes in differcnt experimcnts, whcn almost all ccll s on thc coverslips wcre cx­ amined. T he disruption of microtubules with colchicine did not induce any significant reorganiza tion of keratin intermediate filaments

Figure 4. Effects of a \-h trea tm ent with cytochalas in B (20 p,M) on keratin filam ent (0) and microtubule cytoskeleton (b) is shown by double­ labeled immunoflu orescence usin g antikeratin and antitubulin antibodics. There arc no maj or cffects on the microtubular cy toskeleton by reorga­ niza tion of keratin fi laments cxe rted by cytochalas in B-induced mi cro­ ftIam cllt disruption. A l-h trcatmcnt with cytochalasi n B changed the keratin filament arrangement into a star-likc, latti ce stru cture (0), but did not chan ge microtubule orga ni zation (b).

(not shown). Combined trea tment w ith cytochalas in E and col­ chicine produccd a lattice arrangem cnt of keratin fil am ents almost identica l to that produced by cytochalasin B treatm ent alone (not shown). D ISCUSSION The present study has shown that the disruption of microfilam ents by cytochalas in B induces a reorganization of keratin fil ament networks in cultured human keratinocytcs in the absence of m aj or reorganiza tion of microtubules (Fig 4a and b). Similar effects of cytochalas in B on keratin filament networks have been observed in monkey kidncy TC7 cell s [21). H owever, the TC7 cell s re­ spond to demecolcine-induced disassembly of microtubules so as to reorganize keratin fil am cnt arra ys, although human keratino­ cytes do not respond in this manner to microtubule disassembly (induced by colchicine in the present study) . On the other hand, it has been shown that the combined disruption of microtubule and microfilament cytoskelctons alters the keratin cytoskeletal Figure 3. Effects of a 1-h treatmcnt with cy tochalasin B (20 p,M) on thc arrangement in mouse carcin o ma ccll s, mouse cpidermal cells, keratin (0 ) and actin (b) fi lament cy tos kelctons in cultured human kerat­ and HeLa-S3 cell s [1 8, 19). These papers have described that ad­ inocytes is shown by doubl e-labeled immunoflu orescence . (antikeratin, and anti actin antibodies). Thc keratin filamcnt arrays are re-form ed into dition of any one drug alone or of cytochalasin D and ,B-Iumi­ a star-like, knotted keratin network (0), while actin fil aments are reso lved colchicine simultaneousl y d oes not elicit significant organiza ti onal into small puncta te aggrega tes (b) . Thc kcratin kn ots appea red to co in cide change in the keratin cytoskelcton and suggested that microfi­ with actin (nlTOII!S) . The ccnter of keratin foci appea rs to be unstained lam ents and microtubules have a combined role in maintaining (a rrows, iI/ set). the arrangement of keratin in H eLa and m ouse epidermal cell s. 568 K ITAJIMA. INOUE. A N D YAOITA TH E JOURNA L O F INVESTIGATIVE DERM AT O LOG Y

In epitheli al PtK2 cell s also, combined treatment with cytochalas in vcl opmcntally rcg ulated cl ass of protcins. Annu Rcv Biochem D and colcemid induced a star-like, lattice-li ke redistribution of 51: 2 19-250, 1980 keratin fil am ents, but neither colcemid nor cytochalasin D alone 3. Frecdbcrg 1M: Epidermal diffcrentiation and kcratiniza ti on, Update: was able to cause a m aj o r reorganiza tio n o f keratin fi lam ents [20] . Dermatology in Ge neral Mcdi cin e. Editcd by TB Fitzpatrick, AZ Since the microfilam ent disruptio n induced the keratin fila ­ Eisc n, K Wolf, 1M Frccdberg, KF Au sten. Ncw York, McG ra w­ m ents to fo rm a lattice structure connected with star-like kno ts, Hill , 1983, pp 159-1 73 w hich arc also stained w ith antiactin antibody (Fi g 3a and b) , it 4. Stcincrt PM, Canticri j S: Epidcrmal keratins, Biochemistry and is pro bable that the o rgani za tion of keratin filaments is supported Ph ys iology of th c Skin. Editcd by LA Goldsmith. Ncw Yo rk, w ith actin arrangem ents in cultured human keratinocytes. How­ Oxford Univ Prcss . 1983, pp 135- 169 ever, when the keratin fil am ents w cre disrupted with anti keratin 5. AubinjE, Osborn M, Frankc WW , Wcbcr K: Intermediatc fi laments antibod y, the micro fil am entous array did not change in PtK and of vimcntin-type and the cy tokeratin-typc arc di stributed differ­ I cntl y during mitosis. Exp Cell Rcs 129: 149-165 , 1980 PtK2 cells [1 7]. T he disruptio n of microfilam ents alo ne did no t change the keratin fi lam ent o rgani za tio n in H eLa cell s, fetal mouse 6. Horwitz B, Kupfer H, Eschar Z, Gcigc r B: Heorga ni zation of arrays epidcrmal cell s, PtK and PtK cells [1 8-20,23]. Therefore, it of prckcratin fi laments during mitos is: immuno Au orcsccncc mi­ I 2 croscopy with multicl onal and monocl ona l prekeratin antibodies. appea rs that the keratin cytoskeleton has different relationships Exp Cell Hcs I34:281-290, 1981 w ith micro fil am cnt syste m s am o ng diffcrent cell types. 7. Frank c WW , Schmid E, Grund C: Intcrmcdiate filam cnt protcins in T hc kno ts of star-like arrangem ents o fkcratin filaments showed nonfilamcntous structures : transicnt disintegrati on and inclusion w hat appearcd to be central unstained spots (Fi g 3a and b), w hile of subunit protcin s in granular aggrcga tcs. Cell 30: 103-113, 1982 thcse kno ts are stain ed completely w ith antiactin antibody . This 8. Taylor-Papadimitriou j , Purkis P, Lanc EB, McKa y lA , C hang SE: suggests that kcratin filaments may aggregate or fo cus onto the Effects of SV40-transforlll ation on thc cy toskel eton and behav­ sm all mass o f actin o r may be trappcd at the sites commo n to ioural propcrtics of human kcratinocytcs. Cell Differ 11: 169-1 80. actin aggregatio n. [n this res pcct, it m ay be worthwhile to no te 1982 that intermediate fil am ents (vimentin o r kcratin) in PtKI cell s 9. Lanc EB, Goodman SL, Trcjdosic wcz LK : Di sruption of the kcratin appeared to be interwoven into m asses o f uno rdered microfila­ fil amcnt nctwork during cpith el ial cc ll division. EMBO j m ents after treatmcnt w ith cytochalasin B [24], and that they 1 :1 365-1372, 1982 wcrc no t rcorganized in to a latticc structure. '10. Brown DT, Anderton BH, While CC: Alterati ons in orga ni zation Concerning the intera ctio n betwecn keratin fi lamcnts and mi­ of cy tokcratin filam ents in normal and mali gnant hum;]n co lonic cro tubules, there havc bcen som e cx pe riments w hich suggest that epithelial cell s during mitosis. Cell Tiss uc Res 233:619-628, 1983 intermediatc fil am ents m ay depend o n microtubules fo r thcir dis­ 11 . Frank c WW , Schmid E, Wcll stccd j, Grund G, Gigl 0 , Geige r B: tributio n in the cyto plasm and m ay act in coordination w ith mi­ Changes of cytokeratin fil ament organi za ti on during the cell cycl c: cro tubules in the positioning o f nuclci. This situatio n has been selccti vc mas kin g of an immunologic dcterminant in intcrphase o bserved in SV5-infccted BHK21-F syncytia [25], altho ug h an PtK2 ce ll s. j Cell BioI 97 :1 255- 1260, 1983 agent, vanadate, has bcen fo und that separatcs intermcdiate 12. jones JCR, Goldman AE, Yang H-Y, Goldman RD: Thc organi­ filaments and microtubules topo logicall y and functionall y (24]. zational fate of intcrmcdi ate fi lamcnt nctworks in two epithelial Also, the coll apsc of intcrmediatc filaments during mitosis and cell typcs during mitosis. J Cell BioI 'lO0:93- 102, 1985 aft cr drug-induced microtubule depo lymerizatio n suggests the 13. Kitajima Y, Inouc S, Yoneda K, Mori S, Yaoita H: Alteration in dependence of intermediate fil am cnt o rganizatio n on microtu­ thc arrangcment of the kcratin-type intcrmedi ate filam ents during buIes in chick g izzard ccll s [26]. R ccently, in te rmcdiate fi lamcnts mitosis in cultured human keratinocytes. Eur J Ccll BioI 38:219-225, 1985 (vimcntin) in gerbil fibro m a cell s have been shown to be co ll apsed in li ving cell s micro injccted with m o no clonal and polyclonal an­ 14. Celi s JE, Larscn PM, Fcy SJ , Cclis A: Phosphorylation of keratin tibo dics against tubu li n [27] . It has been suggested from these and vimcntin polypcptides in normal and transformcd mitotic human cpithelial cclls: behavior of kcratin and vimcntin fil amcnt data that thc antibod ies disrupted postulated intermediate fil a­ during mitosis. J Cell BioI 97 :1 429-1434, 1983 m ent-microtubule intcra ctio ns [27]. 15. Celis j E, Larscn PM, Fey Sj , Cclis A: Prcferential phosphorylation O ur results, however, showed that the o rgani za tio n o f keratin of keratin and vimentin during mitosis in normal and transformed filamcnts appea rcd to be independcnt o f thc microtubule cyto­ human amnion cell s. Ann NY Aca d Sci 455 :268-281, 1985 skeleto n, since thc reorganizatio n o f kcratin fi laments by disru p­ 16. Eckert BS: Alteration of intermediatc filament distribution in PtK I tio n of microfilam cnts did no t sig nifica ntly affect microtubule ccll s by ac rylamide. Elir j Ccll BioI 37:1 69-175, 1985 o rga nizatio n (Fi g 4{/ and b) , and the depo lyme ri zatio n o f micro­ 17. Klymkowsky M, Millcr RH, Lane EB: Morphology, bchavior, an d tubules did no t changc the keratin fi lament networks in cultured intcrac ti on of cultured cpithelial ccll s aftcr th e antibody-induced human keratinocytes. T he inde pendence of thc o rganiza tio n and disruption of keratin filamcnt orga ni za ti on.j Ccll BioI 96:494-509, rcorganiza ti on o f keratin fil am ents from microtubulcs also has 1983 been suggested by do uble-Iabclcd immunoflu o rescencc micro­ 18. Knapp LW , O'Guin WM . Sa wycr RH : Drug-induced altcrati ons of scope studies o n alteratio ns of thcse cytoskeletons during mitosis cy tokcratin organi za tion in cultured cpithclial cell s. Science [5-13] in human keratinocytes [1 3J and o ther epithelial cells [5-13J. 219:501-503, 1983 T hc reason for the difference in susceptibility o f keratin fi laments 19. Knapp LW , O'Guin WM , Sawycr RH: Rcarran gcment ofthc keratin in human keratinocytes to disruptio n b y depolym eriza ti o n of mi­ cytoskel cton after co mbincd trca tmcnt wi th mi crotubule and mi­ crofil am ents and micro rubules are unkno wn. crofilamcnt inhibi tors. j Ccll BioI 97:1788-1794, 1983 20. Kes ki-Oja J, Alitalo K: Hco rganization of plasma membranc-asso­ ciated 36000 dalton protein upon drug-induced redi stribution of We 1Ila lik Dr. F. G rilili ell , Depm'llIl elil of Cell Biolog y, The Ullillersily of cytokcratin . Exp Ccll Rcs 158:86-94, 1985 Texas, Health Sciell ce Cwter at Dallas, f or revie llJillc~ the II/alli/script. We are 21. Celis JE, Small JV, Larsen PM , Fcy SJ, DeMcy J, Cclis A: Inter­ ill dcb led 10 Miss Takako Masllda alld Miss Ayako Shilloz aki fo r experllecllllica l mcdi ate filam ents in monkey kidncy T C7 cclls: fo cal centers and help , al/(I Miss Noriko [shiz aki fo r secretarial assislall ce . interrelationship with othcr cytoskel ctal systcms. Proc Nat! Acad Sci USA 8 1:1117- 11 21, 1984 22. Rheinwald JG, Grecn H: Serial cultivation of strains of human kc­ ratin ocytcs: the formation of keratini zin g colonies fro m single RE FE RE N CES cel ls. Cell 6:331-344, 1975 1. Laza rides E: In tcrmcdiatc fil amcnts as mechani ca l integrators of ccl­ 23. Franke WW , Grund C, Osborn M, Webcr K: T he intermediate­ lul ar shapc. Naturc 283 :249-258, 1980 sizcd filamcnts in rat kangaroo PtK 2 cell s. I. Morphology in situ. 2. Lazaridcs E: Intcrmcdiatc fil aments: a chcmi ca ll y hetcrogcn ous, dc- Cytobiologie 17:365-391, 1978 VOL. 87, NO. 5 N OVEMBER 1986 REORGANIZ ATIO N OF KERATIN CYTOSKELETON 569

24. Wang E, C hoppin PW: Effects of va nadate on intracellular distri­ 26. Geiger B, Singer Sj: Associati on of microtubules and intermediate bution and function of 10-nm filam ents. Proc Nat! Acad Sci USA filaments in chi ck gizzard cells as detected by double immunoflu­ 78:2363-2367, 1981 orescence. Proc N at! Acad Sci USA 77:4761-4773, 1980 25. Wang E, C ross RK, C hoppin PW: Involvement of microtubules and 27. Blose SH, Meltze r Of, Fcramisco JR: 10-nm fil aments are induced 10-nm filament in the movement and positioning of nuclei in to collapse in li ving cdls microinj ccted w ith monoclonal and poly­ syncytia. J Cell Bioi 83:320-337, 1979 clonal an tibodies against tubulin. J Cell Bioi 98:847- 858, 1984