Proc. NatI. Acad. Sci. USAVol. 82, pp. 6908-6912, October 1985Cell Biology

Multiple forms of tubulin in the cilia and cytoplasm ofTetrahymena thermophila

(isoelectric focusing/two-dimensional gel electrophoresls/one-dinensional peptide mapping)

KATHY A. SUPRENANT, EVA HAYS, EDWARD LECLUYSE, AND WILLIAM L. DENTLERDepartment of Physiology and Cell Biology, University of Kansas, Lawrence, KS 66045

Communicated by Charles D. Michener, June 21, 1985

ABSTRACT Most higher eukaryotic tubulins are separat-ed into a- and J3-tubulin when electrophoresed in NaDodSO4-denaturing gels, while many lower eukaryotic tubulins arepoorly resolved under these conditions, which include a stack-ing gel (pH 6.80) and a separating gel (pH 8.80). By loweringthe pH of the separating gel to 8.25, we have found that tubulinisolated from the protozoan Tetrahymena thermophila is re-solved by one-dimensional polyacrylamide gel electrophoresisinto two a-tubulins and one P-tubulin. Moreover, at least fiveA- and two .3-tubulin isotypes are identified in Tetrahymena byisoelectric focusing and two-dimensional polyacrylamide gelelectrophoresis. Three of these a isotypes and one (3 isotype arefound specifically in ciliary microtubules, while the other twoisotypes are found only in the cytoplasmic tubulin pool that wasisolated and induced to self-assemble into microtubules in vitro.Peptide mapping by limited proteolytic digestion indicates thatthe tubulins are closely related. Possible mechanisms for thegeneration and selection of these tubulin isotypes are discussed.

Microtubules participate in a wide variety of cellular func-tions that include the development and maintenance of cellshape, the movement of chromosomes during mitosis andmeiosis, the translocation of organelles, and the motility ofeukaryotic cilia and flagella. Some of these processes requireprecise spacial and temporal control ofmicrotubule assemblyand disassembly, as in the formation of the mitotic apparatus,and some require the assembly of morphologically distinctmicrotubules, as in the assembly of ciliary and flagellarmicrotubules.

Microtubules are composed primarily of dimers of a- and,B-tubulins, each with a molecular weight of 50,000 (1-4).Within the same cell and among different cell types, a- and/3-tubulins are heterogeneous as judged by protein and DNAsequencing (2-6), electrophoresis, isoelectric focusing, pep-tide mapping (7-15), immunological methods (16-19), andgenetic analysis (20, 21). Whether these tubulin differencescan be correlated with the various functions with whichmicrotubules are associated is an intriguing question (22). Tocorrelate tubulin biochemistry with microtubule function, itis appropriate to study the simplest cell type or organism that(i) can be grown in large quantities to permit biochemicalanalysis, (ii) contains functionally different microtubule sys-tems that can be isolated in vitro, and (iii) is amenable togenetic and molecular biological techniques.The ciliated protozoan Tetrahymena therntophila contains

an elaborate array of microtubules involved in ciliary motil-ity, cell division and conjugation, the organization of the oralapparatus, and the maintenance of cell shape (23). Methodsare available for the independent isolation of cilia (24),cytoplasmic microtubules of the oral apparatus (25, 26), andthe soluble pool of cytoplasmic tubulin (ref. 27 and this

report). In addition, techniques are available for the potentialisolation of tubulin mutants (28) and the analysis of tubulingenes (29). Overall, Tetrahymena is an ideal organism for thestudy of tubulin expression, heterogeneity, and function.

In this report, methods are presented for the electropho-retic separation of Tetrahymena tubulin and for the purifi-cation and in vitro assembly of cytoplasmic tubulin. Ciliarytubulin is compared to the assembly-competent pool ofcytoplasmic tubulin by one-dimensional (1-D) and two-dimensional (2-D) PAGE, isoelectric focusing (IEF), andlimited proteolytic digestion. Several different a- and f3-tubulins are identified in the ciliary axoneme, some of whichare distinct from those in the soluble cytoplasmic pool.

MATERIALS AND METHODSMaterials. Ultrapure ammonium sulfate was purchased

from Schwarz-Mann and ultrapure urea was from BethesdaResearch Laboratories. Carrier ampholytes for IEF (pHranges 4-6, 5-7, 3.5-10) were obtained from LKB. Electro-phoretic grade acrylamide and NN'-methylenebisacryla-mide were obtained from Bio-Rad.

Preparation of Protein. Ciliary axonemes were isolatedfrom Tetrahymena thermophila, clone SB911 (30), as de-scribed (24, 31).Cytoplasmic tubulin was purified from deciliated cell

bodies by ion-exchange chromatography on DEAE-Seph-adex (27). Phenylmethylsulfonyl fluoride (PMSF) (1 mM) andleupeptin hemisulfate (synthetic) (10 pug/ml) [a competitiveinhibitor of serine and thiol endopeptidases in Tetrahymena(32, 33)] were added to all solutions. The isolated tubulin wasconcentrated by ammonium sulfate precipitation [50% satu-ration (wt/vol), 313 g/liter] and desalted into assembly buffer(100 mM Pipes adjusted to pH 6.80 with NaOH/1 mMMgCl2/1 mM EGTA/1 mM GTP/1 mM PMSF/10 pug ofleupeptin/ml) by gel filtration chromatography on SephadexG-25. Microtubules were assembled at 300C by either twocycles of reversible assembly (34) or the addition of 10 ,uMtaxol (35).Bovine and chicken brain tubulins were isolated by two

cycles of reversible assembly as described (36).Analytical Methods. Proteins, resolved on 1-D PAGE [10o

T (total acrylamide)/0.5% C (bis-acrylamide, crosslinker)](37), were fixed and stained with Coomassie blue R-250 (38).For some experiments, the pH of the separating gel wasvaried from 9.25 to 8.25 by the addition of HC1, resulting ina calculated increase in ionic strength of 0.01 over thestandard Laemmli formulation (37).Tubulins were resolved by 2-D PAGE (39) with a first-

dimension IEF gel with 2% ampholytes (0.8%, pH 5-7; 0.8%,pH 4-6; 0.4%, pH 3.5-10; vol/vol). Electrophoresis in thesecond dimension was carried out under denaturing condi-

Abbreviations: IEF, isoelectric focusing; 1-D, one-dimensional; 2-D,two-dimensional.

6908

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Dow

nloa

ded

by g

uest

on

Dec

embe

r 7,

202

0

Proc. Natl. Acad. Sci. USA 82 (1985) 6909

tions (37) with a pH 8.25 separating gel. Gels were stainedwith Coomassie blue R-250 (38).

Isoelectric focusing in 0.75-mm-thick slab gels was carriedout vertically in a Bio-Rad SE 600 gel apparatus using theampholyte composition described above. IEF slab gels werestained with crocein scarlet and Coomassie blue R-250 (40).Peptide maps of tubulins were generated by limited

proteolytic digestion with type XVII Staphylococcus aureusV8 protease (Sigma) in NaDodSO4/polyacrylamide gels (41).Tetrahymena 83-tubulin was visualized on nitrocellulose

blots by immunoperoxidase methods as modified by Binderet al. (42), using a monoclonal antibody (clone Tu9B) againstmammalian brain f3-tubulin (43).

Pellets of in vitro-assembled cytoplasmic microtubuleswere fixed and embedded for electron microscopy as de-scribed (36).

RESULTSPurification and Assembly of Cytoplasmic Tubulin. Purified

Tetrahymena cytoplasmic tubulin was assembled into intactmicrotubules by either two cycles of temperature-dependentassembly and disassembly or taxol-induced polymerization(Fig. 1). The yield of tubulin obtained after taxol-inducedassembly is 4-fold higher (3-5 mg per 1010 cells) than thatobtained by self-assembly (0.8-1.5 mg per 1010 cells); how-ever, no differences are observed in the composition of thetubulin found in microtubules assembled by either method, asassayed by 2-D PAGE, IEF, or peptide mapping. Taxol-induced assembly, therefore, did not select for a specificisotype of cytoplasmic tubulin.

Separation and Identification of Tubulins. As a first step inthe analysis of Tetrahymena tubulin heterogeneity, condi-

tions were established for maximal resolution of the a- and,3-tubulins in acrylamide gels. Separation of Tetrahymenatubulin on 1-D denaturing gels was found to be dependent onthe pH of the separating gel. Identical tubulin samples wereelectrophoresed on five separate gels that differed only in thepH of the separating gels (Fig. 2). As the pH of the separatinggel decreased, the separation of Tetrahymena a- and f3-tubulin increased with maximum resolution occurring in pH8.25 gels. In contrast, the resolution of chicken and bovinebrain tubulin decreases with decreasing pH, such that at pH8.25 there is no separation of these brain tubulins.For most tubulins, the faster migrating polypeptide in 1-D

PAGE is ,3-tubulin (45); however, under our conditions, theTetrahymena a-tubulin is the faster polypeptide. This isconfirmed by immunoblot procedures using a monoclonalantibody against mammalian brain 83-tubulin (Fig. 3) and bythe migration of the two tubulins on IEF and 2-D PAGE (seeFigs. 4 and 5), where the slowest migrating tubulin in theNaDodSO4 direction is the most acidic ,B-tubulin.With the increased resolution provided by the low pH

separating gel, differences can be observed between ciliaryand cytoplasmic a-tubulins by 1-D PAGE. Ciliary tubulinmigrates as a single 1- and a-tubulin complex on these gels,whereas cytoplasmic tubulin migrates as a single fB-tubulinand a doublet of a-tubulins (see arrowheads, Fig. 2). Theslower migrating, cytoplasmic a-tubulin comigrates with theaxonemal a-tubulin under these conditions. The apparent Mrof the tubulin polypeptides in pH 8.25 1-D gels are 53,000 for/3- and 50,000 and 48,000 for the cytoplasmic a-tubulindoublet.

Isoelectric Focusing and 2-D Gel Electrophoresis. In order tocompare ciliary and cytoplasmic tubulins more closely, thesamples were electrophoresed in adjacent lanes in high-

of/~6~

t0 W~~~~~0, ,,.,,04?>¢_*/ ,.>..

l ift.* !

Op.el

|-ONMA. ...

._I. *ew

Jor,;"::

_4gmiw Ajo Tu

-Owmw -4

MAW..: ,'..

'aWI

A B C D E F G

FIG. 1. Purification of Tetrahymena cytoplasmic tubulin. Tubulin was purified by DEAE-Sephadex-batch chromatography and induced toassemble into microtubules by the addition of taxol. (Left) The predominant polymer formed was microtubules; however, in selected regionsof the microtubule pellet, S-shaped sheets, macrotubules, and doublet-like microtubules were seen in thin sections for electron microscopy.(Right) The purification of Tetrahymena cytoplasmic tubulin was monitored by NaDodSO4/PAGE (pH 8.75 separating gel). Cytoplasmic tubulinwas purified from the total protein in the 40,000 x g supernatant fluids (lane A) by DEAE batch absorption. The nontubulin proteins were eithernot absorbed to the DEAE (lane JO) or were eluted with 0.35 M NaCl (lane C). Cytoplasmic tubulin was eluted with 0.7 M NaCI (lane D) andwas concentrated by ammonium sulfate precipitation and desalted into a microtubule-reassembly buffer (lane E). Microtubules, induced toassemble in the presence of taxol, were pelleted at 40,000 x g (lane G); the proteins remaining in the supernatant fluids are shown in lane F.Tubulin (Tu) is the major protein in the microtubule pellet; however, there is also a significant amount of a 46,000 Mr protein (arrowhead) thatco-sediments with the tubulin. Preliminary experiments have indicated that the 1-D peptide map of the Mr 46,000 protein resembles actin (62).This protein is enriched in the taxol-microtubule pellet (lane G), but it is not found in microtubule preparations obtained by two cycles oftemperature-dependent assembly (compare lane G, to Fig. 3 Left, lane C).

~v n

& t8

Cell Biology: Suprenant et al.

Dow

nloa

ded

by g

uest

on

Dec

embe

r 7,

202

0

6910 Cell Biology: Suprenant et al.

Ch B Cy Aay- " Ut

1A- ~ 9 9-/3@saS25

9...._ *-3,

:59a.- 9........ ....

a,P- 0

825 s

Ci_3 /2 1/1

A

C I

132a5a4

% -oi -a

FIG. 2. pH effect on tubulin separation in 1-D NaDodSO4/PAGE. The migration of chicken (Ch) and bovine (B) brain tubulinis compared to Tetrahymena cytoplasmic (Cy) and ciliary axonemal(A) tubulin on five separate gels where the pH of the separating gelwas decreased from 9.25 to 8.25 (as indicated). The Tetrahymenacytoplasmic tubulin fraction was obtained by ion-exchange chroma-tography and was not assembled into microtubules (this fractioncorresponds to Fig. 1, lane D). The position of the a- and (3-tubulinsare shown for chicken and bovine brain tubulin to the Left and forTetrahymena tubulin to the Right.

resolution isoelectric focusing slab gels (Fig. 4) and on 2-DNaDodSO4 gels (Fig. 5). Under these conditions, the singleaxonemal a-tubulin observed in 1-D gels is resolved into 3-5isotypes (Fig. 4A). Since only 3 a-tubulin spots are seenclearly in 2-D gels (Fig. 5A), two of the IEF bands must benontubulin polypeptides or have merged in the seconddimension of electrophoresis. The a-tubulin isotypes, desig-nated al, a2, or a3 (from acidic to more basic), are not foundin cytoplasmic tubulin preparations. In vitro-assembled cy-toplasmic microtubules are composed of two a-tubulinisotypes, a major a5 and a less abundant a4 (Fig. 5 C and C').

A C B A C B

* _a4It 1-Tu

FIG. 3. Immunoblot (Right) and NaDodSO4/polyacrylamide gel(Left) of Tetrahymena axonemal tubulin (lanes A), cytoplasmictubulin purified by ion-exchange chromatography and two cycles ofassembly (lanes C), and bovine brain tubulin (lanes B). The tubulinswere resolved onpH 8.25 gels and stained for protein with Coomassieblue R-250 or electrophoretically transferred to nitrocellulose andstained with the monoclonal antibody Tu9B against mammalian brain(-tubulin, using the peroxidase-antiperoxidase technique. The posi-tion of the ,-tubulin (13-Tu) is indicated.

.I

4-

GEE

FIG. 4. Axonemal tubulin (A) (30 and 50 lug) and cytoplasmictubulin purified by ion-exchange chromatography and two cycles ofassembly (C) (30 and 50 ,ug) are compared in adjacent lanes in a highresolution IEF slab gel. Five a-tubulins, designated a,, a2, a3, a4, oras (from acidic to basic), and two 3-tubulins, (i and ,2w are identifiedin Tetrahymena. Only three of the five bands in the a-tubulin regionin A are identified as tubulins on 2-D gels and so are labeled a,-3.

The (3-tubulins are well separated from the a-tubulins in theIEF and NaDodSO4 dimensions. Only one isotype of 83-tubulin, 132, is found in cytoplasmic microtubule prepara-tions, while axonemal microtubules contain two (3-tubulinisotypes: 12, which comigrates with the cytoplasmic 1-tubulin, and f31, a more acidic and less abundant tubulin (Figs.4 and 5).The apparent pI values of the tubulins, obtained from the

high-resolution slab-IEF gels (Fig. 4), are a,, pI 5.38; a2, pI5.40; a3, pI 5.42; a4, pI 5.50; a5, pI 5.55; 13, pI 5.20; and 32,pI 5.28.

1-D Peptide Mapping. Peptide mapping by limited prote-olysis (41) indicates that the ciliary and cytoplasmic tubulinsare closely related. Fig. 6 illustrates the pattern of peptidesresolved on a pH 8.25 separating gel, obtained by S. aureusV8-protease digestion of a- and (3-tubulins from the cilia andthe cytoplasm. The peptide maps produced by V8-proteasedigestion of ciliary and cytoplasmic a-tubulins are qualita-tively similar, but not identical, in pH 8.25 gels (Fig. 6). Twoof the high Mr cytoplasmic a-tubulin peptides (arrowheads)do not comigrate with the major ciliary a peptides. Theunique high Mr peptides are produced from the faster mi-grating cytoplasmic a-tubulin in 1-D gels. Close inspection ofthe two gels reveals that a small amount of these peptides isfound in the ciliary-microtubule preparation (Fig. 6A). Thepeptide map of the slower migrating cytoplasmic a-tubulin,

0U)00coz

IEF

C 'U

as.<:;: K x. a,45 aa2t

A C

a3 02 1 2a3 "CZ5 C104

FIG. 5. Cytoplasmic tubulin, purified by ion-exchange chroma-tography and two cycles of assembly, and ciliary axonemal tubulinare compared by 2-D NaDodSO4/PAGE. Cytoplasmic tubulin isresolved into a single a5 and (2 tubulin at moderate protein loadings(25 ,ug, C). At higher protein loadings (100 ,ug, C'), two cytoplasmica-tubulins, a4 and a5, are apparent, as is a single (2 tubulin.Axonemal tubulin (50 Ag, A) is resolved into more acidic a-tubulins(al, a2, and a3) and two P-subunits (p3k and (2). The axonemaltubulins (50 Ag) are distinct from the cytoplasmic tubulins (25 ,ug)when co-electrophoresed in the same gel (M).

Proc. Natl. Acad Sci. USA 82 (1985)

41

.,Iit, A

W, I..::. 1.f, ?..".4

Dow

nloa

ded

by g

uest

on

Dec

embe

r 7,

202

0

Proc. Natl. Acad. Sci. USA 82 (1985) 6911

NaDodSO4 0

CO)~0a

0

(a

w~~~~~~- e

FIG. 6. One-dimensional peptide maps of Tetrahymena tubulin

digested with V8 protease and stained with Coomassie blue. (A)

Axonemal tubulin (150 At.g) and (C) cytoplasmic tubulin purified by

ion-exchange chromatography and taxol-induced assembly (150 A.g),were resolved on 1-D gels (10% acrylamide, pH 8.25). The entire gellane was cut out and transferred at right angles to a second slab gel

(15% acrylamide, pH 8.25). Partial proteolysis was performed in the

stacking gel of the second electrophoresis and the resulting side byside pattern of peptides produced by the a- andf/-tubulins is shown.

The "e" indicates the location of V8 protease (20 jig per 150 ,A.gtubulin). The a-tubulin peptides, enriched in the cytoplasmic tubulin

preparations, are marked with arrowheads.

just to the left of the unique peptides, is similar if not identical

to the major ciliary a-tubulin peptide map. The ciliary and

cytoplasmic f-tubulin peptide maps are identical in pH 8.25

gels.

DISCUSSION

Tetrahymena is an ideal organism for studying the regulationof microtubule assembly because it contains a diverse pop-

ulation of microtubules that are assembled at specific placesand at specific times during the cell cycle. Biochemicalanalysis of these microtubules has, however, been limited

because methods had only been developed to isolate the

ciliary microtubules or very small quantities of soluble

cytoplasmic tubulin (27). The methods developed during this

study substantially increase the yield of cytoplasmic tubulin

isolated and show, for the first time, that the cytoplasmictubulin can be further purified by cycles of in vitro assemblyinto microtubules without the addition of exogenousmicrotubule "seeds" (27). With these methods, Tetrahy-mena tubulin can now be isolated in quantities sufficient for

biochemical analysis and comparison with other tubulins in

Tetrahymena such as the cortical and ciliary mirotubules, as

well as with tubulins from other organisms.Analysis of Tetrahymena tubulins by 1-D NaDodSO4/

PAGE reveals that the migration of a- and eo-tubuns is

considerably different from that found in mammalian or avian

tubulins. In this respect, Tetrahymena tubulins behave in a

similar manner to tubulins isolated from Physarum (46, 47),yeast (44), Chlamydomonas (48, 49), and trypanosomes (13)but unlike tubulins from these species, the resolution ofTetrahymena tubulin is not improved by the presence of ureaor alterations in the source or chain length ofNaDodSO4 usedin the gel (unpublished data). Resolution of both ciliary andcytoplasmic a- and B-tubulins in Tetrahymena required lowpH separating gels, of pH 8.0-8.3, while resolution ofchicken and bovine brain tubulins required separating gels ofpH 8.8 or greater (see also refs. 27, 50, and 51). An additionaldifference between brain and Tetrahymena tubulins is themigration of the 8-tubulin. At high pH (where brain a- and,8-tubulins are best resolved), brain 83-tubulin migrates fasterthan a-tubulin (45), but at the low pH, Tetrahymena a-tubulinmigrates faster than f8-tubulin. Since there is no establishedbasis for the binding ofNaDodSO4 to proteins at different pHvalues, it is uncertain whether the separation of Tetrahymenatubulin at low pH results from an electrostatic interactionbetween NaDodSO4 and ionized cationic side chains or fromdifferential NaDodSO4 binding to the a- and f-tubulins.Lowering the pH of the separating gels may be of generalinterest for the analysis of tubulins that are poorly resolvedon standard Laemmli gels (37).Tetrahymena contains as many as seven distinct isoelectric

variants of tubulin, five of which are a-tubulins. Two of thea-tubulins with the most basic pI values are found in thecytoplasmic pool of tubulin and the three remaining a-tubulins are specific for ciliary microtubules. These resultspose four significant questions. (i) Are the tubulin isotypesproduced by posttranslational modification or transcriptionof different tubulin genes? (ii) How does the cell sort theorganelle-specific tubulins? (iii) If the tubulin is posttransla-tionally modified, where does this occur? (iv) Do the differentisotypes of tubulin function differently in terms of theirassembly properties or interaction with other cellular com-ponents?The a-tubulin heterogeneity may result from posttransla-

tional modification, since the modification of cytoplasmictubulin to become ciliary tubulin is not without precedent.a-Tubulin is posttranslationally modified as it is assembledinto flagellar microtubules of Chlamydomonas (52, 53) by theacetylation of the e amino group of a lysine (54, 55). Similarposttranslational modifications of flagellar a-tubulins havebeen observed in Polytomella (15), and the trypanosome,Crithidia fasciculata (56), although the sites of the modifi-cation(s) in these organisms are unknown.

In Chlamydomonas, Polytomella, and Crithidia, a singlecytoplasmic a-tubulin appears to be converted to a singlemodified flagellar a-tubulin. In Tetrahymena, however, thepresence of five distinct isotypes of a-tubulin indicates thatthere would have to be a series of potentially differentmodifications of a single a-tubulin precursor. Moreover, ifthe a-tubulin diversity in Tetrahymena is generatedposttranslationally, the modification(s) would have to alterthe migration of the tubulin peptides in the pH 8.25 peptidemaps, since the peptide map of the faster migrating cytoplas-mic a-tubulin is clearly different than the slower migratingcytoplasmic a- and the major ciliary a-tubulins. These resultsraise the possibility that there are at least two a-tubulin genesthat give rise to the different a-tubulins. To date, however,only one a-tubulin gene has been identified in Tetrahymenausing Southern blot analyses with heterologous tubulinprobes (29).Whether the isotypes are generated at the transcriptional or

posttranslational level, the altered ciliary tubulins must eitherbe modified (i) in the cytoplasm and find their way to theaxoneme, (ii) in the ciliary matrix, or (iii) directly at theassembly end of the ciliary microtubules, distal to the cellbody (57, 58). The distal ends of Chlamydomonas, Tetrahy-mena, and virtually every cilia examined terminate in distinct

Cell Biology: Suprenant et al.

Dow

nloa

ded

by g

uest

on

Dec

embe

r 7,

202

0

6912 Cell Biology: Suprenant et al.

morphological structures, referred to collectively as micro-tubule-capping structures (59-61). It is tempting to speculatethat these structures, which are present throughout ciliarygrowth, are somehow involved in the posttranslational mod-ification of flagellar and ciliary tubulin isotypes or that theyact as molecular sieves, filtering out unwanted tubulins. Inthis context, we feel that Tetrahymena will be a valuableorganism for further studies, particularly since we havedemonstrated the presence of organelle-specific tubulins andhave presented methodologies in which to study them.

We thank Dr. L. Binder for kindly providing antibodies and Dr. E.Orias for providing Tetrahymena cultures, as well as Drs. P. Burtonand R. Himes for helpful discussions during the course of this work.This research was supported by grants from the National Institutesof Health (AM21672, GM32556), Biomedical Sciences and GeneralResearch grants funded by the University of Kansas (W.L.D.) andthe American Cancer Society (PF-2415) (K.A.S.). The taxol used inthese studies was generously supplied by the Natural ProductsBranch, Division of Cancer Treatment, National Cancer Institute.

1. Luduena, R. F., Shooter, E. M. & Wilson, L. (1977) J. Biol.Chem. 252, 7006-7014.

2. Krauhs, E., Little, M., Kempf, T., Hofer-Warbinek, R., Ade,W. & Ponstingl, H. (1981) Proc. Natl. Acad. Sci. USA 78,4156-4160.

3. Ponstingl, H., Krauhs, E., Little, M. & Kempf, T. (1981) Proc.Natl. Acad. Sci. USA 78, 2757-2761.

4. Valenzuela, P., Quiroga, M., Zaldivar, J., Rutter, W. J. &Kirschner, M. W. (1981) Nature (London) 289, 650-655.

5. Hall, J. L., Dudley, L., Dobner, P. R., Lewis, S. A. &Cowan, N. J. (1983) Mol. Cell. Biol. 3, 854-862.

6. Sullivan, K. & Cleveland, D. W. (1984) J. Cell Biol. 99,

1754-1760.7. Bibring, T., Baxandall, J., Denslow, S. & Walker, B. (1976) J.

Cell Biol. 69, 301-312.8. Gozes, I. & Sweadner, K. J. (1981) Nature (London) 294,

477-480.9. Nelles, L. P. & Bamburg, J. R. (1979) J. Neurochem. 32,

477-489.10. Field, D. J., Collins, R. A. & Lee, J. C. (1984) Proc. Natl.

Acad. Sci. USA 81, 4041-4045.11. Stephens, R. E. (1978) Biochemistry 14, 2882-2891.12. Little, M., Luduena, R. F., Langford, G. M., Asnes, C. F. &

Farrell, K. (1981) J. Mol. Biol. 149, 95-107.13. Russell, D. G., Miller, D. & Gull, K. (1984) Mol. Cell Biol. 4,

779-790.14. Burland, T. G., Gull, K., Schedl, T., Boston, R. S. & Dove,

W. F. (1983) J. Cell Biol. 97, 1852-1859.15. McKeithan, T. W., Lefebvre, P. A., Silflow, C. D. &

Rosenbaum, J. L. (1983) J. Cell Biol. 96, 1056-1063.16. Gundersen, G. G., Kalnoski, M. H. & Bulinski, J. C. (1984)

Cell 38, 779-789.17. Thompson, W. C., Asai, D. J. & Carney, D. H. (1984) J. Cell

Biol. 98, 1017-1025.18. Cumming, R., Burgoyne, R. D. & Lytton, N. A. (1984) J. Cell

Biol. 98, 347-351.19. Gozes, I. & Barnstable, C. J. (1982) Proc. Natl. Acad. Sci.

USA 79, 2579-2583.20. Kemphues, K. J., Kaufman, T. C., Raff, R. A. & Raff, E. C.

(1982) Cell 31, 655-670.21. Oakley, R. R. & Morris, R. (1981) Cell 24, 837-845.22. Fulton, C. & Simpson, P. A. (1976) in Cell Motility, eds.

Goldman, R., Pollard, T. & Rosenbaum, J. (Cold SpringHarbor Laboratory, Cold Spring Harbor, NY), pp. 987-1005.

23. Williams, N. E. (1975) Int. Rev. Cytol. 41, 59-86.24. Thompson, G. A., Baugh, L. & Walker, L. F. (1974) J. Cell

Biol. 61, 253-257.25. Wolfe, J. (1970) J. Cell Sci. 6, 679-700.26. Rannestad, J. & Williams, N. E. (1971) J. Cell Biol. 50,

709-720.27. Maekawa, S. & Sakai, H. (1978) J. Biochem. (Tokyo) 83,

1065-1075.28. Orias, E. & Bruns, P. J. (1976) Methods Cell Biol. 13, 247-282.29. Callahan, R. C., Shalke, G. & Gorovsky, M. A. (1984) Cell 36,

441-445.30. Orias, E., Flacks, M. & Satir, B. H. (1983) J. Cell Sci. 64,

49-67.31. Dentler, W. L. (1984) J. Cell Sci. 66, 167-173.32. Umezawa, H. (1976) Methods Enzymol. 45, 678-695.33. Levy, M. R., Sisskin, E. E. & McConkey, C. L. (1976) Arch.

Biochem. Biophys. 172, 634-647.34. Borisy, G. G., Marcum, J. M., Olmsted, J. B., Murphy, D. B.

& Johnson, K. A. (1975) Ann. N. Y. Acad. Sci. 253, 107-132.35. Shiff, P. B., Fant, J. & Horowitz, S. B. (1979) Nature (Lon-

don) 277, 665-667.36. Dentler, W. L., Granett, S., Witman, G. B. & Rosenbaum,

J. L. (1974) Proc. Natl. Acad. Sci. USA 71, 1710-1714.37. Laemmli, U. K. (1970) Nature (London) 227, 680-685.38. Fairbanks, G. T., Steck, T. L. & Wallach, D. F. H. (1971)

Biochemistry 10, 2606-2617.39. O'Farrell, P. H. (1975) J. Biol. Chem. 250, 4007-4021.40. Crowle, A. J. & Cline, L. J. (1977) J. Immunol. Methods 17,

379-381.41. Bordier, C. & Crettol-Jarvinen, A. (1979) J. Biol. Chem. 254,

2565-2567.42. Binder, L. I., Frankfurter, A. & Rebhun, L. (1985) J. Cell Biol.

101, 1371-1378.43. Caceres, A., Binder, L. I., Payne, M. R., Bender, P., Rebhun,

L. & Steward, 0. (1984) J. Neurosci. 4, 394-410.44. Kilmartin, J. V. (1981) Biochemistry 20, 3629-3633.45. Bryan, J. & Wilson, L. (1971) Proc. Natl. Acad. Sci. USA 68,

1762-1766.46. Roobol, A., Pogson, C. I. & Gull, K. (1980) Exp. Cell Res. 130,

203-215.47. Clayton, L., Quinlan, R. A., Roobol, A., Pogson, C. I. & Gull,

K. (1980) FEBS Lett. 115, 301-305.48. Jarvik, J. & Rosenbaum, J. L. (1980) J. Cell Biol. 85, 258-272.49. Remillard, S. P. & Witman, G. B. (1982) J. Cell Biol. 93,

615-631.50. Bryan, J. (1974) Fed. Proc. Fed. Am. Soc. Exp. Biol. 33,

152-157.51. Williams, N. E., Vandaux, P. E. & Skriver, L. (1979) Exp.

Cell Res. 123, 311-320.52. L'Hernault, S. W. & Rosenbaum, J. L. (1983) J. Cell Biol. 97,

258-263.53. Brunke, K. A., Collis, P. S. & Weeks, D. P. (1982) Nature

(London) 297, 316-318.54. L'Hernault, S. W. & Rosenbaum, J. L. (1985) J. Cell Biol.

100, 457-462.55. L'Hernault, S. W. & Rosenbaum, J. L. (1985) Biochemistry

24, 473-478.56. Russell, D. G. & Gull, K. (1984) Mol. Cell. Biol. 4, 1182-1185.57. Rosenbaum, J., Moulder, J. & Ringo, D. L. (1969) J. Cell Biol.

41, 600-619.58. Witman, G. B. (1975) Ann. N. Y. Acad. Sci. 253, 178-191.59. Sale, W. S. & Satir, P. (1977) Cell Biol. Int. Rep. 1, 56-63.60. Dentler, W. L. & Rosenbaum, J. L. (1977) J. Cell Biol. 74,

747-759.61. Dentler, W. L. (1981) Int. Rev. Cytol. 72, 1-47.62. Mitchell, E. J. & Zimmerman, A. M. (1985) J. Cell Sci. 73,

279-297.

Proc. Natl. Acad Sci. USA 82 (1985)

Dow

nloa

ded

by g

uest

on

Dec

embe

r 7,

202

0