Proc. Natl. Acad. Sci. USAVol. 93, pp. 1735-1742, March 1996

Review

Going mobile: Microtubule motors and chromosome segregationNelson R. Barton and Lawrence S. B. GoldsteinHoward Hughes Medical Institute, Division of Cellular and Molecular Medicine, Department of Pharmacology, University of Califomia San Diego, 9500 GilmanDrive, La Jolla, CA 92093-0683

ABSTRACT Proper chromosomesegregation in eukaryotes depends uponthe mitotic and meiotic spindles, whichassemble at the time of cell division andthen disassemble upon its completion.These spindles are composed in large partof microtubules, which either generateforce by controlled polymerization anddepolymerization or transduce force gen-erated by molecular microtubule motors.In this review, we discuss recent insightsinto chromosome segregation mecha-nisms gained from the analyses of forcegeneration during meiosis and mitosis.These analyses have demonstrated thatmembers of the kinesin superfamily andthe dynein family are essential in all or-ganisms for proper chromosome andspindle behavior. It is also apparent thatforces generated by microtubule polymer-ization and depolymerization are capableof generating forces sufficient for chro-mosome movement in vitro; whether theydo so in vivo is as yet unclear. An impor-tant realization that has emerged is thatsome spindle activities can be accom-plished,by more than one motor so thatfunctional redundancy is evident. In ad-dition, some meiotic or mitotic move-ments apparently occur through the co-operative action of independent semire-dundant processes. Finally, the molecularcharacterization of kinesin-related pro-teins has revealed that variations both inprimary sequence and in associationswith other proteins can produce motorcomplexes that may use a variety of mech-anisms to transduce force in associationwith microtubules. Much remains to belearned about the regulation of these ac-tivities and the coordination of opposingand cooperative events involved in chro-mosome segregation; this set of problemsrepresents one of the most important fu-ture frontiers of research.

"The cell has no other mode of origin thanby division of a preexisting cell" (E. B.Wilson, ref. 1). As Wilson and his con-temporaries realized over a century ago,the primary function of cell division (mi-

The publication costs of this article were defrayedin part by page charge payment. This article musttherefore be hereby marked "advertisement" inaccordance with 18 U.S.C. §1734 solely to indicatethis fact.

tosis) is to ensure the equipartition ofchromosomes between the two daughtercells; meiosis uses similar structures andmechanisms to accomplish an accuratereduction of chromosome number. In thisreview, we will describe recent advances inour understanding of the role of microtu-bule motors in chromosome segregationin mitosis and the related process nieiosis.In particular, we will focus on the likelyactions of members of the microtubule-associated molecular motor superfamilieskinesin and dynein, which convert thechemical energy in ATP into force andmovement during cell division (2, 3). Infact, it is safe to state that the rapiddiscoveries of numerous members of thekinesin superfamily and the identificationand molecular cloning of cytoplasmicforms of dynein have rekindled the fervorof efforts'to sort out the mechanism,s ofmitosis and meiosis (4-11).We will begin with a brief discussion of

the structure and behavior of the mitoticand meiotic spindles and note those eventsthat are likely to use molecular motors.We will then summarize recent studies oflikely mitotic and meiotic motors and dis-cuss their possible roles. Finally, we willhighlight some of the gaps in our under-standing that must be filled before we canfully understand the basis for the accuratesegregation of chromosomes at mitosisand meiosis.

Microtubules and the Spindle

The goal of mitosis is to equally divide thechromatin of the mother cell between thedaughter cells. The goal is accomplishedby the mitotic spindle, a macromolecularassembly that attaches to, organizes, anddirects the movements of chromosomesduring mitosis. The spindle itself is a

highly dynamic structure, forming anewduring the cell cycle when it is time todivide and disassembling when division isdone (12-14). The mitotic spindle, and itssibling structure, the meiotic spindle, ismost prominently composed of microtu-bules. Microtubules are assembled fromheterodimeric tubulin subunits and dis-play both structural and kinetic polarity.This polarity is reflected in the designa-tion of the microtubule ends: the minus

end, where tubulin subunits are added andlost more slowly; and the plus end, wheretubulin subunits are rapidly added and lost

1735

(15). Spindle microtubules are orientedwith their minus ends at the spindle polesand their plus ends pointing toward thecenter of the spindle or the cell cortex(Fig. 1). In addition, the polarity of mi-crotubules is recognized by microtubulemotors. A motor is classified as eitherminus-end-directed, with an activity thatwould move a cargo from the plus end tothe minus end of the microtubule, orplus-end-directed, with an activity thatwould move a cargo from the minus end tothe plus end of a microtubule (Fig. 2a).Microtubules have been suggested eitherto generate forces for chromosome move-ment by actively changing polymer lengthor to transduce directed forces generatedby molecular motors such as kinesin ordynein.

Mitotic and Meiotic Phenomena

The stages of mitosis are well known, asare the' major events that define them.Although the demarcations between thesestages are not precise, they are useful forbreaking the process down into manage-able segments (shown in Fig. 1). The firstdivision of meiosis is generally composedof similar stages and events, the obviousdifference being that prophase is veryextended and includes a period in whichhomologous chromosomes pair and un-dergo recombination. As a result of therecombination events, most homologouschromosomes remain attached to eachother by virtue of chiasmata. These at-tachments then constrain the homologuesto adopt and maintain orientations toopposite spindle poles during promet-aphase and metaphase, thus ensuring re-ductional division at anaphase (16, 17).

Inspection of the mitotic process indi-cates'that there are several movementsthat must be accounted for, and whichmay be powered by microtubule motors:(i) separation of spindle poles duringprophase and the maintenance of centro-some separation during the later stages ofmitosis; (ii) movement of chromosomestoward, and away, from the poles duringprometaphase, metaphase, and anaphase;(iii) mov'ement of the spindle poles apartduring anaphase; and (iv) a recently dis-

Abbreviations: KLP, kinesin-like proteins;MCAK, mitotic centromere-associated kinesin.

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 27

, 202

1

1736 Review: Barton and Goldstein

Prophase: Centrosome separatiorDuplicated ( \II /centrosomes 0-separate. KAR3 family

&_4(8 ~~~~Cytoplasmic dynein

Prometaphase: MT capture by

Nuclear kineochoreenvelope breaks _

H , MCAK

down. MTa are Chro___m

captured by QcongressoAnkinetochores. congression

Chromosomes XkIpl

congress to

metaphase plate. MKLP1

Spindle stabilization~~~~~~~~bimCfamily

Metaphase: KA- ufam\ly

chromosomes (+)\XklplK/bm family

aligned at-Xk

metaphase lt kpplate.

Anaphase A: Chromosome to polelinkage betweenI movement

sister \ / CENP-Echromatids is - °fmsevered and / KAR3 family

chromosomes abegin to move tothe poles.

Sp2indle elongationAnaphase B( MKLPI

elongates as

distance / \ Cytoplasmic dynein

poles increases. /

FIG. 1. Stages of mitosis (left column), major microtubule-dependent movements (centerdiagrams), and motors that have been suggested to play a role in these movements (right column).The motor functions presented represent suggested roles for the various motors, and theirassignment is based on the experimental evidence discussed in the text.

covered "flux" of tubulin subunits in themicrotubules of the spindle (e.g., as a

result of the addition of tubulin subunits atthe plus end and removal of subunits atthe minus ends in an assembled spindle).Any of these movements could actually bepowered by one or more motors, and inmany cases could be the sum of oppositelyacting forces as well. In addition, micro-tubule polymerization and depolymeriza-tion have long been recognized as poten-tial sources of force (12, 18-22).

In principle, a minimum set of forcegenerators would include a plus- and a

minus-end-directed motor at the chromo-somes, a plus-end-directed motor in thepole to drive microtubule flux or spindlepole migration in the cytoplasm, a plus-end-directed motor between overlappingmicrotubules in the central spindle topower spindle elongation, and perhaps a

minus-end-directed motor acting in thecytoplasm to drive spindle pole move-ments (Fig. 2). Recent studies of molecu-lar motors have turned up many candidatemotors to drive these events, and in somecases they have turned up new ways oflooking at motor function in mitosis andmeiosis (5, 6, 9, 23-29).

Potential Mitotic and Meiotic Motors

There are two classes of microtubule-based motors, the dyneins and the kine-sins. These two superfamilies of motorsdiffer in most significant properties, in-cluding primary sequence, size, pharma-cology, and in general, direction of move-ment along the polar microtubule (2, 30,31).True dynein was discovered over 30

years ago in cilia and flagella, where itpowers sliding displacement forces in theaxoneme by generating minus-end-di-rected microtubule movement (32-34).The existence of cytoplasmic forms ofdynein was confirmed with the purifica-tion of dynein from bovine brain (35) andCaenorhabditis elegans (36). These discov-eries suggested that cytoplasmic dyneinmay power organelle movements and per-haps mitotic movements as well. Unlikethe diverse number of kinesin-relatedgenes, thus far there appears to be but onecytoplasmic dynein heavy chain gene inmost organisms (37-42). Recent evidencesuggests that cytoplasmic dynein func-tional diversity is achieved through theassociation of the heavy chain with a va-

riety of accessory proteins that target and

regulate its activity (8, 43, 44). The mo-lecular characterization of the dynein su-perfamily is progressing, yet the extent towhich different members have diverged instructure and function is obscure.True kinesin was first found in neural

tissue, where it appears to generate plus-end-directed movements needed for ax-onal transport (45). Subsequent studieshave suggested that while true kinesinmight have no essential role in mitosis ormeiosis (46), other relatives in the kinesinsuperfamily do have essential mitotic andmeiotic functions. These kinesin-like pro-teins (KLPs) share a common core motordomain attached to different tail domains,which are thought to confer diverse asso-ciations with other cellular constituentsand hence different functions (Table 1;refs. 4, 82-84). On the basis of the con-servation ofprimary sequence within theirkinesin motor domains, one can definefour major families of meiotic/mitotic mo-tors: the bimC* family, the kar3 family, thekif2 family, and the kif4 family (Table 1;refs. 4, 29, 70, 84). In addition, there arethree "orphan" KLPs that may eventuallydefine additional families of mitotic andmeiotic motors. These include NOD (85),MKLP1 (86), and CENP-E (87).The bimC Family: bimC, cut7, cin8, kip],

Eg5, and klp61F. Phenotypic analyses ofmutants in the bimC family (bimC, As-pergillus nidulans; cin8/kipl, two near-equivalent and redundant bimC familymembers in Saccharomyces cerevisiae;cut7, Schizosaccharomyces pombe; andklp61F, Drosophila melanogaster), and ofEg5 (Xenopus laevis) immunodepletionexperiments, suggest that the bimC family-like KLPs play homologous roles in one ofthe earliest events of mitotic spindle as-sembly, namely the proper separation ofduplicated spindle poles (refs. 46-51); forreview, see ref. 25. In support of the ideathat members of this family perform sim-ilar functions, the two bimC family KLPsthat have been analyzed biochemically,EgS and KLP61F, show very similar invitro motility and microtubule interactioncharacteristics (50, 88, 89). These bio-chemical characteristics are also shared byKRP130, possibly a homologue of EgS,purified from Drosophila embryo extracts(52). In addition, immunolocalization ofCUT7 (90), Eg5 (50, 91), KLP61F (89),and epitope-tagged CIN8 or KIP1 (48,49), shows that these bimC family mem-bers are distributed along the length ofspindle microtubules during mitosis. Addi-

*To avoid confusion caused by organism-specific genetic nomenclatures, we haveadopted a simplified nomenclature system forthis review. Wild-type genes will be written initalicized lowercase letters (e.g., bimC). Refer-ences to a mutant gene will include the word"mutant" or "mutation" (e.g., "bimC mutants"or "mutations in bimC"). References to theprotein encoded by the gene will be written incapital letters (e.g., BIMC).

Proc. Natl. Acad. Sci. USA 93 (1996)

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 27

, 202

1

Proc. Natl. Acad. Sci. USA 93 (1996) 1737

(-)9 4(+) (MINUS END DIRECTED MOTOR PLUS END DIRECTD MOTOR

b W+ _ W+ -

PLUS END DIRECTED MOTOR GENERATES SLDING BETWEENANTI-PARALLEL MICROTUBULES

cTUBUUN SUBUNIT TUBUUN SUBUNITLOSS )

ADDITION

P.LU -EN .I-

PLUS END DIRECTED MOTOR DRIVES MICROTtsULE FLUX

d

e(-) <(+) 1 ~ NMOTOR COUPLES CHROMOSOMAE MOVMNTO MICROTUBULE DEPOLYMERIZATION

e W+ +

(-) (+) (+)MINUS END DIRECTED MOTOR FOCUSES MICROTORULE MINUS ENDS

FIG. 2. Some possible mechanisms that motors may use to generate a variety of microtu-bule-associated movements. (a) Motors can attach to a cargo (including vesicles and chro-mosomes) and transport it along the microtubule. Minus-end-directed motors move the cargofrom the plus end to the minus end of the microtubule; plus-end-directed motors move thecargo from the minus end to the plus end of the microtubule. (b) A plus-end-directed motorcross-links antiparallel microtubules and causes them to slide apart as the motor tries to moveto the microtubule plus ends. (c) A plus-end-directed motor could drive microtubule flux. Astubulin subunits are added to the microtubule plus end, the plus-end-directed motor, anchorednear the microtubule minus end, generates force against the microtubule leading to the lossof subunits from the microtubule minus end. (d) A motor couples chromosome movement tochanges in microtubule assembly rates. As the microtubule shortens due to loss of tubulinsubunits at its plus end, chromosome-associated microtubule motors maintain a dynamicattachment to the end of the shortening microtubule. As the microtubule shortens, the attachedchromosome moves to the microtubule minus end. (e) A minus-end-directed motor could focusthe minus ends of microtubules. The motor cross-links parallel microtubules and moves towardtheir minus ends, drawing them together. (f) Minus-end-directed and plus-end-directedmotors work together to move a cargo toward the cell periphery. The minus-end-directedmotor, anchored at the cell cortex, pulls on the microtubule plus end as it tries to move to themicrotubule minus end. The plus-end-directed motors, anchored to the cargo (i.e., centro-some), drive loss of tubulin subunits from the microtubule minus ends, causing the microtubuleto shorten. As the microtubule shortens, the cargo moves toward the anchored minus-end-directed motor.

tional similarities between BIMC, CUT7,KLP61F, and Eg5 emerged from the iden-tification of a conserved sequence elementin their globular carboxyl-terminal tail do-mains, further suggesting that these proteinsare involved in similar types of interactions(51). Using site-directed mutagenesis andtransient transfection assays, Sawin andMitchison (92) have analyzed the functionalsignificance of a highly conserved threonineresidue in this conserved region ofXenopusEgS. This threonine is a potential phosphor-ylation site for a p34Cdc2 or mitogen-activated protein kinase, and changing it to

a serine has no observable effect on thespindle localization of Eg5; however, whenEgS has an alanine or glutamic acid at thisposition, it does not localize to the spindleand instead remains cytoplasmic (92). Theseresults strongly suggest that this small do-main plays an important role in the propertargeting of EgS motor activity.Although the primary function of bimC

family members is generally considered tobe the separation of spindle poles, addi-tional genetic analyses of cin8/kipl func-tion in S. cerevisiae, using a temperature-sensitive mutant allele of cin8, revealed

that the proteins encoded by these geneshave important roles throughout mitosis(54). This work demonstrated that CIN8/KIP1 activity is necessary to stabilize thespindle prior to anaphase by providing a

counterbalance to forces that may be gen-erated by minus-end-directed motors, in-cluding members of the kar3 family of KLPs(53, 54). Specifically, when a kipl-deletion/cin8-temperature sensitive mutant strainwas shifted to non-permissive temperature,after spindle pole body separation occurredbut prior to anaphase, the separated spindlepoles collapsed back together (54). How-ever, deletion of kar3 in this genetic back-ground partially suppressed the collapsedspindle phenotype of cin8/kipl mutants,suggesting that the activities of KAR3 andCIN8/KIP1 are antagonistic (54). Finally,recent work demonstrated that CIN8/KIP1appear to work in concert with cytoplasmicdynein to drive anaphase B spindle elonga-tion (55). These results raise the possibilitythat other members of the bimC family mayserve similar roles during mitosis in otherorganisms.The kar3 Family: kar3, IdpA, ncd, and

kat4. All kar3 family members have theconserved kinesin motor domain element atthe carboxyl terminus; this is in contrast tomost kinesin-related proteins, which haveamino-terminal or central motor domains(4, 84). In addition, KAR3, NCD, andCHO2, the kar3 family members with dem-onstrated motor activity, generate minus-end directed motility in vitro (57, 58, 60, 93).

kar3 and ncd are the most extensivelycharacterized KLPs of this family. Geneticanalysis of kar3 reveals that it performs a

redundant function in mitosis and an es-sential function in karyogamy, suggestingthat it, and possibly other members of thisfamily, perform a task fundamental to anumber of microtubule-based processes(56). In mitosis, KAR3 appears to providea counterbalance to the activities of thebimC-related KLPs CIN8 and KIP1 (53,54). A similar motor interaction is seen inAspergillus, where the activity of KLPAmay antagonize the activity of BIMC dur-ing mitosis (59). Analyses of KAR3 (56,93), NCD (57,94), KLPA (59), and CHO2(60) suggest that kar3 family members maypossess a microtubule cross-linking activ-ity that contributes to the assembly andstabilization of the spindle.The importance of this cross-linking

activity is clearly seen in the analyses ofNCD function during the assembly of thefemale meiotic spindle in Drosophila (27,62, 95, 96). As shown by Theurkauf andHawley (97), this spindle is assembled"inside out," apparently by drawing to-gether the ends of chromosome-capturedmicrotubules to form focused poles thatlack centrioles (Fig. 2e). In ncd mutants,the analyses of fixed oocytes shows thatthe meiotic spindles often appear barrel-shaped, without focused poles, suggestingthat NCD function is required for proper

L+ (-W- _-MINUS END DIRECTED MOTOR PULLS ON MICROTUBULE PLUS END. -

PLUS END DIRECTED MOTORS PROMOTE LOSS OF TUBUUNSUBUNITS FROM THE MICROTUBULE MINDS END.

Review: Barton and Goldstein

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 27

, 202

1

1738 Review: Barton and Goldstein

Table 1. Major meiotic/mitotic microtubule motor families

FAMILY PROPERTIES FUNCTION(S)

bimC FamilybimC A. nidulans (46) Spindle pole separation (see text)cut7 S.pombe (47)Cin8 S.cerevisiae (48) SI.ZEKip8 S.cerevisiae (49) | 120-135 kDa Spindle stabilization (53, 54)

Eg5 X.Iaevis (50) MOTILITYKLP61 F D. melanogaster (51) () Anaphase B movements (55)KRP130 D. melanogaster (52) 1-2 gm/min.

kar3 Family --Ak ar Fail _ Anti-parallel MT sliding (56,57)

Kar3 S.cerevisiae (56)ncd D.melanogaster (57, 58) SIZEKIpA A.nidulans (59) 65-85 kDa Focus MTs at spindle poles (62, 63)

CH02 CHO cell line MOTILITY Anaphase A motor (64, 65)(Chinese hamster ovary) (60) (

katA A.thaliana (61) 1-15 ,um/min. Spindle stabilization (53)

KIF4 Family Stabilize bipolar spindle (67)

KIF4 M. musculus (66) SIZE Central spindle assembly (67, 68)

Xklpl X. Iaevis (67) 140 kDa Vesicle transport (70)KLP3A Dmelanogaster (68) MOTILITYChromokinesin G. domesticus (69) NA. Signal to begin cytokinesis (68)

KIF2 Family Modulation of kMT dynamics (71)MCAK Chinese hamster

ovary cell line (CHO) (71) 80-90 kDaKIF2 M. musculus (66, 72) MOTILITY Vesicle transport (72)

(+)28 gm/min. (72)

Cytoplasmic DyneinS.cerevisiae (38) Centrosome separation (79)R. norvegicus (39, 40) SIZEA. nidulans (73)AnpaeAmtr(0 1D. discoideum (37) 450-550 kDa Anaphase A motor (80, 81)N. crassa (74) MOTILITYD. melanogaster (75) (H Anaphase B motor (55)T. gratilla (76) 75 gm/min. (35)A. crassispina (77)75Jmm.()P. tetraurelia (78)_ _ _ _ _ _ _ _ _

The motility rates presented for the bimC and KAR3 families represent the range of velocitiesreported for those family members with demonstrated in vitro motility. The motor structurecartoons show each family member as a homodimer. For the KLPs, the amino terminus is to theleft. The dark balls represent the conserved, kinesin motor domain or "head"; the linked ovalsrepresent the coiled-coil dimerization domain or "stalk"; and the remaining sequence is referredto as the "tail" domain. For cytoplasmic dynein, the molecular structure is less clear, but the motordomain is also shown as a dark ball. The proposed functions represent suggested roles based onthe experimental evidence presented in the text. MT, microtubule; KMT, kinetochore MT.Complete species names are Aspergillus nidulans, Schizosaccharomyces pombe, Saccharomycescerevisiae, Xenopus laevis, Drosophila melanogaster, Arabidopsis thaliana, Mus musculus, Gallusdomesticus, Rattus norvegicus, Dictyostelium discoideum, Neurospora crassa, Tripneustes gratilla,Anthocidaris crassispina, and Paramecium tetraurelia.

pole formation (62, 95, 96). Recently,real-time analyses of spindle formation inncd mutants reveal that these spindlesundergo dynamic morphological transi-tions during metaphase of meiosis I, fromnormal-appearing spindles to dramaticallyunstable, multipolar spindle configura-tions (H. J. G. Mathies, H. McDonald,L.S.B.G., and W. E. Theurkauf, unpub-lished results). The ability of ncd mutantsto transiently assemble a normal-looking

meiotic spindle suggests that NCD func-tion may be partially redundant with thecross-linking activity of another KLP. Im-portantly, this type of motor function wasfirst suggested for cytoplasmic dynein byVerde and co-workers (98), who deter-mined that cytoplasmic dynein is requiredfor the in vitro assembly of microtubuleasters in Xenopus extracts. Therefore, al-though the pole-forming role of NCD maybe redundant with a dynein-like activity,

the live analyses support the suppositionthat NCD's primary function is the stabi-lization of the assembled spindle and itspoles (ref. 63; H. J. G. Mathies, M. Mc-Donald, L.S.B.G., and W. E. Theurkauf,unpublished results). In this respect, mem-bers of the bimC family and the kar3family may have similar, though counter-balanced, mitotic roles as spindle-stabiliz-ing elements (6, 53).

Finally, the analyses of kar3 suggest thatthe activity of this motor may be adaptedto perform a variety of functions duringmitosis. The characterization of recombi-nant KAR3, KAR3 fusion proteins, andkar3 mutants suggest that this KLP can,depending on the conditions, have a vari-ety of effects on microtubules. First, re-combinant KAR3 protein has been shownto both cross-link taxol-stabilized micro-tubules and destabilize microtubule minusends in vitro (93). The cross-linking activ-ity could be used to generate the sliding ofanti-parallel microtubules in vivo, whilethe microtubule-destabilizing activity sug-gests that KAR3 could facilitate microtu-bule flux during mitosis (93). Second, theexpression of KAR3-,B-galactosidase fu-sion proteins in S. cerevisiae results in theenhanced stabilization of cytoplasmic mi-crotubules (56). Meluh and Rose (56)suggest that this activity is the result of,B-galactosidase-induced oligomerizationof these KAR3 fusion proteins. However,it is unclear whether these KAR3 fusionproteins retain motor activity, and there-fore it is not known if motor activity isrequired for the microtubule-stabilizingactivity. Nonetheless, these results lead tothe supposition that KAR3 (and perhapsother KLPs) could stabilize the spindle invivo by forming higher-order complexes.Together, these results suggest that mem-bers of the kar3 family are involved pri-marily in microtubule cross-linking-basedmotor activities (Fig. 2 b and e) that areused to generate microtubule movementand spindle stabilization.The KIF4 Family: KIF4, XKLP1, Chro-

mokinesin, and kIp3A. The KIF4 family isa newly discovered family of mitotic KLPs(29, 70). Xenopus XKLP1 and mouseKIF4 are very similar throughout theirentire length (approximately 60% identi-cal), suggesting that these two proteinsmay be homologues (67). The kinesin-likemotor domains of Drosophila KLP3A (68)and chicken chromokinesin (69) are verysimilar to those of XKLP1 and KIF4, andthere are two motifs in the nonmotordomains of these KLPs that are conservedbetween XKLP1, KIF4, KLP3A, andchromokinesin (29). These include a shortsequence that contains putative nuclearlocalization signals and a highly conservedcysteine-rich sequence that resembles azinc finger motif.

Despite these KLPs having very similarprimary structures, their localization pat-terns are somewhat dissimilar. While

Proc. Natl. Acad. Sci. USA 93 (1996)

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 27

, 202

1

Proc. Natl. Acad. Sci. USA 93 (1996) 1739

XKLP1 is primarily localized on chromo-some arms (67), KIF4 is apparently vesicleassociated throughout mitosis (8). Duringmale meiosis, Drosophila KLP3A is onlytransiently associated with chromatin inearly prophase, becomes excluded fromchromatin during late prophase, and isdispersed in the cytoplasm from meta-phase to anaphase (68). At the transitionfrom anaphase to telophase, a fraction ofXKLP1 and chromokinesin and essen-tially all of KLP3A become localized tothe developing midbody (67-69).

Functional analyses of XKLP1 in Xeno-pus embryos and in vitro spindle assemblyextracts suggest that this KLP is primarilyinvolved in the positioning of chromo-somes and the stabilization of the assem-bled bipolar spindle (67). Injection ofXKLP1 antisense oligonucleotides intoXenopus embryos blocks cytokinesis andresults in the formation of abnormal spin-dles with displaced chromosomes. Thisphenotype and the localization of XKLP1to chromosomes prompted Vernos andco-workers (67) to suggest that XKLP1may have a role, similar to that proposedfor NOD, in generating an anti-polewardforce during chromosome congression. Inaddition, after introduction of XKLP1-specific antisera, Xenopus egg extracts as-semble mitotic spindles that contain a

reduced density of microtubules in thecentral spindle (67). Drosophila klp3A mu-tations also lead to the assembly of spin-dles with a severely decreased density ofmicrotubules in the central spindle (68).The numerous microtubules of the centralspindle are generally believed to be in-volved in the antiparallel microtubule slid-ing that results in the elongation of thespindle at anaphase B. Therefore, the lossof central spindle microtubule density inthe absence of XKLP1 or KLP3A activitywould be expected to diminish anaphase Bspindle elongation. Surprisingly, despitethis disruption of central spindle organi-zation, anaphase B elongation appears tobe unaffected in klp3A mutants (68).The biochemical analyses of the KIF4

family suggests that these KLPs also as-semble into a variety of different motorcomplexes. Xenopus XKLP1 appears to beable to associate with both chromosomesand membrane-bound vesicles (67).Mouse KIF4 appears to associate primar-ily with membrane-bound vesicles, but invitro analyses of the recombinant proteinshow that it has a strong microtubulebundling activity (70). Finally, DrosophilaKLP3A is not obviously associated withvesicles or chromosomes during meiosis ormitosis (68). However, its localization tothe central spindle suggests that it mightpromote interactions between microtu-bules, consistent with the in vitro activityof recombinant KIF4. Therefore, the lo-calization of KIF4 family members tochromosomes, vesicles, and central spin-dle microtubules raises the possibility that

these KLPs may have the capacity to forma variety of complexes that participate insuch apparently diverse processes as spin-dle organization and vesicle transport pro-cesses during mitosis and meiosis. In ad-dition, the observation that KIF4 alsoappears to function in axonal transport(70) suggests that KLPs with importantmitotic functions may not be exclusivelymitotic motors.The KIF2 Family: KIF2 and MCAK.

KIF2 and MCAK are unusual in that thekinesin-like motor domains of these pro-teins are centrally located in their primarysequences (66, 71). Mitotic centromere-associated kinesin (MCAK) was identifiedin an immunological screen designed toidentify novel KLPs in a Chinese hamsterovary (CHO) cell line (71). Immunolocal-ization of MCAK shows it to be associatedwith the centromere from prophase throughanaphase. Its localization to both the innerand outer kinetochore plates promptedWordeman and Mitchison (71) to suggestthat MCAK plays a role in chromosomecongression or the regulation of kineto-chore microtubule end dynamics.The deduced amino acid sequence of

mouse KIF2 is approximately 65% similarto the sequence of MCAK over its entirelength (71). Functional analysis of KIF2has focused on its possible role in axonalvesicle transport and not on its potentialrole during mitosis. In adult brain, immu-noaffinity purification of KIF2 with aKIF2-specific monoclonal antibody re-veals that KIF2 is associated with a pop-ulation of vesicles that appear to be dis-tinct from synaptic vesicles (72). However,analysis of KIF2 protein expression showsthat KIF2 levels are highest in developingbrain and lowest in adult brain, possiblysuggesting a role for KIF2 in cell prolif-eration (72). Further analyses for bothMCAK and KIF2 are required to deter-mine the extent of functional overlap be-tween these two very similar proteins.NOD. Mutations in the gene encoding

NOD lead to the misbehavior of chromo-somes that have not experienced recom-bination during female meiosis in Dro-sophila melanogaster (98, 99). Classicalgenetic studies and confocal microscopicanalyses reveal that NOD function, al-though nonessential for exchange chro-mosomes, provides an activity necessaryonly for the proper segregation of achias-mate chromosomes (85, 97, 99). Recently,Afshar et al. (100) have shown that theNOD protein is localized along the lengthof condensed chromosome arms and thatsequences in the NOD nonmotor domainshow sequence-specific DNA binding invitro. This observation suggests that NODmay be a major component contributingto the phenomenon known as the polarejection force (21, 97, 101), providing a

counterbalance to poleward-directedforces likely positioned at the kinetochore(20, 21).

MKLP1. MKLP1 (mitotic kinesin-likeprotein) was identified by using an immu-nological approach designed to generatemonoclonal antibodies against compo-nents of CHO cell mitotic spindles (102,139). Immunofluorescent analysis ofMKLP1 distribution during mitosis showsit to be localized to the spindle poles atmetaphase and to the central spindle atanaphase (86). Biochemical analysis ofrecombinant MKLP1, cloned from HeLacells, suggests that it can bundle microtu-bules in vitro (86). Additionally, in a mod-ified motility assay, MKLP1 was able tocause Chlamydomonas axonemes to movealong microtubules grown from Tetrahy-mena pellicles. These properties, inter-preted in light of the immunocytochemis-try, prompted Nislow et al. (86) to suggestthat MKLP1 may be involved in generat-ing sliding forces between anti-parallelmicrotubules during anaphase B. How-ever, microinjection of an anti-MKLP1monoclonal antibody, CHO1, into PtK1cells or first-division sea urchin embryoscauses mitotic arrest before metaphase(103, 104). Analyses of the arrested cellssuggest that anti-MKLP1 partially inhibitschromosome congression and disrupts theconvergent organization of spindle micro-tubules near the poles. These results sug-gest that MKLP1 may use its microtubulecross-linking activity to stabilize mitoticspindle poles. Interestingly, anti-MKLP1has no effect on anaphase movements,suggesting that MKLP1 is either inacces-sible to the antibody during anaphase or isnot essential for anaphase progression(103, 104).CENP-E. The CENP-E protein was also

identified in an immunological screen de-vised to identify monoclonal antibodiesspecific for components of the humankinetochore (105). Analyses of CENP-Elevels show that the protein is under tightcell cycle control, as it is rapidly degradedat the end of mitosis (87, 106). The dis-tribution of CENP-E during mitosis showsit to be localized to kinetochores fromprometaphase until early anaphase. Atthis point, CENP-E shows a dramatic re-distribution to the central spindle andlater to the midbody at telophase (105).CENP-E's redistribution from the kinet-ochore to the central spindle may be con-trolled by phosphorylation-dependentregulation of its nucleotide-independentmicrotubule-binding domain (107).The injection of the CENP-E-specific

monoclonal antibody mAbl77 into pro-metaphase human pancreatic epithelialcells caused a block or delay at the meta-phase/anaphase transition of mitosis, butit did not disrupt spindle organization orthe attachment of the chromosomes to themetaphase spindle (105).The localization of CENP-E to the ki-

netochore prompted Lombillo et al. (108)to investigate the possibility that kineto-chore-associated CENP-E is involved in

Review: Barton and Goldstein

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 27

, 202

1

1740 Review: Barton and Goldstein

maintaining a dynamic attachment be-tween the kinetochore and the depoly-merizing plus ends of kinetochore-associated microtubules. This possibilitywas investigated in an assay system inwhich the association of the chromosomewith the ends of the depolymerizing mi-crotubule causes the attached chromo-some to move toward the microtubuleminus end, the direction of anaphase Amovement (see Fig. 2d). Interestingly,chromosome movement in this in vitroassay is ATP independent, suggesting thata nonconventional motor activity is in-volved in the process. Antibodies toCENP-E, but not antibodies to MCAK or

dynein, block the microtubule disassem-bly-dependent movement of chromo-somes and suggest that CENP-E is directlyinvolved in coupling changes in microtu-bule dynamics to chromosome movements(108).Although motor activity for recombi-

nant CENP-E has not been directly dem-onstrated, recently Thrower et al. (109)partially purified CENP-E from HeLacells; surprisingly, this preparation showsminus-end directed motility. If theCENP-E protein is responsible for thisminus-end-directed motility, this would bethe first instance of which we are aware inwhich a motor with its kinesin-like motordomain at the amino terminus producedminus-end-directed motility.

Dynein. Due in large part to the con-siderable size of dynein, the identificationand characterization of cytoplasmic dy-neins has not progressed as rapidly as theanalyses of members of the kinesin super-family. However, recent accomplishmentsin the field, most importantly the cloningof full-length dynein heavy chains from a

number of organisms, have opened thedoor to the full characterization of cyto-plasmic dynein and perhaps cytoplasmicdynein-related proteins (37, 39, 40, 41,73-77).The initial suggestion that cytoplasmic

dynein might play a role in mitosis comesfrom striking immunological studies thatshow cytoplasmic dynein localized to kin-etochores and diffusely associated withspindle poles and spindle fibers duringmitosis (80, 81). This localization to kin-etochores prompted the suggestion thatcytoplasmic dynein functions as an an-

aphase A motor, providing force for themovement of the chromosomes to thepoles (80, 81, 110, 111). However, recentgenetic analyses of dynein function duringmitosis in S. cerevisiae strongly suggest thatdynein activity is not essential for chro-mosome movement during mitosis (38).Instead, it appears that dynein function isnecessary for the proper alignment of themitotic spindle relative to the bud neck(38, 112). In addition, analyses of theinteraction of dynein with CIN8 and KIP1mutant-encoded proteins (members ofthe bimC family) reveal that these proteins

work cooperatively to achieve anaphase Bspindle elongation (55). These results ob-tained in the yeast system are consistentwith results obtained in higher eu-karyotes. The microinjection of motility-blocking dynein antibodies into mamma-lian PtK1 cells does not inhibit chromo-some movement, but it does cause thecollapse of separating centrosomes,strongly suggesting that cytoplasmic dy-nein is involved both in centrosome sep-

aration and stabilization of the pre-

anaphase spindle (79). Together, theseresults suggest that dynein is involved incentrosome separation, positioning of themitotic spindle, and anaphase B spindleelongation. However, it cannot be con-cluded that dynein activity is not involvedin chromosome movement but rather that,if involved, activity is likely to be redun-dant with another mitotic motor or seg-regation process.Dynamic Microtubules. It has been sug-

gested that microtubules can performwork by changing polymer length. Suchchanges in microtubule dynamics havelong been proposed to provide force forspindle and chromosome movement dur-ing mitosis (9, 12, 18, 19, 21). For example,there is evidence to support the idea thatgrowing microtubule exert an anti-poleward (polar ejection) force on chro-mosomes during prometaphase chromo-some congression (12, 19, 20). While it isnot difficult to imagine how the plus endof a growing spindle microtubule couldpush a chromosome towards the spindlemidzone, it is harder to understand how a

microtubule, depolymerizing at its plusend, could pull a chromosome from themetaphase plate toward the microtubuleminus end at the spindle pole. Neverthe-less, it has recently been shown that de-polymerizing microtubules can generateminus-end-directed movement of chro-mosomes in vitro (108, 113, 114). Surpris-ingly, this work has shown that kinesin andKLPs can attach a chromosome to thedepolymerizing end of a microtubule andmaintain this dynamic attachment even atMgATP concentrations that are too low topermit kinesin or dynein to generate force(108). This capacity of microtubules tomove chromosomes could be functionallyredundant with the activity of an anaphaseA motor in vivo. These results demon-strate that (i) microtubule depolymeriza-tion can generate chromosome move-ment; and (ii) kinesin-like motor proteinsmay not need to generate force to serve anactive role in chromosome segregation.

Emerging Principles

The recent analyses of the many microtu-bule motors that appear to function duringmitosis and meiosis provide a glimpse ofsome of the mechanisms that are likely tobe operating during chromosome segre-gation. These inferred motor activities

suggest a variety of ways that microtubulemotors could accomplish the numerousmovements involved in chromosome seg-regation. For example, the separation ofcentrosomes involves motors that are plus-end-directed (bimC family) and minus-end-directed (kar3 family, cytoplasmic dy-nein). A plus-end-directed motor couldfacilitate centrosome separation by gen-erating antiparallel sliding between micro-tubules initiated at each centrosome (Fig.2b). In addition, a plus-end-directed mo-tor, positioned at the centrosome, couldwork in concert with a minus-end-directedmotor, anchored at the cell cortex, tomove centrosomes apart (Fig. 2f). Foranastral spindles, minus-end-directed mo-tors (kar3 family, cytoplasmic dynein)could organize microtubule minus endsand stabilize spindle poles (Fig. 2e). Dur-ing prometaphase, plus-end-directed mo-tors (KIF4 family, NOD) located on chro-mosome arms could participate in chro-mosome congression (Fig. 2a). Atanaphase A, various types of motors (kar3family, CENP-E) could couple chromo-some movements to changes in microtu-bule dynamics, allowing the chromosomesto move to the poles (Fig. 2d). Finally,plus-end-directed motors (bimC family,MKLP1) and minus-end-directed motors(cytoplasmic dynein) could carry out move-ments similar to those suggested for centro-some separation to accomplish anaphase Bspindle elongation (Fig. 2b and f).A recurring theme in studies of the

motors involved in mitosis and meiosis isthat some of the motor activities involvedin these movements are redundant. Whilesome of these activities are directly redun-dant (e.g., those encoded by cin8 and kiplin S. cerevisiae), other activities are appar-ently redundant because spindles may si-multaneously use different, cooperativemechanisms to generate movement (Fig.2). For example, both antiparallel micro-tubule sliding (Fig. 2b) and minus-end-directed motor pulling (Fig. 2f) mecha-nisms may be used to accomplish spindlepole separation and anaphase B spindleelongation.

In thinking about these mechanisms, weneed to keep in mind an underappreciatedaspect of mitosis and meiosis, namely theirextraordinary accuracy. Errors in mitoticchromosome segregation occur at the rateof approximately 10-5 per generation,while errors in reductional segregation atmeiosis occur on the order of 10-3 pergeneration (16, 115). This unusual preci-sion may have been selected for over longperiods of evolutionary time. As a result,the slow rate of chromosome segregationmay reflect the presence and operationof an extensively redundant system opti-mized for accuracy.The variety of microtubule motor activ-

ities used to segregate chromosomes ap-pears to include activities positioned at thekinetochore, between overlapping micro-

Proc. Natl. Acad. Sci. USA 93 (1996)

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 27

, 202

1

Proc. Natl. Acad. Sci. USA 93 (1996) 1741

tubules, and at the spindle poles. How-ever, recent evidence suggests that motorspositioned in less obvious locations none-theless play important roles in meiosis andmitosis. These include (i) motors posi-tioned along the length of chromosomearms that appear to be involved in thestabilization of the bipolar spindle and theproper positioning and segregation ofchromosomes (Fig. 2a; refs. 67, 69, 100);and (ii) vesicle-associated motors involvedin the organization of the central spindleand perhaps the transmission of the signalto initiate cytokinesis (67, 68, 70). Fur-thermore, it appears that some KLPs,particularly members of the XKLP1 andKIF2 families, may be associated withmultiple "cargoes," ranging from vesiclesto chromosomes, during mitosis (67, 70-72). Some of these motors do not appearto be mitosis-specific, but rather recruitedto their mitotic roles during cell prolifer-ation. Later in development, these KLPsappear to function in other microtubule-dependent processes, including axonaltransport (Fig. 2a; refs. 70, 72).

Gaps in Our Understanding

The analyses of KLPs and cytoplasmicdynein have greatly extended our under-standing of the likely roles of microtubulemotors in mitosis and meiosis. However,these investigations have also illuminateda number of gaps that will serve as im-portant areas for further work in the fu-ture.The functionally redundant nature of

many of the activities involved in accuratechromosome segregation has made it dif-ficult to assign functions to individualKLPs (23, 24). Nonetheless, the identifi-cation of the specific motor or motorsinvolved in carrying out the numerousmovements of mitosis and meiosis willcontinue to be a major challenge. Classicalgenetic approaches, particularly in yeast,have proven to be extremely valuable inidentifying the functions of cin8/kipl,kar3, and cytoplasmic dynein (53-55, 116).However, in systems where multipleknockouts are more difficult or not pos-sible with genetic approaches, we will haveto look for other approaches. In this re-gard, the identification of KLP familieswith distinct motor properties may lead tothe identification of KLP "family-specif-ic" inhibitors that can be used to target theelimination of an entire KLP family dur-ing cell division. The ability of injected"pan" antibodies to inhibit mitosis, whendirected against the most conserved re-gions of the kinesin motor domain, andhence against multiple kinesins simulta-neously, shows the potential for this typeof approach (104, 117, 118).To determine how motor activities are

targeted to specific spindle locations, weneed to characterize active, native motorcomplexes. Such work has already pro-

vided evidence that the force-generatingcapacity of the conserved kinesin motordomain itself has been modulated duringevolution so that it can produce signifi-cantly different motor activities, includ-ing the following: movements at varyingvelocities; movement toward either mi-crotubule plus or minus ends (i.e.,changes of direction); sliding betweenantiparallel microtubules; modulation ofmicrotubule assembly dynamics; and as-sociation with the ends of dynamicallyunstable microtubules. These studieshave also led to the view that variationsin motor output generated by the con-served kinesin-like motor domain aredependent not only on its motor and tailsequences but also on localization, asso-ciations with interacting proteins, andoligomeric architecture. The importanceof these latter concepts has been clearlydemonstrated by the pioneering work ofScholey and co-workers (119, 120). Us-ing a biochemical approach, they puri-fied an active heterotrimeric microtubulemotor complex (KRP85/95) from Strongy-locentrotus purpuratus that contains twodifferent kinesin-like motor subunits,SpKRP85 and SpKRP95, and a 115-kDaassociated protein. The novel feature ofKRP85/95 is the ability of its two kinesin-like subunits to heterodimerize, possiblythrough their coiled-coil stalk domains(121). This may be a general principle,since related KLPs have been identifiedin mouse (66, 122, 123), Drosophila(124), Chlamydomonas (125), and Cae-norhabditis (44, 124), suggesting thatthese KLPs are members of a new kinesinfamily (4, 84). Therefore, the potentialfor KLPs to form different types of oli-gomers is another mechanism that maybe used to generate variation in motoroutput, and it stresses the importance ofidentifying and characterizing native mo-tor protein complexes (126).Another poorly understood facet of

chromosome segregation is the mecha-nism that coordinates and regulates themultitude of motor activities that areapparently working during meiosis andmitosis. Recent cytological analyses haveprovided strong evidence suggesting thatthe generation and sensing of tension onthe spindle is an important "checkpoint"mechanism used to verify that chromo-somes have made a proper bipolar con-nection to the spindle prior to the initi-ation of anaphase (127-131). This ten-sion is quite possibly generated byopposing microtubule motor activities(6). In addition, there is evidence tosuggest that the phosphorylation state ofcertain kinetochore components is sen-sitive to tension (132, 133). These obser-vations lead to the supposition that theactivity or output of meiotic and mitoticmotors may also be regulated by, orsense, changes in tension (128). If micro-tubule motors sense and respond to

changes in tension along microtubules,this property would effectively enablemotors at the spindle poles to commu-nicate, through tension on spindle mi-crotubules, with motors on the chromo-somes and, through tension on astralmicrotubules, with motors localized atthe cell cortex.

Finally, many of the signals that lead tocell proliferation must ultimately be"read-out" by changes in the activities ofthe motor proteins that generate forcesduring cell division. Thus, it will be im-portant to focus attention on those regu-latory events that integrate motor activityinto the signal transduction cascades ofthe cell cycle.

L.S.B.G. is an Investigator with the HowardHughes Medical Institute.

1.

2.

3.

4.

5.

6.

7.8.

9.

10.

11.

12.

13.14.

15.

16.

17.18.

19.

20.

21.

22.

23.

24.25.

26.

27.28.

29.

30.

Wilson, E. B. (1925) The Cell in Developmentand Inheritence (Macmillan, New York), 2ndEd.Walker, R. A. & Sheetz, M. P. (1993) Annu.Rev. Biochem. 62, 429-451.Endow, S. A. & Titus, M. A. (1992) Annu. Rev.Cell. Bio. 8, 29-66.Goldstein, L. S. B. (1993) Ann. Rev. Genet. 27,319-351.Sawin, K. E. & Scholey, J. M. (1991) Trends CellBio. 1, 122-129.Fuller, M. T. & Wilson, P. G. (1992) Cell 71,547-550.Vallee, R. (1991) Trends Cell Biol. 1, 25-29.Schroer, T. A. (1994) Curr. Opin. Cell Bio. 6,69-73.McIntosh, J. R. & Pfarr, C. M. (1991) J. CellBiol. 115, 577-585.Roof, D. M., Meluh, P. B. & Rose, M. D. (1991)Cold Spring Harbor Symp. Quant. Biol. 56, 693-703.Sawin, K. E. & Endow, S. A. (1993) BioEssays15, 399-407.Salmon, E. D. (1989) Mitosis: Molecules andMechanisms, eds. Hyams, J. S. & Brinkley, B. R.(Academic, London), pp. 119-181.Karsenti, E. (1991) Semin. Cell Biol. 2, 251-260.Wadsworth, P. (1993) Curr. Opin. Cell Biol. 5,123-128.Cassimeris, L. U., Walker, R. A., Pryer, N. K. &Salmon, E. D. (1987) Bioessays 7, 149-154.Hawley, R. S. (1988) Genetic Recombination,eds. R. Kucherlapati & Smith, G. R. (Am. Soc.Microbiol., Washington, DC), pp. 497-527.Carpenter, A. T. C. (1994) Cell 77, 959-962.Inoue, S. & Sato, H. (1967) J. Gen Physiol. 50,259-292.Salmon, E. D. (1989) Cell Movement: Kinesin,Dynein, and Microtubule Dynamics (Liss, NewYork), Vol. 2, pp. 431-440.Cassimeris, L., Rieder, C. L. & Salmon, E. D.(1994) J. Cell Sci. 107, 285-297.R;eder, C. L. & Salmon, E. D. (1994) J. CellBiol. 124, 223-233.Desai, A., and Mitchison, T. J. (1995) J. CellBiol. 128, 1-4.Goldstein, L. S. B. (1991) Trends Cell Biol. 1,93-98.Goldstein, L. S. B. (1993) J. Cell Biol. 120, 1-3.Saunders, W. S. (1993) Trends Cell Biol. 3,432-437.Vale, R. D. (1992) Trends Biochem. Sci. 17,300-304.Endow, S. A. (1993) Trends Genet. 9, 52-55.Hoyt, M. A. (1994) Curr. Opin. Cell Biol. 6,63-68.Vernos, I. & Karsenti, E. (1995) Trends CellBiol. 5, 297-301.Goldstein, L. S. B. & Vale, R. D. (1991) Nature(London) 352, 569-570.

Review: Barton and Goldstein

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 27

, 202

1

1742 Review: Barton and Goldstein Proc. Natl. Acad. Sci. USA 93 (1996)

31. Shimizu, T., Furusawa, K., Ohashi, S., Toyo-shima, Y. Y., Okuno, M., Malik, F. & Vale,R. D. (1991) J. Cell Biol. 112, 1189-1197.

32. Gibbons, 1. R. & Rowe, A. (1965) Science 149,424-426.

33. Gibbons, 1. R. (1988)JBiol. Chem. 263, 15837-15840.

34. Witman, G. B., Wilkerson, C. G. & King, S. M.(1994) Microtubules, eds. Hyams, J. S. & Lloyd,C. W. (Wiley, New York), pp. 229-249.

35. Paschal, B. M., Shpetner, H. S. & Vallee, R. B.(1987) J. Cell Biol. 105, 1273-1282.

36. Lye, R. J., Porter, M. E., Scholey, J. M. & McIn-tosh, J. R. (1987) Cell 51, 309-318.

37. Koonce, M. P., Grissom, P. M. & McIntosh,J. R. (1992) J. Cell Biol. 119, 1597-1604.

38. Eshel, D., Urrestarazu, L. A., Vissers, S., Jau-niaux, J.-C., Van Vliet-Reedijk, J. C., Planta,R. J. & Gibbons, I. R. (1993) Proc. Natl. Acad.Sci. USA 90, 11172- 11176.

39. Zhang, Z., Tanaka, Y., Nonaka, S., Aizawa, H.,Kawasaki, H., Nakata, T. & Hirokawa, N.(1993) Proc. Natl. Acad. Sci. USA 90, 7928-7932.

40. Mikami, A., Paschal, B. M., Mazumdar, M. &Vallee, R. B. (1993) Neuron 10, 787-796.

41. Gibbons, B. H., Asai, D. J., Tang, W.-J. Y.,Hays, T. S. & Gibbons, I. R. (1994) Mol. Biol.Cell 5, 57-70.

42. Rasmusson, K., Serr, M., Gepner, J., Gibbons,I. & Hays, T. S. (1994) Molec. Biol. Cell 5, 45-55.

43. Holzbaur, E. L. F., Mikami, A., Paschal, B. M.& Vallee, R. B. (1994) Microtubules, eds. Hy-ams, J. S. & Lloyd, C. W. (Wiley, New York),pp. 251-267.

44. Tabish, M., Siddiqui, Z. K., Nishikawa, K. &Siddiqui, S. S. (1995)J. Mol. Biol. 247, 377-389.

45. Vale, R. D., Reese, T. S. & Sheetz, M. P. (1985)Cell 42, 39-50.

46. Enos, A. P. & Morris, N. R. (1990) Cell 60,1019-1027.

47. Hagan, I. & Yanagida, M. (1990) Nature (Lon-don) 347, 563-566.

48. Hoyt, M. A., He, L., Loo, K. K. & Saunders,W. S. (1992) J. Cell Biol. 118, 109-120.

49. Roof, D. M., Meluh, P. B. & Rose, M. D. (1992)J. Cell Biol. 118, 95-108.

50. Sawin, K. E., LeGuellec, K., Philippe, M. &Mitchison, T. J. (1992) Nature (London) 359,540-543.

51. Heck, M. S., Pereira, A., Pesavento, P., Yan-noni, Y., Spradling, A. C. & Goldstein, L. S. B.(1993) J. Cell Biol. 123, 665-679.

52. Cole, D. G., Saxton, W. M., Sheehan, K. B. &Scholey, J. M. (1994) J. Biol. Chem. 269, 22913-22916.

53. Hoyt, M. A., He, L., Totis, L. & Saunders, W. S.(1993) Genetics 135, 35-44.

54. Saunders, W. S. & Hoyt, M. A. (1992) Cell 70,451-458.

55. Saunders, W. S., Koshland, D., Eshel, D., Gib-bons, I. R. & Hoyt, M. A. (1995) J. Cell Biol.128, 617-624.

56. Meluh, P.B. & Rose, M. D. (1990) Cell 60,1029-1041.

57. McDonald, H. B., Stewart, R. J. & Goldstein,L. S. B. (1990) Cell 63, 1159-1165.

58. Walker, R. A., Salmon, E. D. & Endow, S. A.(1990) Nature (London) 347, 780-782.

59. O'Connell, M. J., Meluh, P. B., Rose, M. D. &Morris, N. R. (1993) J. Cell Biol. 120, 153-162.

60. Kuriyama, R., Kofron, M., Essner, R., Kato, T.,Dragas-Granoic, S., Omoto, C. K. & Khodja-kov, A. (1995) J. Cell Biol. 129, 1049-1059.

61. Mitsui, H., Yamaguchi-Shinozaki, K., Shi-nozaki, K., Nishikawa, K. & Takahashi, H.(1993) Mol. Gen. Genet. 238, 362-368.

62. Hatsumi, M. & Endow, S. A. (1992) J. Cell Sci.101, 547-559.

63. Endow, S. A., Chandra, R., Komma, D. J.,Yamamoto, A. H. & Salmon, E. D. (1994) J.Cell Sci. 107, 859-867.

64. Hyman, A. A., Middleton, K., Centola, M.,Mitchison, T. J. & Carbon, J. (1992) Nature(London) 359, 533-536.

65. Middleton, K. & Carbon, J. (1994) Proc. Natl.Acad. Sci. USA 91, 7212-7216.

66. Aizawa, H., Sekine, Y., Takemura, R., Zhang,Z., Nangaku, M. & Hirokawa, N. (1992)J. CellBiol. 119, 1287-1296.

67. Vernos, I., Raats, J., Hirano, T., Heasman, J.,Karsenti, E. & Wylie, C. (1995) Cell 81, 117-127.

68. Williams, B. C., Riedy, M. F., Williams, E. VB.,Gatti, M. & Goldberg, M. L. (1995) J. Cell Biol.129, 709-723.

69. Wang, S.-Z. & Adler, R. (1995) J. Cell Biol. 128,761-768.

70. Sekine, Y., Okada, Y., Noda, Y., Kondo, S.,Aizawa, H., Takemura, R. & Hirokawa, N.(1994) J. Cell Biol. 127, 187-201.

71. Wordeman, L. & Mitchison, T. J. (1995) J. CellBiol. 128, 95-105.

72. Noda, Y., Sato-Yoshitake, R., Kondo, S., Nan-gaku, M. & Hirokawa, N. (1995) J. Cell Biol.129, 157-167.

73. Xiang, X., Beckwith, S. M. & Norris, N. R.(1994) Proc. Natl. Acad. Sci. (USA) 91, 2100-2104.

74. Plamann, M., Minke, P. F., Tinsley, J. H. &Bruno, K. S. (1994) J. Cell Biol. 127, 139-149.

75. Li, M., McGrail, M., Serr, M. & Hays, T. S.(1994) J. Cell Biol. 126, 1475-1494.

76. Gibbons, I. R., Gibbons, B. H., Mocz, G. &Asai, D. J. (1991) Nature (London) 352, 640-643.

77. Ogawa, K. (1991) Nature (Londoni) 352, 643-645.

78. Asai, D. J., Beckwith, S. M., Kandl, K. A., Keat-ing, H. H., Tjandra, H. & Forney, J. D. (1994) J.Cell Sci. 107, 839-847.

79. Vaisberg, E. A., Koonce, M. P. & McIntosh,J. R. ( 1993) J. Cell Biol. 123, 849 -858.

8(). Steuer, E. R., Wordeman, L., Schroer, T. A. &Sheetz, M. P. (1990) Nature (London) 345,266-268.

81. Pfarr, C. M., Coue, M., Grissom, P. M., Hays,T. S., Porter, M. E. & McIntosh, J. R. (1990)Nature (London) 345, 263-265.

82. Yang, J. T., Saxton, W. M., Stewart, R. J., Raff,E. C. & Goldstein, L. S. B. (1990) Science 249,42-47.

83. Vale, R. D. & Goldstein, L. S. B. (1990) Cell 60,883-885.

84. Bloom, G. & Endow, S. (1994) Protein Profile 1,issue 10.

85. Zhang, P., Knowles, B. A., Goldstein, L. S. B. &Hawley, R. S. (1990) Cell 62, 1053-1062.

86. Nislow, C., Lombillo, V. A., Kuriyama, R. &McIntosh, J. R. (1992) Nature (London) 359,543-547.

87. Yen, T. J., Li, G., Schaar, B. T., Szilak, I. &Cleveland, D. W. (1992) Nature (London) 359,536-539.

88. Sawin, K. E., Mitchison, T. J. & Wordeman,L. G. (1992) J. Cell Sci. 101, 303-313.

89. Barton, N. R., Pereira, A. J. & Goldstein,L. S. B. (1995) Mol. Biol. Cell, in press.

90. Hagan, I. & Yanagida, M. (1992) Nature (Lon-don) 356, 74-76.

91. Houliston, E., Le Guellec, R., Kress, M., Phil-ippe, M. & Le Guellec, K. (1994) Dev. Biol. 164,147-159.

92. Sawin, K. E. & Mitchison, T. J. (1995) Proc.Natl. Acad. Sci. USA 92, 4289-4293.

93. Endow, S. A., Kang, S. J., Satterwhite, L. L.,Rose, M. D., Skeen, V. P. & Salmon, E. D.(1994) EMBO J. 13, 2708-2713.

94. Chandra, R., Salmon, E. D., Erickson, H. P.,Lockhart, A. & Endow, S. A. (1993) J. Biol.Chem. 268, 9005-9013.

95. Wald, H. (1936) Genetics 21, 264-281.96. Davis, D. G. (1969) Genetics 61, 577-594.97. Theurkauf, W. E. & Hawley, R. S. (1992)J. Cell

Biol. 116, 1167-1180.98. Carpenter, A. T. C. (1973) Genetics 73, 393-

428.99. Zhang, P. & Hawley, R. S. (1990) Genetics 125,

115-127.100. Afshar, K., Barton, N. R., Hawley, R. S. &

Goldstein, L. S. B. (1995) Cell 81, 129-138.101. Carpenter, A. T. C. (1991) Cell 64, 885-890.102. Schroer, T. A. (1994) J. Cell Biol. 127, 1-4.

103. Nislow, C., Sellitto, C., Kuriyama, R. & McIn-tosh, J. R. (1990) J. Cell Biol. 111, 511-522.

104. Wright, B. D., Terasaki, M. & Scholey, J. M.(1993) J. Cell Biol. 123, 681-689.

105. Yen, T. J., Compton, D. A., Wise, D.,Zinkowski, R. P., Brinkley, B. R., Earnshaw,W. C. & Cleveland, D. W. (1991) EMBO J. 10,1245-1254.

106. Brown, K. D., Coulson, R. M., Yen, T. J. &Cleveland, D. W., (1994) J. Cell Biol. 125, 1303-1312.

107. Liao, H., Li, G. & Yen, T. J. (1994) Science 265,394-398.

108. Lombillo, V. A., Nislow, C., Yen, T. J., Gelfand,V. 1. & McIntosh, J. R. (1995) J. Cell Biol. 128,107-115.

109. Thrower, D. A., Jordan, M. A., Schaar, B. T.,Yen, T. J. & Wilson, L. (1995) EMBO J. 14,918-926.

110. Vallee, R. (1991) Nature (London) 351, 187-188.

111. Vallee, R. (1990) Nature (London) 345, 206-207.

112. Li, Y.-Y., Yeh, E., Hays, T. & Bloom, K. (1993)Proc. Natl. Acad. Sci. USA 90, 10096-10100.

113. Koshland, D. E., Mitchison, T. J. & Kirschner,M. W. (1988) Nature (London) 331, 499-504.

114. Coue, M., Lombillo, V. A. & McIntosh, J. R.(1991) J. Cell Biol. 112, 1165-1175.

115. Brown, M., Garvik, B., Hartwell, L., Kadyk, L.,Seeley, T. & Weinert, T. (1991) Cold SpringHarbor Symp. Quant. Biol. 56, 359-365.

116. Kilmartin, J. V. (1994) Curr. Opin. Cell Biol. 6,50-54.

117. Hogan, C. J., Wein, H., Wordeman, L., Scholey,J. M., Swain, K. E. & Cande, W. Z. (1993) Proc.Natl. Acad. Sci. USA 90, 6611-6615.

118. Rodionov, V. I., Gelfand, V. I. & Borisy, G. G.(1993)J. Cell Sci. 106, 1179-1188.

119. Cole, D. G., Chinn, S. W., Wedaman, K. P.,Hall, K., Vuong, T. & Scholey, J. M. (1993)Nature (London) 366, 268-270.

120. Cole, D. G., Cande, W. Z., Baskin, R. J., Skou-fias, D. A., Hogan, C. J. & Scholey, J. M. (1992)J. Cell Sci. 101, 291-301.

121. Rashid, D., Wedaman, K. P. & Scholey, J. M.(199S) J. Mol. Biol. 252, 157-162.

122. Kondo, S., Sato-Yoshitake, R., Noda, Y.,Aizawa, H., Nakata, T., Matsuura, Y. & Hiro-kawa, N. (1994) J. Cell Biol. 125, 1095-1107.

123. Yamazaki, H., Nakata, T., Okada, Y. & Hiro-kawa, N. (1995) J. Cell Biol. 130, 1387-1399.

124. Pesavento, P. A., Stewart, R. J. & Goldstein,L. S. B. (1994) J. Cell Biol. 127, 1041-1048.

125. Walther, Z., Vashishtha, M. & Hall, J. L. (1994)J. Cell Biol. 126, 175-188.

126. Cole, D.G., & Scholey, J. M. (1995) Trends CellBiol. 5, 259-261.

127. Nicklas, R. B. (1988) Annu. Rev. Biophys. Bio-phys. Chem. 17, 431-449.

128. Nicklas, R. B. & Ward, S. C. (1994)J. Cell Biol.126, 1241-1253.

129. Jang, J. K., Messina, L., Erdman, M. B., Arbel,T. & Hawley, R. S. (1995) Science 268, 1917-1919.

130. Li, X. & Nicklas, R. B. (1995) Nature (London)373, 630-632.

131. Murray, A. W. (1995) Nature (London) 373,560-561.

132. Campbell, M. S. & Gorbsky, G. J., (1995)J. CellBiol. 129,1195-1204.

133. Gorbsky, G. J. (1995) Trends Cell Biol. 5, 143-147.

134. Yeh, E., Skibbens, R. V., Cheng, J. W., Salmon,E. D. & Bloom, K. (1995) J. Cell Biol. 130, 687-700.

135. Saxton, W. M., Hicks, J., Goldstein, L. S. B. &Raff, E. C. (1991) Cell 64, 1093-1102.

136. Paschal, B. M. & Vallee, R. B. (1987) Nature(London) 330, 181-183.

137. Verde, F., Berrez, J.-M., Antony, C. & Karsenti,E. (1991) J. Cell Biol. 112, 1177-1187.

138. Shakir, M., Fukushige, T., Yasuda, H., Miwa, J.& Siddiqui, S. S. (1993) NeuroReport 4, 891-894.

139. Sellitto, C., & Kuriyama, R. (1988) J. Cell Biol.106, 431-439.

Dow

nloa

ded

by g

uest

on

Feb

ruar

y 27

, 202

1