f-actin ring formation in fission yeast · (for a review see le goff et al., 1999). some are...

IntroductionIn cytokinesis of animal cells, assembly of the actin-basedcontractile ring occurs in the cell cortex at the division planeduring late anaphase to early telophase, and the cells dividethrough contraction of the contractile ring (Mabuchi, 1986).It has been proposed that positional signals for contractilering assembly may be transferred from astral or centralmicrotubules to the cell cortex (Rappaport, 1986).Furthermore, it has generally been accepted that variousregulators including signaling or actin-binding proteins areinvolved in dynamic rearrangement of the actin cytoskeletonthroughout the cell cycle. Although various proteins that areinvolved in cytokinesis have recently been identified, little isknown about the molecular mechanism of the contractile ringassembly including determination of the division plane.

The fission yeast Schizosaccharomyces pombehas threesimple F-actin structures: cortical patches, cables and rings(Marks and Hyams, 1985). Dynamic behaviour of these actincytoskeletons during the cell cycle has been investigated(Marks and Hyams, 1985; Arai et al., 1998). The F-actinpatches localize to both growing ends of the cylindrical cellduring interphase. It is suggested that the F-actin patches playa role in the deposition of cell wall materials and in themaintenance of polarized cell growth (Kobori et al., 1989;

Ishiguro and Kobayashi, 1996). The localization of F-actinpatches has been shown to be controlled by small GTP-bindingproteins and various actin-modulating proteins (for reviews,see Ishiguro, 1998; Le Goff et al., 1999). The F-actin cablesrun longitudinally during interphase and seem to be linked tosome F-actin patches. During mitosis, the F-actin patchesdisappear from the ends of the cell, and the F-actin ring isformed at the medial region of the cell. The F-actin cables areseen to attach to the F-actin ring during mitosis (Arai et al.,1998). It has been proposed that F-actin ring formation initiatesbefore anaphase in S. pombe (Marks and Hyams, 1985),being different from animal cells (Mabuchi, 1994) orSchizosaccharomyces japonicus(Alfa and Hyams, 1990), inwhich the contractile ring is formed during late anaphase totelophase. Recently, it has been reported that the actin filamentsfirst appear as a faint and broad F-actin ring in the wild-typeS. pombecell before anaphase. However, how the actinfilaments accumulate at the medial region remains unknown.Then the ring becomes a sharp one as anaphase progresses(Arai et al., 1998; Bähler et al., 1998). As the ring shrinksduring cytokinesis, septum invaginates centripetally from thecell surface. The F-actin patches relocalize near the septumregion in this stage.

Several cell division mutants that cannot form the F-actin

887

Cells of the fission yeast Schizosaccharomyces pombedivideby the contraction of the F-actin ring formed at the medialregion of the cell. We investigated the process of F-actinring formation in detail using optical sectioning and three-dimensional reconstruction fluorescence microscopy. Inwild-type cells, formation of an aster-like structurecomposed of F-actin cables and accumulation of F-actincables were recognized at the medial cortex of the cellduring prophase to metaphase. The formation of the aster-like structure seemed to initiate from branching of thelongitudinal F-actin cables at a site near the spindle polebodies, which had been duplicated but not yet separated. Asingle cable extended from the aster and encircled the cellat the equator to form a primary F-actin ring duringmetaphase. During anaphase, the accumulated F-actincables were linked to the primary F-actin ring, and then allof these structures seemed to be packed to form the F-actinring. These observations suggest that formation of theaster-like structure and the accumulation of the F-actincables at the medial region of the cell during metaphase

may be required to initiate the F-actin ring formation. Inthe nda3 mutant, which has a mutation in ß-tubulin andhas been thought to be arrested at prophase, an F-actinring with accumulated F-actin cables similar to that ofanaphase wild-type cells was formed at a restrictivetemperature. Immediately after shifting to a permissivetemperature, this structure changed into a tightly packedring. This suggests that the F-actin ring formationprogresses beyond prophase in the nda3cells once the cellsenter prophase. We further examined F-actin structures inboth cdc12and cdc15early cytokinesis mutants. As a result,Cdc12 seemed to be required for the primary F-actin ringformation during prophase, whereas Cdc15 may beinvolved in both packing the F-actin cables to form theF-actin ring and rearrangement of the F-actin afteranaphase. In spg1, cdc7and sid2 septum initiation mutants,the F-actin ring seemed to be formed in order.

Key words: F-actin ring, F-actin cable, Aster-like structure,Deconvolution, 3D reconstruction

Summary

F-actin ring formation and the role of F-actin cables inthe fission yeast Schizosaccharomyces pombeRitsuko Arai 1 and Issei Mabuchi 1,2,*1Division of Biology, Department of Life Sciences, Graduate School of Arts and Sciences, University of Tokyo, Tokyo, 153-8902, Japan2Department of Cell Biology, National Institute for Basic Biology, Okazaki, 444-8585, Japan*Author for correspondence (e-mail: [email protected])

Accepted 19 November 2001Journal of Cell Science 115, 887-898 (2002) © The Company of Biologists Ltd

Research Article

888

ring and grow into large multinucleate cells at the restrictivetemperature have been obtained in S. pombe, and severalproteins that are involved in the F-actin ring formation havebeen reported and characterized by analyzing these mutants(for a review see Le Goff et al., 1999). Some are considered tofunction in early steps of the ring formation process. Plo1, apolo kinase homolog, is an important protein to regulateand initiate the F-actin ring formation. When Plo1 isoverexpressed, F-actin ring formation is induced even ininterphase (Ohkura et al., 1995). mid1/dmf1/pos1mutant cellsform F-actin rings and septa at a random position and angles(Chang et al., 1996; Sohrmann et al., 1996; Edamatsu andToyoshima, 1996). Thus, Mid1 is required to determine theposition of the F-actin ring formation. It localizes on thenuclear membrane during interphase (Sohrmann et al., 1996;Bähler et al., 1998), but relocalizes as a broad ring at the medialregion of the cell during prophase, and then the ring becomessharp (Bähler et al., 1998). Recently, it has been reported thatPlo1 is necessary for the proper Mid1 function (Bähler et al.,1998).

Mutants of the cdc12 and cdc15 genes, which encode adiaphanous family protein and an SH3-containing protein,respectively, cannot form the normal F-actin ring duringmitosis, but display normally localized F-actin patches duringinterphase (Fankhauser et al., 1995; Chang et al., 1996;Balasubramanian et al., 1998). Cdc12 localizes as a spot in thecytoplasm during interphase when it is overexpressed in the cell(Chang et al. 1997; Chang, 1999). During prophase, the Cdc12spot positions at the medial cortex. A single Cdc12 strandextends from the spot and then encircles the cell at the equator(Chang et al., 1997). However, when Cdc15 is overexpressed inG2-arrested cells, F-actin ring formation is induced (Fankhauseret al., 1995). These phenotypes suggest that both Cdc12 andCdc15 play specific roles, in an early step of F-actin ringassembly. In mutants of cdc3 and cdc8 genes, which encodeprofilin and tropomyosin, respectively, both formation of the F-actin ring and localization of the F-actin patches are impaired(Balasubramanian et al., 1992; Balasubramanian et al., 1994;Chang et al., 1996). Tropomyosin is also required inorganization of the F-actin cables (Arai et al., 1998).

Type II myosin heavy chains, Myo2 (Kitayama et al., 1997;May et al., 1997) and Myp2/Myo3 (Bezanilla et al., 1997;Motegi et al., 1997), and light chains Cdc4 (McCollum et al.,1995) and Rlc1 (Le Goff et al., 2000; Naqvi et al., 2000) arelocalized only to the F-actin ring and required in its formation.They are thought to be involved in a late step of the F-actinring assembly. An IQGAP-like protein has also beenconsidered to be involved in a late step of the F-actin ringassembly (Chang et al., 1996; Eng et al., 1998).

At the onset of cytokinesis, Spg1, a small GTP-bindingprotein, regulates initiation of F-actin ring contraction orseptum formation, or both, through controlling localization andfunction of downstream effectors, Cdc7 and Sid2 proteinkinases (Schmidt et al., 1997; Sohrmann et al., 1998; Sparkset al., 1999). It has been reported that the septum formationnever occurs in mutants of these genes, although the F-actinring is formed during mitosis (Fankhauser and Simanis, 1994;Schmidt et al., 1997; Balasubramanian et al., 1998).

We are interested in the molecular process of the early stepof the F-actin ring formation. From the above results, twodistinct events seem to occur in the early step of the F-actin

ring formation: one is accumulation of Mid1 followed by theaccumulation of F-actin as the medial broad ring, and then thering narrows. The other is positioning of the Cdc12 as a spotin the medial cortex. The spot then forms the ring by extendinga strand. However, the relationship between these eventsremains unclear. Furthermore, little is known about the processof structural rearrangement of F-actin during the F-actin ringformation including how actin filaments are arranged in the‘broad F-actin ring’. Thus, we investigated the process of theF-actin ring formation by optical sectioning and three-dimensional (3D) microscopy in the wild-type S. pombecellsand some cytokinesis mutant cells. Our observations suggestthat the F-actin ring formation is initiated by the accumulationof the F-actin cables and the formation of an aster-like structurecomposed of F-actin cables at the cell equator during prophase.

Materials and MethodsYeast strains and growth conditionsAll the strains were cultured in YPD liquid medium. JY1h– cells wereused as wild-type cells and grown to exponential phase at 30°C. nda3-KM311cold-sensitive mutant (National Collection of Yeast Cultures,Norwich, UK) were grown at 36°C, a permissive condition, and thenincubated at 20°C for 8 hours, a restrictive condition. Temperature-sensitive mutants, cdc7-24, cdc12-112, cdc15-140, spg1-B8(giftsfrom Viesturs Simanis, ISREC, Switzerland) and sid2-250(a gift fromDannel McCollum, Massachussetts University, USA), were grown at25°C and then incubated at 36°C, a restrictive condition, for 3-4 hours.

Fluorescence microscopyImmunostaining was performed as previously described (Arai et al.,1998). Mitotic stages were defined by localization of spindle polebodies (SPBs) according to Hagan (Hagan, 1998). Representativeimages are shown in the following figures. The SPBs were labeledwith rabbit anti-Sad1 antibody (a gift from Iain Hagan) and then withrhodamine-conjugated anti-rabbit IgG antibody (Cappel, Durham,NC). For F-actin and nuclear staining, Bodipy (4,4-difluoro-4-bora-3a,4a-diaza-s-indacene)-phallacidin (Molecular Probes Inc., Eugene,OR) and 4′,6-diamidino phenylindole (DAPI) were used, respectively.For usual observations, the specimens were examined under a ZeissAxioscope fluorescence microscope using a Plan Apochromat 63×lens and then photographed on Kodak T-Max films of ASA 400 (Fig.6a-h). To obtain deconvoluted optical sections and 3D reconstructedimages, we used a Delta Vision system (Applied Precision Inc.,Issaquah, WA) with an Olympus IX70 fluorescence microscopeequipped with a UPlanApo 100× lens according to the manufacturer’sprotocol (Figs 1-9). Original fluorescence micrographs of serialoptical sections were taken every 0.2 µm in the direction of Z-axisand digitized. The original images were deconvoluted, and thenreconstructed to obtain 3D images of the cells. For the 3Dreconstruction, we examined two calculation modes, ‘Max Intensity’and ‘Additive’. In the former mode, major fluorescences wereemphasized but minor ones were removed. The minor fluorescenceswere retained, however, in the latter mode. Because the Additive modewas suitable for visualization of F-actin structures, especiallycontinuous F-actin cables, we used this mode to show whole F-actincables in 3D images in this paper. To visualize F-actin structures morein detail, we showed deconvoluted serial sections.

ResultsBecause it was difficult to visualize the detailed structure of theactin cytoskeleton during F-actin ring formation in the S. pombe

Journal of Cell Science 115 (5)

889F-actin ring formation in fission yeast

cell by using conventional fluorescence microscopy, we analyzedthe localization of F-actin especially in prophase-to-metaphasecells using optical sectioning and 3D reconstruction microscopy.About 70 prophase to metaphase cells were analyzed. About 30cells were analyzed in interphase, anaphase or cytokinesis.

Accumulation of F-actin cables at the medial region inwild-type cells during early mitosisAn interphase cell that passed through ‘new end take off’

(Marks and Hyams, 1985) showed a single spot of SPB (Fig.1). In this stage, several F-actin cables ran in the longitudinaldirection of the cell from one end of the cell to the other, andseveral short F-actin cables were also observed. F-actin patcheswere concentrated in both growing ends and at least one endof each F-actin cable was linked to an F-actin patch.

In the cells showing two SPB spots in one optical section(Fig. 2), a few major F-actin cables were seen to adjoin theSPBs at the mid region of the cells and branch in this region.The stage of these cells was considered to be pre-prophase toearly prophase, for the following reasons. First, the length ofthese cells was in the range of 13-14 µm, which was the sizeof mitotic cells we used. Second, two adjacently positionedSPB spots were recognized in these cells in a 0.2 µm section.Third, maturation of duplicated SPBs, which are not separatedyet, occurs in S. pombecells during this stage, and the size ofone matured SPB is about 0.2 µm (Ding et al., 1997). Althoughmost of the F-actin patches were localized near the ends, aconsiderable number were seen in the cytoplasm away fromthe ends. Some of them were attached to the F-actin cables.

In a metaphase cell (Fig. 3), the density of the F-actin cablesincreased at the medial region of the cell. Most of these F-actincables seemed to emanate radially from a focus forming anaster-like structure, whereas there were some cables that werenot involved in this structure. A single F-actin cable extendsfrom the aster and encircles the cell at the equator. We call thisF-actin cable a ‘leading F-actin cable’. In Fig. 3A, this cableonce branched into two cables and then they merged again.Two leading cables extending from the aster and moving in twoopposite directions were occasionally seen (Fig. 3B). At thereverse side of the medial cortex, a clear longitudinal cable wasoften observed (Fig. 3). Serial sectioning showed that theaccumulation of the F-actin cables and the formation of theaster-like structure occur near the cell surface but not in theinner cytoplasm.

Association of F-actin cables with the F-actin ringIn a cell at anaphase A (Fig. 4), a thicker F-actin ring wasobserved. A fairly distorted portion is seen at one side of theF-actin ring in contrast to the portion at the other side of thering (Fig. 4d-f). This distorted portion may have been the aster-like structure at the previous stage. Heavy accumulation of F-actin cables were seen at the medial region, and they werelinked to the F-actin ring. In a cell in anaphase B (Fig. 5a), asharp and ‘straight’ F-actin ring was seen, to which some F-actin cables were always linked. In a late anaphase cell (Fig.5b) in which SPBs were completely separated towards bothends, an F-actin ring accompanied by fewer F-actin cables wasseen.

The F-actin patches showed relatively random distributionduring prophase to mid anaphase, being sparse at the midregion as compared with interphase. However, the F-actinpatches were scarcely seen, especially at the medial region ofthe cell during late anaphase.

As the cells initiated cytokinesis (Fig. 5c), the F-actin ringstarted to contract, and the F-actin cables became clearlyvisible again extending from the division site. The F-actinpatches also reappeared at the medial region of the cells (Fig.5c). In either a late cytokinesis cell (Fig. 5d) or separatingdaughter cells (Fig. 5e), many F-actin patches emerged around

Fig. 1.F-actin structures in a wild-type cell during interphase. Green,red and blue represent F-actin, SPBs and DNA, respectively. Arepresentative 3D image is shown in a. (b) Illustration of the F-actincables in the cell. Thick lines represent the F-actin cables at one sideof the cell, whereas thin lines represent those at the other side of thecell. Deconvoluted serial sections are shown in c-n. A magnifiedimage of SPB is shown in e. Three major F-actin cables are seen inthis cell (white arrows). Some of the F-actin patches are linked to theF-actin cables (pink arrows). Bars, 2 µm (a,n); 1 µm (e).

890

the newly formed septa, and the F-actin cables elongated fromthe septa. Serial sections showed that the F-actin cables werelinked to either the shrinking ring or the F-actin patches at themedial region (data not shown).

Broad F-actin ring in nda3 conditional mutantThe broad F-actin ring similar to the one seen in the wild-typecells during prophase (Arai et al., 1998) has also been observedin the nda3cold-sensitive mutant at a restrictive temperature(Chang et al., 1996). This mutant is thought to be arrested atprophase, and duplicated SPBs cannot separate from each otheron the nucleus (Hiraoka et al., 1984; Kanbe et al., 1990). At 2-4 minutes after release from the arrested condition, the SPBsinitiate separation on the nucleus, and then chromosomesegregation (~10 minutes after the release) and cytokinesis(~14 minutes after the release) proceed in order (Hiraoka et al.,1984; Kanbe et al., 1990). However, it has not yet beenexamined how this broad F-actin ring is related to that seen inthe wild-type cells and reorganization of the actin cytoskeletonafter the release.

We observed localization of F-actin in both the arrested nda3mutant cells and the released cells using Bodipy-phallacidinstaining. In the arrested cells showing condensedchromosomes, a broad F-actin ring and some elongated cableswere seen as observed by conventional fluorescencemicroscopy (Fig. 6a,b). As analyzed by the optical sectioningand 3D reconstitution microscopy, an F-actin ringaccompanied by several F-actin cables was visualized in thearrested cells (Fig. 6i). This appearance is very similar to thatof early anaphase cell. These rings seemed to be loosely packedas compared to those seen in released cells (see below).

The F-actin rings were narrowed at 4 minutes after shift topermissive temperature (Fig. 6c,d). As analyzed by the 3Dreconstruction microscopy (Fig. 6j), packed F-actin rings were

clearly visible. These rings did not seem to be accompanied bythe F-actin cables. F-actin patches appeared all over the cellcortex. After the completion of the chromosome segregation(Fig. 6e-h), the F-actin ring started to contract at ~10 minutesafter the release.

Actin cytoskeleton in several temperature-sensitivecytokinesis mutantsBecause we considered that cytokinesis mutants might havedefects in organization of the F-actin cables, we examined actincytoskeleton in some temperature-sensitive mutant cells. Weexamined the second mitosis after temperature shift to observethe exact phenotype of the mutant cells. Detailed mitotic stageswere determined both by the extent of condensation ofchromosomes and the position of the SPBs on the nuclei.During anaphase, two sister nuclei move towards both ends ofthe cell. However, they are pulled back towards the cell centerafter anaphase, and a septum is formed between the closelylocated nuclei during cytokinesis in the wild-type cell (Hagan,1998). The SPBs always localize on the moving side of thenucleus. In most of cytokinesis mutant cells examined in thisstudy, the sister nuclei separated after the first mitosis at arestrictive temperature were located close to each other nearthe center of the elongated cells. They stayed in these positionswithout a septum during the next interphase to metaphase. Asa result, two spindles tended to elongate parallel to each otherduring the following anaphase in the second mitotic cycle.

F-actin structures in cdc12 and cdc15 mutant cells wereanalyzed by the optical sectioning microscopy. Duringinterphase, F-actin cables seemed to be normally arranged in boththe cdc12and the cdc15mutant cells (data not shown). In acdc12-112mutant cell at metaphase (Fig. 7A), two aster-likestructures were seen in the medial cell cortex. However, noleading F-actin cable was recognized. During anaphase (Fig. 7B),


Fig. 2.F-actin structures in three wild-type cells at early prophase. Green, redand blue (a-f) represent F-actin, SPBsand DNA, respectively. Arepresentative 3D image of one of thesecells is shown in a. Deconvoluted serialsections around the SPBs of these cellsare shown in b-f, g-j and k-o,respectively. The SPBs just initiate toseparate (pink arrows; see magnifiedimages in insets of d, i and k,respectively), and some F-actin cables(white arrows) are branched near theSPBs at the medial region of the cell.Arrowheads indicate F-actin patchesattached to F-actin cables. Bars, 2 µm(o); 1 µm (k).


the aster-like structures were still recognizable and F-actin cableswere accumulated at the medial region. However, again, noleading F-actin cable was seen, in contrast to anaphase wild-typecells in which the aster-like structure could notclearly be recognized; instead, a distorted (Fig. 4)or a packed (Fig. 5a) F-actin ring was observed.After anaphase (Fig. 7C), some F-actin cables werelongitudinally elongated like interphase wild-typecells, and the accumulation of the F-actin cables atthe medial region was not seen any longer.

In a cdc15-140mutant cell at metaphase (Fig.8A), two aster-like structures, each possessing aleading F-actin cable, were seen. During anaphase(Fig. 8B), two distorted F-actin rings, whichseemed to be partially connected with each other,were formed. After anaphase (Fig. 8C), severallongitudinal F-actin cables were clearly seen.These cables were elongated from a point at themedial cell cortex. A remnant of an F-actin ring,which did not seem to have shrunk after anaphase,was also seen at the cell equator, and it was linkedto the medial point from which the F-actin cableswere elongated. In the first mitotic cycle after shiftto the restrictive temperature, only one aster-likestructure was observed both in the cdc12 andcdc15mutant cells (data not shown). Therefore, itis confirmed that one nucleus formed one aster-like structure, and that the two aster-like structuresseen during the second mitosis were not fragmentsderived from one single aster-like structurethrough its degradation.

We further analyzed F-actin structures in spg1,cdc7and sid2mutant cells during and after the F-actin ring formation. These mutant cells showedsimilar phenotypes. During metaphase, we sawtwo aster-like structures near the nuclei. Aleading cable seemed to extend from each of theaster-like structures (data not shown). Then, asingle F-actin ring was formed in about half ofthe cells at late anaphase (Fig. 9a,b), whereasdouble rings were also observed in the remainder(Fig. 9c,d). We could never find cells that had noF-actin ring in this stage. Most of the double ringsseemed to be connected to each other at onepoint. Moreover, one of the double rings was a

complete ring, whereas the other one was incomplete (Fig. 9d).In the first division cycle after shift to the restrictivetemperature, only one single F-actin ring was observed in these

Fig. 3.F-actin structures in wild-type cells atmetaphase. Green, red and blue represent F-actin,SPBs and DNA, respectively. (A) A cell in which asingle leading cable extends from the aster-likestructure. (B) A cell in which two leading cablesextend from the aster-like structure. Rotating 3Dimages are shown in a and b (0° and 12°, respectively)of each panel. Direction of rotation is marked near a.(c) The F-actin cables in each cell. Thick linesrepresent the F-actin cables, including the aster at oneside of the cell, whereas thin lines represent those atthe reverse side of the cell. Deconvoluted serialsections are shown in d-o (A) or d-s (B). Pink arrowsindicate an aster-like structure, and white arrowsindicate leading F-actin cables. White arrowheadsindicate longitudinal F-actin cable. Bars, 2 µm.

892

mutant cells during anaphase (data not shown). After anaphase,24 out of 34 spg1mutant cells, 22 out of 31 cdc7mutant cellsand 6 out of 32 sid2 mutant cells still possessed the F-actinring – which did not seem to have contracted (Fig. 9g,h, rightcell) – whereas the rest of the cells did not (Fig. 9e,f). As theSPBs were pulled back further towards the cell center, thepersisted F-actin ring disappeared (Fig. 9g,h, left cell). Beingdistinct from the cdc15cells at the same stage, the longitudinalF-actin cables were not seen in these mutant cells. Curiously,F-actin patches reappeared at the medial region (Fig. 9f,h),although septum formation was never observed in these cells.

DiscussionMutual relationship among F-actin structures inS. pombe cellsSchizosaccharomyces pombecells have three major actin

cytoskeletons, namely the F-actin patches, the F-actin cablesand the F-actin ring, which form during mitosis. Structuralrelationships among these cytoskeletons have not well beenknown. We have previously reported that the F-actin cablesseem to be linked to some of the F-actin patches and to the F-actin ring (Arai et al., 1998). In this study, optical sectioningof the cell followed by deconvolution and 3D reconstructionallowed us a more detailed analysis of the actin cytoskeleton.The linking of the F-actin cables to some of the F-actin patches,at least at one end, was confirmed during interphase and earlyprophase. In addition, attachment of the patches on thecables was observed in metaphase cells. This is similar toSaccharomyces cerevisiae, in which the patches and the cablesare often linked, as observed by 3D reconstruction (Karpova etal., 1998). Recently, it has also been reported that the actinpatches move in a directed manner along the F-actin cables inS. pombe(Pelham and Chang, 2001). However, the F-actin ringis formed from the F-actin cables during mitosis in wild-typeS. pombecells (see below). Therefore, these three F-actinstructures in S. pombeare closely related to each other.Because these cytoskeletons are dynamic structures such thatthey may disassemble and reassemble in the course of the cellcycle, and as only a small amount of G-actin seems to existin the cell (R.A. and I.M., unpublished), it could be that F-actin interchanges between these cytoskeletons withoutdepolymerization, or that very rapid depolymerization andrepolymerization take place in the cell.

Protein constituents other than actin in these structures arereported to be different. Tropomyosin is present in both the F-actin cables and the F-actin ring, but is absent in the majorityof the F-actin patches (Arai et al., 1998). By contrast, Arp3 ispresent in the F-actin patches but is absent from both the F-actin cable and the F-actin ring (McCollum et al., 1996; Araiet al., 1998). Type II myosin light chains (Cdc4 and Rlc1)(McCollum et al., 1995; Le Goff et al., 2000; Naqvi et al.,


Fig. 4.F-actin structures in a wild-type cell at early anaphase. Green,red and blue represent F-actin, SPBs and DNA, respectively.Rotating 3D images are shown in a and b (0° and 24°, respectively).Direction of rotation is marked on the left of a. Deconvoluted serialsections are shown in c-m. F-actin ring is formed but partiallydistorted (pink arrow). Some F-actin cables are linked to the ring(white arrows). Bar, 2 µm.

Fig. 5.F-actin structures inwild-type cells duringanaphase and cytokinesis.Green, red and blue representF-actin, SPBs and DNA,respectively. 3D images ofmid and late anaphase cells(a and b, respectively), earlyand late cytokinesis cells(c and d, respectively) andseparating cells (e) are shown.White arrows, F-actin cables.Pink arrows, F-actin patches.Bar, 2 µm.


2000), type II myosin heavy chains (Kitayama et al., 1997;May et al., 1997; Motegi et al., 2000) and an IQGAP-likeprotein Rng2 (Eng et al., 1998) are present in the F-actin ringbut not in the F-actin patches and cables. α-actinin (Ain1) (Wuet al., 2001) and fimbrin (Fim1) (Wu et al., 2001; Nakano etal., 2001) are present in the F-actin ring and in the patches butnot in the cables. These specific components may characterizeand define organization and functions of the three F-actincytoskeletons in S. pombe.

Formation of aster-like structure of F-actin cablesMajor F-actin cables were branched near the SPBs at earlyprophase, and then an aster-like structure composed of the F-actin cables was seen in the middle region during metaphase.The branching of the F-actin cables seemed to be a first stepin the formation of the aster-like structure. Because theseevents occur at a position very close to the SPBs, which arenot yet separated from each other, it is tempting to speculatethat the nonseparated SPBs control the formation of the aster-like structure during early prophase. The formation of thisstructure is a novel finding in S. pombeduring early mitosis. Ithas also been reported that radial F-actin arrays, which aresimilar to the aster-like structure described here, are observedat a future site of new hypha formation and at a futureelongation tip of each spore in the oomyceteSaprolegnia ferax (Bachewich and Heath, 1998).The relationship between this structure and theSPBs/nucleus, however, has not been investigated.

Accumulation of F-actin cables at the medialregion of the cell during metaphaseA broad and faint F-actin ring has been observedin early mitotic wild-type cells using conventionalfluorescence microscopy (Chang et al., 1996; Araiet al., 1998; Bähler et al., 1998). In this study, thedensity of the F-actin cables increased in thisregion in metaphase wild-type cells. Thus, it isreasonable to consider that the broad F-actin ringvisualized by the conventional microscopy actuallyrepresents the accumulated F-actin cables. Wheredo these F-actin cables come from? The firstpossibility is that the F-actin cables are originatedfrom the center of the aster-like structure.However, our observation suggests that not all thecables emanated from the aster-like structure. Asecond possibility is that actin polymerization andsubsequent bundling might have occurred in thebroad medial region. However, we could not detectnewly polymerized actin filaments, not yetbundled, in the metaphase wild-type, as analyzedby optical sectioning and 3D reconstructionmicroscopy. Thus, this theory is weakened,although the polymerized actin filaments could beremoved from the image as background by thedeconvolution process and only the stronglyfluorescent F-actin cables could have remained.The third possibility is that the F-actin cables movefrom both sides of the cell to the medial region.This is probable because the density of the F-actin

cables move from the ends to the medial region as mitosisprogresses. It is necessary to live-record the movement of thecables to prove this hypothesis and to know how the cablesmove in the cell by using expression of GFP-actin fusionprotein, for example.

Process of F-actin ring formation A schematic model for the F-actin ring formation is presentedin Fig. 10, which is based on the present observations. In thismodel, it is supposed that a positional signal(s) (see below) isgenerated from the SPBs or nucleus, or both, and reaches thecortex of the medial region of the cell at pre-prophase. Thesignal(s) induces the formation of the aster-like structureduring early prophase. The accumulation of the F-actin cablesis initiated during prophase.

During the formation of the aster-like structure, a leading F-actin cable seems to extend from this structure, and encirclethe cell at the equator to form the primary F-actin ring duringmetaphase. F-actin cables accumulated in the medial regionseem to be connected to the leading cable during earlyanaphase and then all the cable structures are packed to formthe complete F-actin ring during late anaphase.

It has been shown that the type II myosin heavy chain Myo2accumulates around the mid region independently of its head

Fig. 6.F-actin structures in nda3mutant cells. nda3mutant cells were incubated at20°C for 8 hours to arrest. For release experiment, once arrested, cells were re-incubated at 36°C and collected every 2 minutes. These cells were immediatelyfixed at each time, and then stained simultaneously with DAPI and Bodipy-phallacidin. (a-h) Photographs obtained by conventional fluorescence microscopy.Left, DAPI stainings; right, Bodipy-phallacidin stainings. (a,b) Arrested cells.(c-h) Released cells: 4 minutes (c,d), 10 minutes (e,f) and 14 minutes (g,h) aftertemperature shift-up. (i,j) 3D images. Green and blue represent F-actin and DNA,respectively. (i) Arrested cells. (j) Cells at 4 minutes after the release. Arrowsindicate accumulated F-actin cables. Bar, 2 µm.

894

domain (Naqvi et al., 1999) and F-actin (Motegi et al., 2000)during early mitosis. Myo2 forms spot structures that do notcolocalize with the aster-like structure. Subsequently, theMyo2 spots change into a spot-fiber network, and then thenetwork is packed to form the ring along with the packing ofthe F-actin cables to form the F-actin ring. Colocalization ofMyo2 with the F-actin cable is seen during this stage (Motegiet al., 2000), and this process may require ATPase activity ofMyo2 (Naqvi et al., 1999) and the presence of F-actin (Motegiet al., 2000).

Conversely, F-actin cables can accumulate in the mid region

of the cell in the cdc4 mutant at the restrictive temperature(Motegi et al., 2000), although both Cdc4 (McCollum et al.,1995) and myosin II heavy chains (Myo2 and Myp2/Myo3)(Motegi et al., 1997; Motegi et al., 2000) are required for theformation of the F-actin ring. These observations suggest thatthe changes in F-actin organization at the mid region of thecell, which include the accumulation of F-actin cables andformation of the aster-like structure, and changes in myosin IIspots at the mid region involve processes that are independentof each other. The next step to complete the F-actin ringformation may require the interaction of these structures.


Fig. 7. F-actin structures in cdc12-112mutant cells. Green, red andblue represent F-actin, SPBs and DNA, respectively. (A) Metaphasecell. (a) 3D image of chromosomes and SPBs. (b) 3D image of F-actin structures. (c) F-actin cables in this cell. Thick lines representthe cables at one side of the cell, whereas thin lines represent thoseat the other side of the cell. (d-q) Deconvoluted serial sectionsaround a medial region of this cell. (B) Anaphase cell. (a) 3D imageof chromosomes and SPBs. (b) 3D image of F-actin structures. (c) F-actin cables in this cell. (d-q) Deconvoluted serial sections around amedial region of this cell. (C) A cell after anaphase. (a) 3D image ofchromosomes and SPBs. (b) 3D image of F-actin structures. Pinkand white arrowheads represent pairs of sister SPBs, respectively.Pink arrows, aster-like structures. White arrows, F-actin cables.Bars, 2 µm


Positional signals and the aster-like structureRecently, it has been proposed that the positional signal(s) forthe F-actin ring formation is generated from the nucleus in themiddle of the cell (Chang and Nurse, 1996) and thattransduction of such signal(s) may be mediated by SPBs duringprophase (Bähler et al., 1998) in S. pombe. In fact, it has beenreported that Mid1, which has been considered to position thesite of the F-actin ring formation (Chang et al., 1996;Sohrmann et al., 1996), localizes on the nuclear membraneduring interphase and relocalizes to the broad medial region ofthe cells during prophase (Sohrmann et al., 1996; Bähler et al.,1998). For this relocalization, function of Plo1 is required, andPlo1 localizes to the SPBs during prophase to anaphase (Bähleret al., 1998). Therefore, these factors are probably involved inthe signaling pathway for the F-actin ring formation. However,it has not been clarified how the signal(s) is transferred fromthe nucleus or the SPBs, or both, to the cortex of the cell.Although the SPBs are attached on the nuclear membraneduring interphase, they are buried in the nuclear membrane

during mitosis in S. pombe: a part of the nuclear membrane justbeneath the SPBs is collapsed at pre-prophase and then theSPBs are settled in the nuclear membrane (Ding et al., 1997).The partial collapse of the nuclear membrane at pre-prophasemay be a trigger to generate signals for initiation of the F-actinring formation. The unknown signal(s) from the SPBs maymark a spot on the medial cortex at pre-prophase or earlyprophase, and the formation of the aster-like structure may beinitiated from this spot. The accumulated F-actin cables thatare not involved in the aster-like structure may be regulated byother signals that localize to the broad medial region duringprophase to anaphase, such as Mid1.

Possible role of Cdc12 in the ring formationIt has been reported that overexpressed Cdc12-GFP forms a spotand this spot moves in the cell during interphase, probably alongmicrotubular tracks and F-actin tracks. The spot arrives at themedial region of the cell during prophase and a ring of Cdc12-

Fig. 8.F-actin structures in cdc15-140mutant cells.Green, red and blue represent F-actin, SPBs andDNA, respectively. (A) Metaphase cell. (a) 3Dimage of chromosomes and SPBs. (b,c) Rotating 3Dimages of F-actin structures (0° and 24°,respectively). Direction of rotation is marked on theright of c. (d-r) Deconvoluted serial sections arounda medial region of this cell. Pink arrows, aster-likestructures. White arrows, leading F-actin cables.(B) Anaphase cell. (a) 3D image of chromosomesand SPBs. Pink and white arrowheads representpairs of sister SPBs, respectively. (b) 3D image of F-actin structures. (c-p) Deconvoluted serial sectionsaround a medial region of this cell. Double F-actinring-like structures (arrows) are formed. (C) A cellafter anaphase. (a) 3D image of chromosomes andSPBs. Pink and white arrowheads represent pairs ofsister SPBs, respectively. (b) 3D image of F-actinstructures. (c-n) Deconvoluted serial sections arounda medial region of this cell. A part of the F-actin ringstill remains (white arrows). Some F-actin cables areelongated from a point (pink arrows). Bars, 2 µm.

896

GFP is formed from it by the spot extending a Cdc12-GFP strandalong the cell equator during prophase to metaphase (Chang etal., 1997; Chang, 1999). It has also been reported that a domainof Cdc12 possesses an ability to interact with components ofSPBs (Petersen et al., 1998). It is possible that the Cdc12-GFPspot mediates the positional signal(s) from the SPBs to themedial cortex, or that the Cdc12-GFP spot recognizes thesignal(s) that had been transferred on the medial cortex, andbecomes the center of the aster-like structure. By contrast, thisstructure was observed in cdc12mutant cells during prophase tometaphase in this study, suggesting that Cdc12 may not benecessary in the formation of the aster. However, it could be thatthe mutated Cdc12 still possesses function to form the aster-likestructure. However, no leading F-actin cable was observed in thecdc12mutant cells during mitosis, although the accumulation ofthe F-actin cables around the mid region occurred. This suggeststhat Cdc12 is required to form the leading F-actin cable andthereby to form the primary F-actin ring. The formation of thestrand that encircles the cell equator from the Cdc12-GFP spotseems to be similar to the leading cable formation from the aster-like structure. Thus, the leading F-actin cable may be formed bybeing influenced by the Cdc12 strand. It has been reported that

Cdc12 interacts with Cdc3 profilin (Chang et al., 1997), andprofilin is concentrated around the mid region during mitosis(Balasubramanian et al., 1994). Actin polymerization may bestimulated by Cdc12 via profilin in the formation of both theaster and the leading F-actin cable. It will be necessary in thefuture to show actual elongation of the leading F-actin cablefrom the aster-like structure and its relationship to the elongationof the Cdc12 strand in living cells by means of expression ofGFP-fusion proteins. In addition, an activity of the accumulationof the F-actin cables in the mid region is retained duringprophase to anaphase, and this accumulation is not very markedafter anaphase in the cdc12mutant cells. This transition of theF-actin cable arrangement does not seem to require the Cdc12function.

Rng2, an IQGAP-like protein, has been considered to play


Fig. 9. F-actin structures in spg1-B8, cdc7-24 and sid2-250mutantcells. Green, red and blue represent F-actin, SPBs and DNA,respectively. (a,b) A cdc7mutant cell at late anaphase. (c,d) A sid2mutant cell at late anaphase. An incomplete F-actin ring (pink arrow)is connected to a complete F-actin ring (white arrow). (e,f) A cdc7mutant cell after anaphase. (g,h) spg1mutant cells. White arrow,remaining F-actin ring. Pink arrow, a remnant of disappearing F-actinring. (a,c,e,g) 3D images of chromosomes and SPBs. Pink and whitearrowheads represent pairs of sister SPBs. (b,d,f,h) 3D images ofF-actin structures. Bar, 2 µm.

Fig. 10.A model for F-actin ring formation in the wild-type S.pombecell. F-actin cables are elongated between both ends of thecell during interphase. As positional signal(s) is generated fromunseparated SPBs and/or nucleus at pre-prophase, formation of anaster-like structure is initiated from a spot during prophase, mediatedby Cdc12 or other unknown proteins. This process includesbranching of the F-actin cable, which is closely associated with theSPBs. During metaphase, the aster formation and accumulation ofthe F-actin cables progress at the medial region. A leading F-actincable extends from the aster and encircle the cytoplasm at theequator. Thus a primary F-actin ring is formed. Cdc12 may beinvolved in this primary ring formation. A thick F-actin ring isformed during anaphase by packing and fusion of the accumulatedF-actin cables with the leading cable. Cdc15 may play a role in thisprocess. During cytokinesis, the F-actin ring initiates contractionbeing regulated by the Spg1-Cdc7-Sid2 pathway, and then F-actincables reappear.


an important role in the signaling of the F-actin ring formation,as it localizes to the SPBs during interphase, and relocalizes tothe medial ring during mitosis (Eng et al., 1998). rng2null cellsform an F-actin spot containing Cdc3 and calmodulin at themid region, although they cannot form the F-actin ring. Thisspot formation is thought to be an intermediate step in thepathway of the F-actin ring formation (Eng et al., 1998).Moreover, rng2mutation shows synthetic lethality with cdc12.Therefore, it would be important to investigate how Cdc12 andRng2 are involved in the formation of the aster-like structure,and whether this structure and the F-actin spot are related.

Arrangement of F-actin cables in cdc15 mutantIn cdc15-140mutant cells, the aster-like structure, the leadingcable and the distorted F-actin ring were almost normallyformed in order during mitosis, although cytokinesis neveroccurred subsequently. It has also been reported that the F-actinring is formed in both cdc15-140 and cdc15-A5mutant cellsduring mitosis, although both recruitment of the F-actinpatches to the medial region and septation after the F-actin ringformation do not occur (Balasubramanian et al., 1998). On thecontrary, it has been reported that the F-actin ring is formedonly in a small population of mitotic cells of cdc15-140andcdc15-287(Fankhauser et al., 1995; Chang et al., 1997). Threedifferent cdc15 mutant strains were used in these reports,including the present study. It may be that both the differentmutations and minute differences in experimental conditionshave resulted in the appearance of different phenotypes. Inaddition, the distorted F-actin ring may not have been regardedas the F-actin ring by some microscopic systems used for theabove observation. From our detailed observation, the cdc15-140 mutant could form the distorted F-actin ring duringanaphase, but the subsequent packing of the ring and itscontraction did not occur. It is suggested that Cdc15 is requiredfor packing of the accumulated F-actin cables to form thecomplete F-actin ring, which is able to contract.

In the cdc15-140cells, a part of the F-actin ring still remainedafter anaphase, suggesting that these cells also have defects inthe disassembly of the F-actin ring after anaphase. It has beenreported that Cdc15 has a structural similarity to Imp2, whichis required for F-actin ring disassembly and cell separation, andthat a double mutant of cdc15-140and imp2 shows syntheticlethality (Demeter and Sazer, 1998). These data suggest thatCdc15 is required for the rearrangements of actin cytoskeletonin the process of cytokinesis, including the packing of the F-actin cables during the formation of the F-actin ring anddisassembly of the F-actin ring after the contraction.

Arrangement and function of F-actin ring in nda3 mutant In the nda3 mutant, the formation of the F-actin ring and theaccumulation of fine F-actin cables associated with the ring wereobserved at the restrictive temperature. The nda3mutant cellsare thought to be arrested at prophase, as duplicated SPBs cannotinitiate separation on the nucleus because microtubules cannotelongate between these SPBs in this mutant (Hiraoka et al.,1984;Kanbe et al., 1990). However, in the arrested nda3mutant cells,we could not detect the aster-like structure, which is normallyformed during prophase in the wild-type cell, and the F-actinring was similar to that of wild-type cells in early anaphase.

Therefore, it could be that the mechanism by which the F-actinring formation is controlled proceeded beyond prophase to earlyanaphase; the activity to accumulate the F-actin cables isprobably stimulated and retained once cells enter prophase. Assoon as the arrested cells were shifted to the permissivetemperature, the tightly packed F-actin ring was formed, whichwas similar to the complete F-actin ring observed in the wild-type cells at late anaphase, which then contracted andcytokinesis progressed. Interestingly, the F-actin ring in thearrested nda3mutant cells never contracted unless temperaturewas increased. These results may suggest that the completion ofthe F-actin ring formation is an essential step for the cell to entercytokinesis, and that function of cytoplasmic microtubules isrequired for this process. However, nda3mutant cells with mad2defect can undergo septation (He et al., 1997). Because Mad2has been known as a component of spindle check point (Li andMurray, 1991; He et al., 1997), the completion of the F-actinring formation and the initiation of the ring contraction may beregulated by such a mitotic check point.

F-actin ring formation during cytokinesis in septuminitiation mutants At late anaphase, the single or double F-actin rings wereformed in the spg1, cdc7or sid2mutant cells. In the latter case,one complete ring and one incomplete ring were connected toeach other at one point. It is possible that the single ring wasformed as a result of fusion of the double rings. The leadingF-actin cable, the aster-like structure and the F-actin ring areformed in these mutant cells, indicating that the componentsof the septum initiation network (McCollum and Gould, 2001)are not involved in the formation of the F-actin ring.

Spg1 localizes only to the SPB (s) at all stages of the cellcycle and recruits Cdc7 to the SPB(s) during mitosis (Schmidtet al., 1997; Sohrmann et al., 1998). However, Sid2 localizesnot only to the SPBs but also to the F-actin ring and both sidesof the forming septum during late anaphase to cytokinesis(Sparks et al., 1999). It has also been reported that Sid2 isrequired for initiation of the F-actin ring contraction or septumformation, or both, as a downstream effector of the Spg1-Cdc7pathway (Balasubramanian et al., 1998; Sparks et al., 1999). Infact, septum formation never occurred in these septum initiationmutant cells. We observed that the majority of the spg1or cdc7mutant cells possessed the F-actin ring after anaphase, whereasthe ring disappeared earlier in most of the sid2 mutant cells.Thus, Sid2 may contribute to retaining the F-actin ring structureduring contraction through localizing to this structure.

We thank Dr V. Simanis of the ISREC, Switzerland and Dr D.McCollum of the University of Massachusetts, USA for supplying uswith mutant strains, and Dr I. Hagan of the University of Manchester,UK for anti-Sad1 antibodies. This work was supported by a researchgrant from the Ministry of Education, Science and Culture of Japan(no.10213202).

ReferencesAlfa, C. and Hyams, J. (1990). Distribution of tubulin and actin through the

cell division cycle of the fission yeast Schizosaccharomyces japonicusvar.versatilis: a comparison with Schizosaccharomyces pombe. J. Cell Sci. 96,71-77.

Arai, R., Nakano, K. and Mabuchi, I. (1998). Subcellular localization andpossible function of actin, tropomyosin and actin-related protein 3 (Arp3)

898

in the fission yeast Schizosaccharomyces pombe. Eur. J. Cell Biol. 76, 288-295.

Bachewich, C. and Heath, I. B. (1998). Radial F-actin arrays precede newhypha formation in Saprolegnia: implications for establishing polar growthand regulating tip morphogenesis. J. Cell Sci. 111, 2005-2016.

Bähler, J., Steever, A. B., Wheatley, S., Wang, Y. L., Pringle, J. R., Gould,K. L. and McCollum, D. (1998). Role of polo kinase and Mid1p indetermining the site of cell division in fission yeast. J. Cell Biol. 143, 1603-1616.

Balasubramanian, M. K., Helfman, D. M. and Hemmingsen, S. M. (1992).A new tropomyosin essential for cytokinesis in the fission yeast S. pombe.Nature360, 84-87.

Balasubramanian, M. K., Hirani, B. R., Burke, J. D. and Gould, K. L.(1994). The Schizosaccharomyces pombecdc3+ gene encodes a profilinessential for cytokinesis. J. Cell Biol. 125, 1289-1301.

Balasubramanian, M. K., McCollum, D., Chang, L., Wong, K. C. Y.,Naqvi, N. I., He, X., Sazer, S. and Gould, K. L. (1998). Isolation andcharacterization of new fission yeast cytokinesis mutants. Genetics149,1265-1275.

Bezanilla, M., Forsburg, S. L. and Pollard, T. D. (1997). Identification of asecond myosin-II in Schizosaccharomyces pombe:Myp2p is conditionallyrequired for cytokinesis. Mol. Biol. Cell8, 2693-2705.

Chang, F. (1999). Movement of a cytokinesis factor cdc12p to the site of celldivision. Curr. Biol. 9, 849-852.

Chang, F. and Nurse, P. (1996). How fission yeast fission in the middle. Cell84, 191-194.

Chang, F., Woollard, A. and Nurse, P. (1996). Isolation and characterizationof fission yeast mutants defective in the assembly and placement of thecontractile actin ring. J. Cell Sci. 109, 131-142.

Chang, F., Drubin, D. and Nurse, P. (1997). cdc12p, a protein for cytokinesisin fission yeast, is a component of the cell division ring and interacts withprofilin. J. Cell Biol. 137, 169-182.

Demeter, J. and Sazer, S. (1998). imp2, a new component of the actin ring inthe fission yeast Schizosaccharomyces pombe. J. Cell Biol. 143, 415-427.

Ding, R., West, R. R., Morphew, M., Oakley, B. R. and McIntosh, J. R. (1997).The spindle pole body of Schizosaccharomyces pombeenters and leaves thenuclear envelope as the cell cycle proceeds. Mol. Biol. Cell8, 1461-1479.

Edamatsu, M. and Toyoshima, Y. Y. (1996). Isolation and characterizationof pos mutants defective in correct positioning of septum inSchizosaccharomyces pombe. Zool. Sci. 13, 235-239.

Eng, K., Naqvi, N. I., Wong, K. C. Y. and Balasubramanian, M. K. (1998).Rng2, a protein required for cytokinesis in fission yeast, is a component ofthe actomyosin ring and the spindle pole body. Curr. Biol. 8, 611-621.

Fankhauser, C. and Simanis, V. (1994). The cdc7protein kinase is a dosagedependent regulator of septum formation in fission yeast. EMBO J. 13,3011-3019.

Fankhauser, C., Reymond, A., Cerutti, L., Utzig, S., Hofmann, K. andSimanis, V. (1995). The S. pombe cdc15gene is a key element in thereorganization of F-actin at mitosis. Cell 82, 435-444.

Hagan, I. (1998). The fission yeast microtubule cytoskeleton. J. Cell Sci. 111,1603-1612.

He, X., Patterson, T. E. and Sazer, S. (1997). The Schizosaccharomycespombespindle checkpoint protein mad2p blocks anaphase and geneticallyinteracts with the anaphase-promoting complex. Proc. Natl. Acad. Sci. USA94, 7965-7970.

Hiraoka, Y., Toda, T. and Yanagida, M. (1984). The NDA3 gene of fissionyeast encodes beta-tubulin: a cold-sensitive nda3mutation reversibly blocksspindle formation and chromosome movement in mitosis. Cell 39, 349-358.

Ishiguro, J. (1998). Genetic control of fission yeast cell wall synthesis: thegenes involved in wall biogenesis and their interactions inSchizosaccharomyces pombe. Genes Genet. Syst. 73, 181-191.

Ishiguro, J. and Kobayashi, W. (1996). An actin point-mutation neighboringthe ‘hydrophobic plug’ causes defects in the maintenance of cell polarityand septum organization in the fission yeast Schizosaccharomyces pombe.FEBS Lett. 392, 237-241.

Kanbe, T., Hiraoka, Y., Tanaka, K. and Yanagida, M. (1990). The transitionof cells of the fission yeast ß-tubulin mutant nda3-311as seen by freeze-substitution electron microscopy. J. Cell Sci. 96, 275-282.

Karpova, T., McNally, J., Moltz, S. and Cooper, J. (1998). Assembly andfunction of the actin cytoskeleton of yeast: relationships between cables andpatches. J. Cell Biol. 142, 1501-1517.

Kitayama, C., Sugimoto, A. and Yamamoto, M. (1997). Type-II myosinheavy chain encoded by the myo2gene composes the contractile ring duringcytokinesis in Schizosaccharomyces pombe. J. Cell Biol. 137, 1309-1319.

Kobori, H., Yamada, N., Taki, A. and Osumi, M. (1989). Actin is associatedwith the formation of the cell wall in reverting protoplasts of the fission yeastSchizosaccharomyces pombe. J. Cell Sci. 94, 635-646.

Le Goff, X., Utzig, S. and Simanis, V. (1999). Controlling septation in fissionyeast: finding the middle, and timing it right. Curr. Genet. 35, 571-584.

Le Goff, X., Motegi, F., Salimova E., Mabuchi, I. and Simanis, V. (2000).The S. pomberlc1 gene encodes a putative myosin regulatory light chain thatbinds the type II myosins myo3p and myo2p. J. Cell Sci. 113, 4157-4163.

Li, R. and Murray, A. W. (1991). Feedback control of mitosis in buddingyeast. Cell 66, 519-531.

Mabuchi, I. (1986). Biochemical aspects of cytokinesis. Int. Rev. Cytol. 101,175-213.

Mabuchi, I. (1994). Cleavage furrow: timing of emergence of contractile ringactin filaments and establishment of the contractile ring by filament bundlingin sea urchin eggs. J. Cell Sci. 107, 1853-1862.

Marks, J. and Hyams, J. S. (1985). Localization of F-actin through the celldivision cycle of Schizosaccharomyces pombe. Eur. J. Cell Biol. 39, 27-32.

May, K. M., Watts, F. Z., Jones, N. and Hyams, J. S. (1997). Type II myosininvolved in cytokinesis in the fission yeast, Schizosaccharomyces pombe.Cell Motil. Cytoskeleton38, 385-396.

McCollum, D. and Gould, K. L. (2001). Timing is everything: regulation ofmitotic exit and cytokinesis by the MEN and SIN. Trends Cell Biol. 11, 89-95.

McCollum, D., Balasubramanian, M. K., Pelcher, L. E. and Hemmingsen,S. M. (1995). Schizosaccharomyces pombecdc4+ gene encodes a novel EF-hand protein essential for cytokinesis. J. Cell Biol. 130, 651-660.

McCollum, D., Feoktistova, A., Morphew, M., Balasubramanian, M. K.and Gould, K. L. (1996). The Schizosaccharomyces pombeactin-relatedprotein, Arp3, is a component of the cortical actin cytoskeleton and interactswith profilin. EMBO J. 15, 6438-6446.

Motegi, F., Nakano, K., Kitayama, C., Yamamoto, M. and Mabuchi, I.(1997). Identification of Myo3, a second type-II myosin heavy chain in thefission yeast Schizosaccharomyces pombe. FEBS Lett. 420, 161-166.

Motegi, F., Nakano, K. and Mabuchi, I. (2000). Molecular mechanism ofmyosin-II assembly at the division site in Schizosaccharomyces pombe. J.Cell Sci. 113, 1813-1825.

Nakano, K., Satoh, K., Morimatsu, A., Ohnuma, M. and Mabuchi, I.(2001). Interactions among a fimbrin, a capping protein, and an actin-depolymerizing factor in organization of the fission yeast actin cytoskeleton.Mol. Biol. Cell 12, 3515-3526.

Naqvi, N. I., Eng, K., Gould, K. L. and Balasubramanian, M. K. (1999).Evidence for F-actin-dependent and independent mechanisms involved inassembly and stability of the medial actomyosin ring in fission yeast. EMBOJ. 18, 854-862.

Naqvi, N. I., Wong, K. C. Y., Tang, X. and Balasubramanian M. K. (2000).Type II myosin regulatory light chain relieves auto-inhibition of myosin-heavy-chain function. Nat. Cell Biol. 2, 855-858.

Ohkura, H., Hagan, I. M. and Glover, D. M. (1995). The conservedSchizosaccharomyces pombe plo1, required to form a bipolar spindle, theactin ring, and septum, can drive septum formation in G1 and G2 cells.Genes Dev. 9, 1059-1073.

Pelham, R. J., Jr and Chang, F. (2001). Role of actin polymerization andactin cables in actin-patch movement in Schizosaccharomyces pombe. Nat.Cell Biol. 3, 235-244.

Petersen, J., Nielsen, O., Egel, R. and Hagan, I. M. (1998). FH3, a domainfound in formins, targets the fission yeast formin Fus1 to the projection tipduring conjugation. J. Cell Biol. 141, 1217-1228.

Rappaport, R. (1986). Establishment of the mechanism of cytokinesis inanimal cells. Int. Rev. Cytol. 101, 245-281.

Schmidt, S., Sohrmann, M., Hofmann, K., Woollard, A. and Simanis, V.(1997). The Spg1p GTPase is an essential dosage-dependent inducer ofseptum formation in Schizosaccharomyces pombe. Genes Dev. 11, 1519-1534.

Sohrmann, M., Fankhauser, C., Brodbeck, C. and Simanis, V. (1996). Thedmf1/mid1 gene is essential for correct positioning of the division septumin fission yeast. Genes Dev. 10, 2707-2719.

Sohrmann, M., Schmidt, S., Hagan, I. and Simanis, V. (1998). Asymmetricsegregation on spindle poles of the Schizosaccharomyces pombeseptum-inducing protein kinase Cdc7p. Genes Dev. 12, 84-94.

Sparks, C. A., Morphew, M. and McCollum, D. (1999). Sid2p, a spindlepole body kinase that regulates the onset of cytokinesis. J. Cell Biol. 146,777-790.

Wu, J. Q., Bähler, J. and Pringle, J. R. (2001). Roles of a fimbrin and analpha-actinin-like protein in fission yeast cell polarization and cytokinesis.Mol. Biol. Cell12, 1061-1077.


f-actin ring formation in fission yeast · (for a review see le goff et al., 1999). some are...

Documents