behaviorof spindles spindleplaques in cycleand conjugation ... · duringthe latter part ofg1....

JOURNAL OF BACTERIOLOGY, OCt. 1975, p. 511-523Copyright 0 1975 American Society for Microbiology

Vol. 124, No. 1Printed in U.S.A.

Behavior of Spindles and Spindle Plaques in the Cell Cycle andConjugation of Saccharomyces cerevisiae

BRECK BYERS* AND LORETTA GOETSCH

Department of Genetics, University of Washington, Seattle, Washington 98195

Received for publication 21 July 1975

The interdependence of spindle plaque behavior with other aspects of celldivision and conjugation in Saccharomyces cerevisiae has been investigated.Three forms of the spindle plaque appear sequentially before the formation of thecomplete, intranuclear spindle. The single plaque is present initially in themitotic cycle; it becomes transformed into a satellite-bearing single plaqueduring the latter part of G1. Subsequently, plaque duplication yields the doubleplaque characteristic of the early phase of budding, which coincides with theperiod of chromosome replication (S). The eventual separation of these plaques toform a complete spindle, with a single plaque at each pole, is nearly coincidentwith the completion of S. The form of the plaque differs in two independent cases

of G1 arrest: the single plaque is found in a cell in stationary arrest of growth,whereas a cell arrested by mating factors in preparation for conjugation containsa satellite-bearing single plaque. The latter form is retained during zygoteformation, where it serves as the initial site of fusion of each prezygotic nucleuswith the other. This fusion results in the formation of a single zygotic nucleuswith a satellite-bearing single plaque, which is subsequently transformed into a

double plaque as the zygote buds. The double plaque is situated adjacent to thesite of bud emergence in both vegetative cells and zygotes.

The cell division cycle of Saccharomycescerevisiae involves the progression of eventsrequired for nuclear division concomitantlywith those involved in bud development. Cyto-logical analysis has revealed that the two spin-dle plaques, densely staining discoidal struc-tures that are situated in the nuclear envelopeand serve as the poles of the intranuclearspindle, arise by duplication of a single spindleplaque by the time of early bud development(8). Chromosomal deoxyribonucleic acid repli-cation occurs at this same phase of the cell cycle(19).Functional interdependence of these events is

suggested by the behavior of temperature-sensi-tive mutants of the cell division cycle character-ized by Hartwell and his colleagues (6). Strainsmutant in cdc genes become arrested at particu-lar stages of the cycle upon transfer to therestrictive temperature. Electron microscopy ofseveral strains has revealed that these specificstages of arrest are reflected in similarly specificconformations of the spindle and spindleplaques (1). All budded strains were found tohave undergone duplication of the spindleplaques. Among the budded strains, those mu-tant in cdc 4 were unique in two pertinentfeatures. On one hand, the spindle plaques

511

remained in a side-by-side configuration (ordouble plaque) rather than separating to form acomplete spindle. In addition, these cells re-tained a capacity for repeated budding in theabsence of nuclear division.These observations suggest that the spindle

plaques may play a role in the control of othercellular events, particularly of bud emergence.We therefore have undertaken an electron mi-croscopy examination of normal vegetativegrowth and conjugation to determine how theseevents may be interrelated.

MATERIALS AND METHODSStrains and media. Haploid strains A364 A a and

S2072 a of Saccharomyces cerevisiae were obtainedfrom L. H. Hartwell, diploid strains 2681-9A a/a and1422-4C a/a were provided by D. C. Hawthorne, anddiploid strain AP-1 a/a by A. K. Hopper. All strainshave several auxotrophic requirements but weregrown on complete media for all experiments de-scribed here. YEPD medium consists of 1% yeastextract, 2% peptone (Difco), and 2% glucose; 1.5%agar was added for plates. PSP2 medium for growth inacetate contains 0.67% yeast nitrogen base, 0.1% yeastextract, 1% potassium phthalate, 1% potassium ace-tate, and 0.004% each of adenine and uracil; themedium was adjusted to pH 5.4 with potassiumhydroxide.

on June 28, 2017 by guesthttp://jb.asm

.org/D

ownloaded from

http://jb.asm.org/

512 BYERS AND GOETSCH

Assay of budding in logarithmic growth. Becauseit was necessary to determine the growth rate of budsunder conditions identical to those employed forcultures fixed for electron microscopy, samples of theasynchronous cultures grown in PSP2 were fixed in 3%glutaraldehyde in 0.1 M cacodylate, pH 6.8, and weredisaggregated by ultrasonication. Portions were thenphotographed by phase contrast microscopy and thefilms were projected for measurement in a NikonProfile Projector. Bud and cell diameters for 200 cellsfrom each sample were determined by the criteria ofWilliamson (19). The frequency distribution of thesevalues was plotted (Fig. 1), thereby permitting deter-mination of the relative age in the cell cycle of anaverage cell.

Mating. The logarithmically growing cultures to bemated were mixed on membrane filters (MilliporeCorp., 0.45-Aum pore diameter) at a density of 107 cellsof each mating type per filter of 25-mm diameter (4).These filters were then incubated on the surface ofYEPD plates until cells were recovered to be scoredand fixed for electron microscopy.

Preconjugatory stages were kindly provided by L.H. Hartwell and Linda Wilkinson. Haploid strainA364A a was blocked at this stage by the addition of amating factor (4) and S2072 a by a similar factor froma culture of a mating type cells (18).

Electron microscopy. Cells were fixed in 3% glu-taraldehyde in a buffer containing 0.1 M cacodylate,pH 6.8, and 5 mM CaCl2 at 20 C for 30 min and thenat 0 C for about 16 h. Fixation of cultures in station-ary phase was preceded by treatment for 10 min at20 C with 0.1 M ,-mercaptoethanol in 0.02 M

ethylenediaminetetraacetic acid (Sigma) and 0.2M tris(hydroxymethyl)aminomethane-hydrochloride(Sigma), pH 8.1, to facilitate later removal of walls.After glutaraldehyde fixation, walls were removed byincubation of the washed cells in !Jo volume glusulase(Endo Laboratories) in 0.2 M phosphate-citratebuffer, pH 5.8. Cells were subsequently postfixed for60 min at 0 C in 2% osmium tetroxide in 0.1 Mcacodylate, pH 6.8, and 5 mM calcium chloride, andthen rinsed with water, treated for 60 min at 20 Cwith 2% aqueous uranyl acetate, dehydrated, andembedded in Spurr resin (15). Blocks were seriallysectioned on a Sorvall MT-2 ultramicrotome; theribbons were picked up on formvar films on 1 by 2-mmoval single hole grids, stained successively with uranylacetate and lead citrate, and viewed in a Philips EM300 electron microscope. Cellular structures weremeasured in electron micrographs at a total mag-nification of 10,000 diameters.

RESULTSSpindles and spindle plaques in the divi-

sion cycle. Our initial observations on themorphology of logarithmically growing yeastlargely confirmed earlier studies (8, 9, 12) andare summarized diagrammatically in Fig. 2a tof. Unbudded yeast cells possess a single, dis-coidal dense structure, the spindle plaque, em-bedded in the nuclear envelope; microtubules ofthe intranuclear spindle end abruptly on its in-tranuclear surface. Near the time of earliest budemergence, the plaque has become duplicatedsuch that two similar plaques lie side by side inthe nuclear envelope, separated from one an-other by a specialized connecting structure, theplaque bridge. This configuration of side-by-side plaques (9), which we shall henceforth terma double plaque, persists during the earlierphase of bud enlargement and eventually sepa-rates into two distinct plaques which form thepoles of a complete intranuclear spindle. Thespindle rapidly achieves a stable length slightlyless than the average diameter of the nucleus,its poles usually lying within cytoplasmic in-dentations of the nuclear envelope. This lengthof spindle persists until the bud reaches a sizesimilar to that of the mother cell. The nucleusthen migrates into the neck between the celland bud. Renewed elongation of the spindlethen occurs, apparently forcing the poles of thenucleus into the distal portions of the cell andbud (8). The nucleus subsequently pinchesapart within the neck and the cytoplasm isdivided by a process involving the fusion ofcytoplasmic vesicles. Cytokinesis is completedby the deposition of wall material in this region(8).From among these events we have concen-

trated our attention on the temporal relation-

FIG. 1. Cumulative distribution of bud size (b/c,bud diameter/cell diameter) in asynchronous popula-tions of haploid (A364A) and diploid (AP-1) cellsgrowing in medium PSP2.

J. BACTERIOL.


.org/D

ownloaded from

http://jb.asm.org/

SPINDLE PLAQUES IN CELL CYCLE AND CONJUGATION

FIG. 2. Diagram indicating the behavior of spindle plaques and satellites (both stippled), half-bridges(bold lines), and microtubules (straight lines) during the budding cycle and the conjugation process of Sac-charomyces cerevisiae in ideal cross-sectional views (except perspective views of nuclear fusion during conjuga-tion in i-k).

ship between plaque duplication and budemergence, attempting to determine whetherthe order of events is consistent with a func-tional interaction between these processes. Be-cause we are constrained to observe the struc-ture of fixed, rather than living, cells, the tempo-ral sequence must be deduced from morphologi-cal clues in cells of an asynchronous population.The extent of progress through the buddingcycle by an individual cell is indicated by theratio of diameters of the bud and cell (19).Figure 1 shows a cumulative frequency distribu-tion of this ratio, determined from measure-ments of light micrographs. The dotted linerepresents the increase in relative bud-to-celldiameter by an average cell as it progressesthrough its cell division cycle. This determina-tion is less precise during the early phase of budemergence when buds too small to be scored bylight microscopy would be resolved by electronmicroscopy. Therefore, bud emergence actuallybegins slightly earlier than the smallest non-zero ratios recorded. Thereafter, a nearly linearrise in the ratio of bud to cell diameter providesan index of cell age in the budding cycle.Samples from the same cultures were embed-

ded and sectioned for electron microscopy. Rep-resentative stages were selected from those cellsin which the spindles and spindle plaques wereadequately oriented for identification of theirforms and dimensions. These features werecompared with the ratio of bud to cell diameters(averaged between the maximal dimension

found in the serial set of images and the dimen-sion perpendicular to it).We find three classes of plaques (represented

by solid horizontal structures in Fig. 3) in thesecells: (I) single, (II) satellite-bearing single, and(III) double plaques. The single plaque (Fig. 3-Iand 4a to b) consists of a dense disk with anadjacent densely staining inflected membraneof the nuclear envelope, the half-bridge, on oneside; microtubules (paired vertical lines in Fig.3) of the spindle extend into the nucleus (strip-pled in Fig. 3) from the plaque proper andextranuclear microtubules are occasionallyfound to extend outward from an outer layer ofmaterial similar in appearance to the plaquebut of lesser thickness and diameter (the outerplaque). Single plaques are found in manyunbudded cells, particularly those which ap-pear to have completed cytokinesis most re-

I-+*1

77-

.; :

I11

CONJUGATION CS

FIG. 3. Diagrams of sections through the funda-mental forms of the spindle plaque present in succes-sive stages of the cell division cycle: (I single plaque,(II) satellite-bearing single plaque (which persistsduring conjugation), (III) double plaque. The com-plete spindle (CS) bears a typical single plaque ateach pole.

513VOL. 124, 1975

11---

--;,tj

.a I .1

---v

27


.org/D

ownloaded from

http://jb.asm.org/


cently as indicated by the irregular contour ofthe cell surface at the glusulase-resistant budscar.

In cells which have lost the irregular contourand are therefore apparently older, a satellite-bearing single plaque (Fig. 3-II and 4c to d) isfound. This form has all the features of thesimple single plaque but in addition carries atthe opposite end of the half-bridge a sphere ofdense amorphous material similar in appear-ance to plaque material. This satellite differsfrom a true second plaque (that is, differs fromthe other half of a double plaque) by threecriteria. First, it is situated wholly on thecytoplasmic side of the half-bridge and contigu-ous nuclear envelope, not embedded in it as is atrue plaque. Second, and probably deriving

a

g.f~b

7'4~ ~4';,, ^'A.t 7as Xt,-

from the first attribute, there are no spindlemicrotubules attached to its intranuclear side.Third, there is no outer plaque-that is, noouter layer of dense material on the cytoplasmicside of the satellite. In addition, it is usually ofsmaller dimensions. As indicated in Fig. 3, thisform of the plaque is also found in conjugation,the details of which are presented later in thisreport.The double plaque (Fig. 3-III and 5b to c)

appears to consist of two single plaques whichshare a common bridge, rather than each bear-ing a separate half-bridge. Both componentsusually display an outer plaque and bear spin-dle microtubules on the intranuclear surface.The two halves of the double plaque differ,however. from typical single plaques with re-

4-.~~~~~~~~~~~~~~~~~~I

'.9 ~ ~ '.s

...Aft

5.-- ,.~~~~~~~~l

.4I4& -'. ' A~~~~~~~~~~~~~~~~~~~~~' ~~~~~eis- e-(%.'I 4

HBr.w

'~~~~~~~~~~~~~~~~~~~- .";.."k-~~~~~~~~~~~

FIG. 4. Detailed electron micrographs of spindle plaques from the earlier phases of the division cycle inlogarithmically growing diploid cells. (a, b) Serial sections of a single plaque. (c, d) Serial sections of asatellite-bearing single plaque. CM, Cytoplasmic microtubule; HBr, half-bridge; OP, outer plaque; S, satellite;SM, spindle microtubule; SP, spindle plaque. Scale lines on this and subsequent plates represent 0.2 gm.

J. BACTERIOL.

i-,' -." t _


.org/D

ownloaded from

http://jb.asm.org/

SPINDLE PLAQUES IN CELL CYCLE AND CONJUGATION 515

gard to their association with extranuclear mi-crotubules. Such microtubules are attached notonly to the outer plaque but also to the centralregion of the external surface of the bridge.Double plaques were observed exclusively in

cells with small buds, never in unbudded cells.This determination depended upon a morestringent definition of early budding than thatapplied for light microscopy, where the smallest

a

buds are not resolved. In the electron micro-scope it is possible to recognize the earliest budas a slight surface evagination containing sev-eral vesicles (usually 40 to 60 nm, but some-times as much as 250 nm, in diameter; 10). Bythis criterion for budding, observations on agrowing diploid culture have revealed the lackof double plaques in 20 unbudded cells and theirpresence in all of 20 budded cells with ratios of

FIG. 5. Diploid cells in early stages of budding. (a) Observations on adjacent sections demonstrated thisspindle plaque (SP) to be a double plaque with cytoplasmic microtubules (CM) directed into the vesicle-ladenearly bud. M, Mitochondrion. (b, c) Serial sections of another cell show both halves of the double plaque andcytoplasmic microtubules (CM) attached to the bridge.

VOL. 124, 1975


.org/D

ownloaded from

http://jb.asm.org/


bud diameter/cell diameter of less than 0.4.Assuming an equal distribution of casesthroughout the age distribution of these unbud-ded and budded phases (each being about 20%of the cell cycle), the elapsed time betweenplaque duplication and bud emergence wouldbe no more than about 2% of the cycle, regard-less of their order of occurrence.The persistence of the double plaque during

bud growth was determined from the ratio ofbud diameter/cell diameter in electron micro-graphs (Fig. 6). The early limit of its appear-ance, as indicated by the presence of a bud, isobscured by the lack of precise timing of thisparameter by light microscopy. The data dem-onstrate, however, that the double plaque re-mains until the ratio of bud to parent cell

6.HAPLOID

_ 5 .

s 4-.&

Z9w .

S~~~~~~~~~~~~~~~~~~-j~~~~~~~~IaLe 2F .

.

1~ .. .

I

O~P __D__

Dl PLOIlDD_.. _ ,sI ;-.

- 0 2 4 6 8 10 0 2 4 6 8 10

b/c (Ovge°ge)

FIG. 6. Correlation between the relative bud size(b/c, bud diameter/cell diameter) and the stage ofspindle development. The dashed line demonstratestwo persistent phases (the double plaque [DP] and aspindle about 1 MAm long) before terminal elongation.

diameter is about 0.35 in the haploid and 0.40 inthe diploid, after which a complete spindleforms. By reference to the frequency distribu-tion in Fig. 1, we may then deduce that thesecells retain double plaques during at least 30% ofthe budded phase of the cycle.As the bud enlarges further, the plaques

separate rapidly from one another to form acomplete spindle (Fig. 2d), which retains astable length of 0.95 + 0.10 ,m for about 65% ofthe budded phase. This length persists until thebud and cell are of equal diameter; by this timethe nucleus has moved into the neck connectingthem (Fig. 7). Then the spindle elongates rap-idly (within about 5% of the cell cycle) to alength of 6 to 8 gm. Nuclear division and cyto-kinesis then ensue as described by Matile et al.(8). The plaques of all stages with completespindles retain the morphology of single plaques(Fig. 3-CS), including the attached half-bridgeand outer plaque with occasional attached ex-tranuclear microtubules.

Association of the double plaque with thebudding site. In making these observations wenoticed that the double plaque was spatiallyas well as temporally associated with the earlybud. To quantify the spatial relationship, wemade measurements on these electron micro-graphs of the distance between the center ofthe double plaque and the center of the neckjoining the early bud to the mother cell.For comparison we measured the distance fromthe center of the neck to the nuclear midpoint(the intersection of the greatest diameter of the

FIG. 7. A complete spindle beginning its terminal phase of elongation in a nucleus lying within the neck of abudded diploid cell. CM, Cytoplasmic microtubule; HBr, half bridge; SM, spindle microtubule; SP, spindleplaque.

J. BACTERIOL.

1.11


.org/D

ownloaded from

http://jb.asm.org/


nucleus with its perpendicular diameter). Allcells analyzed (a total of 34) displayed a smallerdistance from the neck to the double plaquethan to the nuclear midpoint. A set of 13haploids showed an average neck-to-plaque dis-tance of 1.50 Am versus a neck-to-nuclear mid-point distance of 1.86 ,um. A set of 21 diploidsyielded values of 1.26 and 1.95 ,um, respectively.Therefore, the double plaque consistently re-

sides on the side of the nucleus nearer the neck.The orientation of the plaque toward the bud

was verified by measurements of the anglebetween two straight lines from the bridge of thedouble plaque, one directed toward the center ofthe neck and the other drawn perpendicular tothe plane of the bridge (Fig. 8). These dataclearly confirm a favored orientation of thedouble plaque toward the neck of the early bud.We examined the possibility that this associa-tion was maintained most strongly during theearlier phase of bud emergence, but we foundno systematic variation in the magnitude ofthis angle with the age (relative diameter) ofthe buds. It is clear, therefore, that the doubleplaque is oriented toward the bud throughoutthis phase of this cycle.A physical association between the double

plaque and the early bud is further indicated bythe frequent appearance of cytoplasmic mi-crotubules ending in the center of the bridge(Fig. 5). These microtubules, unambiguouslyidentified in 10 of 21 diploid cases examined,emerge from the bridge along the normal (per-pendicular) axis, and curve toward the bud ifnot initially directed within its margins. Theyfrequently enter the bud itself, where their endsare generally ill-defined. Budding vesicles, 40nm or more in diameter and of low electrondensity content, are invariably seen near thedistal (bud) end of these microtubules. But theassociation may be fortuitous because the bud isgenerally filled with these vesicles.

Stationary phase. Yeast cells from culturesgrown in YEPD to stationary phase are foundpredominantly to be unbudded single cells in Gl(20). We have examined both haploids anddiploids under these conditions to determinethe configuration of the spindle plaques withrespect to this phase of the cell cycle. Althoughfixation for electron microscopy was not ideal,all observations demonstrated that these singlecells possess a simple single plaque (Fig. 9). Thehalf-bridge is often prominently stained as arethe spindle microtubules; no satellites indica-tive of satellite-bearing single plaques are seen.The nucleus of a stationary phase cell fre-quently contains an aggregate of elongate struc-

HAPLOID

MUDIPLOID

M) lC0)1

0 30 60 90 120 150 I80ANGLE (DEGREES)

FIG. 8. Frequency distribution of the angle at thebridge of the double plaque between the directiontoward the neck of the early bud and the directionperpendicular to the plane of the bridge.

tures similar to microtubules in form and di-mensions. This aggregate differs from a spindlein that the apparent microtubules are of muchgreater electron density and lack any associa-tion with the spindle plaque. Similar aggregateshave been described in stationary cells previ-ously by Matile and his colleagues (8).Preparation for conjugation. When two

strains of opposite mating type are culturedtogether, the cells form mating pairs whichundergo both cytoplasmic and nuclear fusionbefore the resulting diploid cells commencebudding (13). Synchronization of the haploidcycles, facilitated by mating substances, ap-pears to be required for conjugation, whichoccurs between unbudded cells arrested in G,(4). To obtain haploid cells in preconjugatoryarrest, we applied media enriched for the mat-ing substances of a and a cells to haploid cul-tures of the opposite mating type. Electronmicroscopy (Fig. lOa) of arrested a cells (A364a)revealed that the nucleus invariably contains aplaque indistinguishable from the satellite-bear-ing single plaque formed prior to budding ingrowing cultures. Just as the double plaque isfrequently directed toward the budding sitein growing cultures, the satellite-bearing singleplaque is usually found adjacent to the evagina-tion prominent in a cells arrested by a hormone.This evagination contains a number of vesicles40 nm in diameter and the distal ends ofoccasional extranuclear microtubules arising

VOL. 124, 1975


.org/D

ownloaded from

http://jb.asm.org/

518 BYERS AND GOETSCH J. BACTERIOL.

from the half-bridge of the satellite-bearing Preconjugatory cells also accumulate in mat-single plaque. ing mixtures (described above) of these strains.

Cells of a mating type (S2072a) (18) remain The nuclei of these unbudded cells were foundisodiametric during inhibition of the cell cycle similarly to possess satellite-bearing singlewith a factor, lacking the evagination seen in a plaques.factor-arrested cells of a mating type. But they Conjugation. After the accumulation of un-too bear the same sort of satellite-bearing single budded cells in mating mixtures, zygotes beginplaque seen in the first case (Fig. lOb). to appear. Fixation of these cultures and serial-

-HBr 4*4.s4.

44-

SP6~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ i

* _e* X !

4

0~~~~~~~~~~~~~~~~~~~.1

FIG. 9. The nucleus (outlined by dots) of a diploid stationary cell showing a single spindle plaque (SP) andprominent half bridge (HBr) as well as dense fibers (DF) similar to microtubules. M, Mitochondrion.

singleplaques.S,Satellite; SP, spindle plaque.b.4"'A~~~~~~~~~~~~~~~~~~.

FIG. 10. Haploid cells arrested in the preconjugatory stable by mating substances. Both (a) the a mating typecell (strain A364A) arrested with a factor and (b) the a cell (S2072) arrested with a factor show satellite-bearingsingle plaques. S, Satellite; SP, spindle plaque.


.org/D

ownloaded from

http://jb.asm.org/


section electron microscopy reveals variousstages in the formation of zygotes, the fusion ofthe haploid nuclei, and the commencement ofthe budding cycle by the diploid zygote. Thesestages may be ordered into a sequence of events(Fig. 2g to 1). Stable couples formed between thetwo haploid types before membrane fusion arenot normally observed because our preparationinvolves wall removal, thereby separating thepartners. The earliest paired stage readily foundin our preparations has undergone partial perfo-ration in the region of the fused walls (Fig. 2g).This results in the formation of a medial cyto-plasmic channel similar to that demonstratedfor Hansenula wingei by Conti and Brock (2).No change in nuclear morphology from the stateseen in preconjugatory stages is noted. As the40-nm vesicles begin to fuse with the plasmamembrane at the isthmus connecting the previ-ous haploids, both nuclei retain satellite-bear-ing single plaques. These are always foundadjacent to the isthmus, frequently residing onan evagination of the nuclear envelope orientedin this direction. Both plaques remain orientedtoward the isthmus, and therefore toward oneanother through the passageway, as the nucleimove together (Fig. lla). Cytoplasmic mi-crotubules are frequently seen attached to thesurface of the half-bridge. These microtubulesoften penetrate the isthmus and enter thecytoplasm of the opposite cell, where they mayeither bypass its nucleus or end in the vicinity ofits plaque (Fig. 2h). Cytoplasmic microtubulesappear to have some role in the interactionsbetween the nuclei because stages are found inwhich they clearly interconnect the respectiveplaques of the two nuclei.

Other cases demonstrate that the nuclei thenmove together, led by their plaques, which meetone another near the isthmus. The narrowopening at this point apparently restricts thepassage of much of the nuclear bulk except forthe extended, plaque-bearing region. Frequentobservations of binucleate zygotes with adja-cent satellite-bearing single plaques on the twonuclei indicate a prolonged period in this state.Occasional cases of satellite orientation in anti-parallel directions (Fig. lib) are found; plaquereorientation to a parallel configuration (Fig.llc) may be required at this stage. Formation ofa single zygote nucleus then proceeds by thefusion of these two plaques, thereby resulting infusion of the contiguous nuclear envelopes (Fig.2i and j).The mechanism of plaque fusion was exam-

ined in detail because our original observations(1) had suggested that the first double plaque ofthe zygote arose by end-to-end fusion between

the half-bridges of the two satellite-bearingsingle plaques. Although the small size andcontorted profiles of the fusion plaques in typi-cal crosses defied structural interpretation, thelarger size of plaques formed in crosses betweendiploid strains (2681-9A a/a X 1422-4C a/a)permitted us to interpret the mechanism offusion from serial sections (Fig. 12). Ratherthan fusing end-to-end by their half-bridges,the satellite-bearing single plaques were foundto join laterally along an edge parallel to themajor axis of each plaque as diagrammed (Fig.2h-j). Ordering events from several stages, wefind that fusion occurs first between the twosatellites and the two half-bridges. The plaquesproper subsequently fuse directly together sothe resulting fusion plaque retains the arrange-ment of a satellite-bearing single plaque alongthe axial dimension. In the lateral dimension,the fusion plaque is markedly curved because ofthe angle at which the parental plaques cometogether. This curvature often describes a hem-icylindrical surface about an axis exterior tothe nucleus and parallel to the axial dimension.Profiles in the lateral dimension are sometimesmore V-shaped, each limb of the V beingderived from one parental plaque. This is par-ticularly evident in crosses between a haploid(A364Aa) and a diploid (1422-4C a/a) strain;here, the limbs of the V differ in length,reflecting the different widths of the parentalplaques.Formation of the fusion plaque is followed by

plaque duplication as in a typical cell divisioncycle (Fig. 2k-l), the double plaque eitherretaining the curvature of the fusion plaque orbecoming more nearly planar. The only obviousspecial feature is its location, which is within theisthmus produced by conjugation. Becauseplaque fusion frequently occurs immediatelyafter perforation of the septum between themating cells, the zygote nucleus is presentbefore the isthmus is fully enlarged. This ap-pears to prevent movement out of the narroworifice by the nucleus and the double plaquearising on its surface.The first bud of the zygote usually emerges

from this same region, again revealing a spatialrelationship between the double plaque and thebudding site. The frequency of first buds arisingin this medial location was 83, 74, and 42%,respectively, in crosses described earlier yield-ing diploid, triploid, and tetraploid zygotes.This decrease in the level of medial buddingwith increased ploidy is accompanied by anincrease in the diameter of the isthmus. Sus-pecting that this may relieve the constraint onnuclear movement, we assayed nuclear position

VOL. 124, 1975


.org/D

ownloaded from

http://jb.asm.org/


FIG. 11. Early diploid zygotes before the stage of nuclear fusion. (a) Both satellite-bearing single plaques aredirected toward the developing aperture. (b) Similar plaques in close approach with their axes in antiparallelorientation. (c) As in (b) but in parallel orientation. CM, Cytoplasmic microtubule; HBr, half-bridge; S,satellite; SM, spindle microtubule; SP, spindle plaque.

by Giemsa staining (12) and found nuclei within copy (Fig. 13) reveals that the double plaquethe isthmus less frequently among triploids and lies near the base of the bud, which is entered bytetraploids. More importantly, we observed an cytoplasmic microtubules extending from theinvariant proximity between nuclei and buds: bridge.medial nuclei were found in zygotes with medial DISCUSSIONbuds. Nonmedial nuclei were not only coinci-dent with nonmedial buds, but always occurred The observations reported in this paper de-at the same pole. Moreover, electron micros- fine the manner in which spindle plaque devel-

J. BACTERIOL.


.org/D

ownloaded from

http://jb.asm.org/


a..4

** rVx * .. 2v.

,;M.., . ... ..

4*~~~~~~~~~~~~~~~,-;.4tt'4z e w * A, Zit S,.'SA' ith-a-,-,W,tFkf.zr

FIG. 12. Serial sections of a fusion plaque in an

early stage of nuclear fusion in a tetraploid (diploid xdiploid) zygote. The curved spindle plaque (SP) ispresent in (a) and (b), and the composite satellite (S)is best seen in (c).

opment corresponds with other events of the celldivision cycle and the conjugation process.Similar correspondence with cell cycle eventshad previously been found in the electron mi-croscopy phenotypes (1) of cdc mutants. Takentogether, their findings suggest not only thatspindle plaque behavior is integrated with othercellular processes but also that the spindleplaques play a morphogenetic role in controllingthese processes.At the earliest stage of spindle plaque devel-

opment in the cell division cycle, we find asingle plaque. The same form is present duringstationary phase, which represents an arrest ofthe cell cycle early in G1 (5). In a later portion ofthe unbudded phase of the cycle, the satellite-bearing single plaque arises. This stage is alsoachieved by cells undergoing conjugation. Dur-ing zygote formation, spindle plaques of thisform bear extranuclear microtubules associatedfirst with the developing isthmus and later withthe plaque of the other nucleus as the nucleimigrate together. The formation of a compositesatellite-bearing single plaque, the fusionplaque, marks the end of the prolonged Garrest coincident with conjugation.The next stage of plaque development in the

cycle is the double plaque, which appears toplay a pivotal role in the behavior of bothvegetative and zygotic cells. The present obser-vations on vegetative cultures demonstrate thatthe double plaque is absent in all unbuddedcells and present in all budded ones. Moreover,plaque duplication and bud emergence are alsointerrelated spatially, as indicated by our mea-surements of the position of the double plaqueand its preferred orientation toward the earlybud. But this association does not of itselfpermit us to determine whether one of theseevents may have a role in controlling the other.Clues to their interactions, however, are foundin the temperature-sensitive phenotypes of thecell division cycle mutants (1).The lack of budding by cells mutant in cdc 28

is accompanied by the lack of plaque duplica-tion; these strains retain the satellite-bearingsingle plaque characteristic of unbudded cellsboth in the vegetative cycle and in preconjuga-tory arrest. The double plaque seen in earlybudding was found among these mutants onlyin cdc 4 strains, which show a unique capacityfor repeated budding in the absence of nucleardivision. It is, however, the phenotype of cdc 24which provides perhaps the most compellingevidence that it is plaque duplication which isthe independent event; the arrested cell under-goes complete cycles of plaque development andnuclear division in the absence of any budding,perhaps because of a specific inability to re-

VOL. 124, 1975


.org/D

ownloaded from

http://jb.asm.org/


U... 4..;,....

FIG. 13. An early bud forming on the isthmus of a diploid zygote. Cytoplasmic microtubules (CM) arising atthe double plaque on the medially situated (fusion) nucleus enter the base of the bud.

spond to the stimulus.The distribution of extranuclear mi-

crotubules in vegetative cells suggests, more-

over, that the double plaque controls budemergence. These microtubules extend from thedouble plaque into the bud, their free ends lyingnear budding vesicles, which are thought tocontain enzymes or precursors for the modifica-tion or growth of the bud wall (see reference 8).It seems likely that the microtubules transportthese vesicles to the budding site, as in thecytoplasmic transport by microtubule systemsin a wide variety of organisms (11).

This proposal predicts that the destruction ofmicrotubules should impede bud emergence. S.cerevisiae is quite resistant to colchicine, butthe derivative Colcemid has been shown to bindto a protein similar to other microtubule pro-teins (3). Although loss of microtubules has notbeen proven, appliction of this drug to cellsemerging from stationary phase does indeeddelay the appearance of the first bud, in accordwith the hypothesis.Such control of budding by the double plaque

need not exclude other processes which restrictthe location of budding. Although primulinstaining reveals that the location of budding bydiploid cells is relatively aspecific, haploid cellsmay bud at highly ordered adjacent loci (16). Inother yeasts budding sites may be more pre-cisely determined: the buds of Saccha-

romycodes ludvigii emerge with previous budscars (17) and those of Trigonopsis variabilisoccur at the apices of the triangular cells(14). In such cases of specific bud localiza-tion, the proposed interaction of extranuclearmicrotubules may be restricted to a particularregion of the cell cortex competent to respond tothe stimulus.Bud emergence in logarithmically growing

cells is coincident not only with double plaqueformation but also with chromosomal replica-tion. Williamson (19) determined by autoradi-ography that the replication of nuclear deoxyri-bonucleic acid (DNA) begins approximately atthe time of bud emergence and continues for27% of the total cell cycle time. The data in thisreport demonstrate that the double plaque ispresent during this same period. Therefore, ifthese strains behave similarly, chromosomalDNA replication is coincident with the presenceof the double plaque. Similarly, DNA replica-tion in zygotes has been shown to be coincidentwith bud emergence (13), which we find tocoincide in turn with duplication of the fusionplaque.The cdc mutants reveal, however, that cells

need not retain the double plaque until DNAsynthesis is completed. Strains mutant in cdc 7,which fail even to initiate DNA replication (7),undergo plaque separation to form a completespindle (2). The same spindle behavior is found

J. BACTERIOL.


.org/D

ownloaded from

http://jb.asm.org/


in mutants of cdc 8 and cdc 21 (1), which aredefective in the elongation of replication ofDNA. These strains become arrested with acomplete spindle of length similar to that per-sisting during the period of bud enlargement.Whereas the spindles of normal vegetative cellsagain elongate rapidly as bud growth is com-pleted, the spindles of these mutants do notundergo this final elongation. As in the earlierintegration of plaque duplication with otherearly events of the cell cycle, an integrativemechanism appears to prevent this final elonga-tion of the spindle if chromosome replication isnot completed.

ACKNOWLEDGMENTS

We thank L. H. Hartwell for advice in this research and inpreparation of this manuscript and H. L. Roman for assist-ance with the manuscript.

The research was supported by Public Health Servicegrant GM 18541 from the National Institute of GeneralMedical Sciences. B. B. was supported by a ResearchCareer Development Award from the National Institutes ofHealth.

LITERATURE CITED

1. Byers, B., and L. Goetsch. 1974. Duplication of spindleplaques and integration of the yeast cell cycle. ColdSpring Harbor Symp. Quant. Biol. 38:123-131.

2. Conti, S. F., and T. D. Brock. 1965. Electron microscopyof cell fusion in conjugating Hansenula wingei. J.Bacteriol. 90:524-533.

3. Haber, J. E., J. G. Peloquin, H. 0. Halvorson, and G. G.Borisy. 1972. Colcemid inhibition of cell growth and thecharacterization of a Colcemid-binding activity inSaccharomyces cerevisiae. J. Cell Biol. 55:355-367.

4. Hartwell, L. H. 1973. Synchronization of haploid yeastcell cycles, a prelude to conjugation. Exp. Cell Res.76:111-117.

5. Hartwell, L. H. 1974. Saccharomyces cerevisiae cell cycle.Bacteriol. Rev. 38:164-198.

6. Hartwell, L. H., R. K. Mortimer, J. Culotti, and M.

Culotti. 1973. Genetic control of the cell division cyclein yeast. V. Genetic analysis of cdc mutants. Genetics74:267-286.

7. Hereford, L. M., and L. H. Hartwell. 1974. Sequentialgene function in the initiation of S. cerevisiae DNAsynthesis. J. Mol. Biol. 84:445-461.

8. Matile, Ph., H. Moor, and C. F. Robinow. 1969. Yeastcytology, p. 219-302. In A. H. Rose and J. S. Harrison(ed.), The yeasts, vol. 1. Academic Press Inc.. NewYork.

9. Moens, P. B., and E. Rapport. 1971. Spindles, spindleplaques, and meiosis in the yeast, Saccharomycescerevisiae (Hansen). J. Cell Biol. 50:344-361.

10. Moor, H. 1967. Endoplasmic reticulum as the initiator ofbud formation in yeast (S. cerevisiae). Arch. Mikrobiol.57: 135-146.

11. Olmsted, J. B., and G. G. Borisy. 1973. Microtubules.Annu. Rev. Biochem. 42:507-540.

12. Robinow, C. F., and J. Marak. 1966. A fiber apparatus inthe nucleus of the yeast cell. J. Cell Biol. 29:129-151.

13. Sena, E. P., D. N. Radin, and S. Fogel. 1973. Synchro-nous mating in yeast. Proc. Natl. Acad. Sci. UJ.S.A.70:1373-1377.

14. SentheShanmuganathan, S., and W. J. Nickerson. 1962.Nutritional control of cellular form in Trigonopsisvariabilis. J. Gen. Microbiol. 27:437-449.

15. Spurr, A. R. 1969. A low viscosity epoxy resin embeddingmedium for electron microscopy. J. Ultrastruct. Res.26:31-43.

16. Streiblova, E., and K. Beran. 1963. Demonstration ofyeast scars by fluorescence microscopy. Exp. Cell Res.30:603-605.

17. Streiblova, E., and K. Beran. 1965. On the question ofvegetative reproduction in apiculate yeasts. Folia Mi-crobiol. Praha 10:352-356.

18. Wilkinson, L. E., and J. R. Pringle. 1974. Transient Glarrest of S. cerevisiae cells of mating type a by a factorproduced by cells of mating type a. Exp. Cell Res.89:175-187.

19. Williamson, D. 1965. The timing of deoxyribonucleic acidsynthesis in the cell cycle of Saccharomyces cerevisiae.J. Cell Biol. 25:517-528.

20. Williamson, D. H., and A. W. Scopes. 1962. A rapidmethod for synchronizing division in the yeast, Sac-charomyces cerevisiae. Nature (London) 193:256-257.

VOL. 124, 1975


.org/D

ownloaded from

http://jb.asm.org/

behaviorof spindles spindleplaques in cycleand conjugation ... · duringthe latter part ofg1....

Documents