J.Cell Sci. 24, 81-93 (1977) 81

Printed in Great Britain

CHROMATIN BEHAVIOUR DURING THE

MITOTIC CELL CYCLE OF

SACCHAROMYCES CEREVISIAE

C. N. GORDONDepartment of Molecular Biology and Biocliemistry, University of California,Irvine, California 92717, U.S.A.

SUMMARY

Chromatin behaviour during the cell division cycle of the yeast Saccharomyces cerevisiaehas been investigated in cells which have been depleted of 90 % of their RNA by digestionwith ribonuclease. Removal of large amounts of RNA from the yeast nucleus before treatmentof the cells with heavy metal fixatives and stains permits chromatin to be visualized with ex-treme clarity in thin sections of cells processed for electron microscopy by conventional proce-dures. Spindle pole bodies were also visualized by this treatment, although the associatedmicrotubules were not.

Chromatin is dispersed during interphase and occupies the non-nucleolar region of thenucleus which is known to be Feulgen-positive from light microscopy. Because spindle micro-tubules are not visualized, direct attachment of microtubules to chromatin fibrils could not beverified. However, chromatin was not attached directly to the spindle pole bodies and kineto-chore differentiations were not observed in the nucleoplasm.

During nuclear division chromatin remains dispersed and does not condense into discretechromatids. As the nucleus expands into the bud, chromosomal distribution to the daughtercells is thought to result from the separation of the poles of the spindle apparatus with attachedchromatin fibrils. However, that such distribution is occurring as the nucleus elongates is notobvious until an advanced stage of nuclear division is reached and partition of the nucleus isnearly complete. Thus, no aggregation of chromatin into metaphase or anaphase plates occursand the appearance of chromatin during mitosis is essentially the same as in interphase. Theseobservations indicate that the marked changes in the topological structure of chromatin whichcharacterize mitosis in the higher eukaryotes do not occur in S. cerevisiae.

INTRODUCTION

Cell division in the yeast Saccharomyces cerevisiae has received considerable atten-tion from a number of workers in recent years. Robinow & Marak (1966) discovereda spindle apparatus in the yeast nucleus and succeeding investigations have revealedbroad areas of agreement on the structure of the spindle, its mode of formation and itsbehaviour during the mitotic cell cycle (Moens & Rapport, 1971 ; Byers & Goetsch,1974, 1975 ; Peterson & Ris, 1976). On the other hand, the structure and behaviourof yeast chromatin during the cell cycle is still controversial.

Wintersberger, Binder & Fischer (1975) recently described discrete, condensedbodies seen in smears of yeast sphaeroplasts as 'chromosomes'. The number of' chromosomes' varied from cell to cell in smears of exponential cultures and it wasassumed that this variation represented the topology of chromatin in different stagesof the cell division cycle. Since most of the sphaeroplasts seen in their smears contain

6 CEL 24

82 C. N. Gordon

condensed bodies (e.g. fig. i{a) of Wintersberger et al. 1975) their results suggestthat yeast chromatin is condensed during a substantial part of the cell cycle. By con-trast, Peterson & Ris (1976), from studies using high voltage electron microscopy ofthick (0*25-1 fim) sections and a surface spreading technique, concluded that yeastchromosomes 'do not condense and are not individually visible' during the mitoticcell cycle. On the basis of the number of microtubules counted in cross-sections ofdiploid and haploid cells they concluded that there was probably one non-continuousmicrotubule per genetic linkage group and that mitosis in yeast is essentially orthodox,except for the lack of chromosome condensation.

Condensed chromosomes can be stained for light microscopy in yeasts which havebeen fixed at meiosis I by any one of a range of fixatives; however, none of theseprocedures shows condensed chromosomes in budding yeasts (C. F. Robinow, privatecommunication). In the electron microscope, condensed chromosomes are not ob-served in thin sections of budding yeasts fixed with glutaraldehyde-osmium tetroxide(Robinow & Marak, 1966 ; Moens & Rapport, 1971 ; Byers & Goetsch, 1974, 1975);regions of electron-lucidity can be seen in the nuclei of permanganate-fixed cells andYotsuyanagi (i960) and Williamson (1966) have argued that the lucid regions are'chromosomes' or 'aggregated chromatin'. Lucid regions in permanganate-fixedcells are caused by the preferential leaching out of cellular components (Hayat, 1970).While it is possible that the lucid areas described by Yotsuyanagi (i960) and William-son (1966) may have contained chromatin in the living cell the reaction with per-manganate would preclude a clear judgement as to the state of chromatin aggregationbefore exposure of the cells to the fixative.

The ratio of RNA to DNA in yeast nuclei is about 3, which is nearly 15 timeshigher than the values reported for animal cells (Molenaar, Sillevis-Smith, Rozijn &Tonino, 1970). The inability to see chromatin in thin sections of yeast fixed for elec-tron microscopy by glutaraldehyde-osmium tetroxide might be due to this unusuallyhigh RNA to DNA ratio. RNA and DNA are similar in their uptake of electron-densefixatives and stains and the large amount of nuclear RNA could compete with DNAfor the binding of heavy metal compounds. This paper describes the results of a studyof the yeast nucleus using a procedure in which about 90 % of the cellular RNA ofyeast was removed by ribonuclease before exposure of the cells to heavy metal fixa-tives and stains. When such an RNA-depleted cell was processed for electron micro-scopy, electron-dense material could be seen with extreme clarity in the region of thenucleus known to be Feulgen-positive from light microscopy.

MATERIALS AND METHODSMaterials

Saccharomyces cerevisiae strain SKQ zn was obtained from. Dr Brian Cox, Botany School,Oxford. This strain is a prototrophic diploid with the genotype a/a, ade 1/ +, + /ade 2, + /his 1.Bovine pancreas ribonuclease (RNase) was obtained from Worthington Biochemical Corp.,Freehold, N. J. A stock solution of 10 mg/ml was heated at 80°C for 10 min to destroy deoxy-ribonuclease activity. Glutaraldehyde was obtained from Polysciences, Inc., Warrington, Pa.,or Tousmis Research Corporation, Rockville, Md.

Chromatin behaviour in S. cerevisiae 83

Growth and harvesting of cells; fixation

Cells were grown on YM-i medium (Hartwell, 1967) at 23 CC with shaking to a cell densityof s x io6 cells/ml. The cells were then harvested by low-speed centrifugation, washed oncewith water at room temperature and suspended for 3 h at room temperature in one of thefollowing fixatives: Fixative A, consisting of 4 % glutaraldehyde, o-i M sodium cacodylate,pH 7, 1 mM CaCli; Fixative B, consisting of 4 % glutaraldehyde, 0-5 M sodium acetate, 1mM CaCl,.

The fixed cells were then chilled and either worked up immediately or stored for 1 -2 weeksat 4°C. Aliquots containing 1-5-6 x io8 cells were processed for electron microscopy by oneof the procedures described below.

Processing for electron microscopy

Unless otherwise indicated all operations were carried out in a cold room maintained at 4°C.Procedure A. Cells fixed in fixative A were washed 4 times by alternate suspension and cen-

trifugation in 0 1 M sodium acetate containing 1 mM CaClj,. The cells were then suspended for30 min in a solution consisting of C2 M Tris, 002 M EDTA, pH 94 . After centrifugation,they were washed twice with o-i M sodium acetate, suspended in 4 % uranyl sulphate for 2 h,washed once in 0 1 M sodium acetate and dehydrated and embedded as described below.

Procedure B. Cells fixed in fixative B were washed 5 times with 0 1 M sodium acetate andsuspended in 5 ml 005 M Tris, pH 7 2 . RNase was added to a concentration of 100 /'g/ml andthe cell suspension incubated at room temperature with shaking for 2 h. (During this period,the absorbance at 260 nm released in the supernatant was monitored; a maximum absorbancewas reached in about 90 min). The cells were centrifuged and washed 4 times with OT Msodium, acetate, suspended in 4 % uranyl sulphate for 2 h, washed twice with 0 1 M sodiumacetate and kept overnight at 4°C. The cells were then suspended in 1 % osmium tetroxide,incubated at room temperature for 15 man and centrifuged. After the cells were washed 4 timeswith water, they were suspended in 20 % ethanol and kept at room temperature for 30 min.The cells were then centrifuged and suspended for 25 h in a solution consisting of 4 partsethanol plus 15 parts ethylene glycol. After centrifugation the pellets were processed as des-cribed below under Dehydration and embedding.

Dehydration and embedding

Cells processed by either procedure were washed 3 times with ethylene glycol and once withpropylene glycol by suspension of the cells with a glass rod, followed by centrifugation at 4500rev/min in an angle-head centrifuge. The pellets were suspended in propylene glycol, storedovernight at room temperature, centrifuged, and suspended in 1 ml propylene glycol. To thesuspension in propylene glycol was added 1 ml Spurr's epoxy resin (Spurr, 1969) of the follow-ing composition : ERL-4206, 10 g ; DER 736, 8 g ; nonenylsuccinic anhydride, 26 g ; dimethyl-aminoethanol, 0-2 g. The resulting 2-phase system was warmed briefly to 40-50 °C, stirredwith a glass rod until it formed a single homogeneous phase, and then kept at 40°C for 1 h.Then 2 ml resin were added and after mixing with a glass rod the cell suspension was kept at40 °C for an additional hour. The cells were then centrifuged, washed once with resin by sus-pension and centrifugation, suspended in fresh resin and kept at 40 CC for 1 h. After centrifuga-tion, infiltration was completed by suspending the cells in fresh resin and incubating at 40 °Cfor 2 h. The cells were centrifuged and the pellets were scooped up on a spatula and placed inplastic embedding capsules filled with resin. The resin was polymerized at 70 °C for 16-24 h-

Electron microscopy

Sections were cut with a diamond knife on a Sorval MT2-B microtome with the advancemechanism set at 8co nm (silver to silver-gold). The sections were mounted on mesh gridsor on Formvar-covered 2 x 1 mm slots and allowed to dry. Staining with lead citrate was donein the following way. Lead citrate was prepared as described by Reynolds (1963) and diluted100-fold with o-oi N NaOH. The grids bearing the sections were submerged in the diluted

6-2

84 C. N. Gordon

stain in a teflon container with a screw-cap top. The container top was screwed on tightly,excluding air, and the grids left in the stain solution for 2-3 h. Following this, the grids wereremoved from the stain, rinsed thoroughly with water and allowed to dry.

Specimens were observed in a Philips EM300 electron microscope at 60 kV and with a 50-fim objective aperture. Cells selected for observation and photography were sectioned alongthe long axis, thereby revealing the bud length and age in the cell cycle.

Digestion of fixed cells with RNase; analysis for RNA and DNA

Cells fixed with fixative B for 6 min were washed 5 times with o-i M sodium acetate and sus-pended in 005 M Tris, pH 7-2. RNase was added and the cell suspensions incubated withshaking at room temperature. The reaction was stopped by the addition of diethyl pyrocar-bonate to a concentration of o-i % (Fedorcsak & Ehrenberg, 1966) and the reaction mixturecentrifuged. The absorbance of the supernatant at 260 run provided an index of the extent ofRNase digestion. The pellets were washed 3 times with 01 M sodium acetate and twice withC25 N perchloric acid at 4°C. The nucleic acids were extracted with 0-5 N perchloric acid at70 °C. After clarification of the extracts by centrifugation, total nucleic acids were determinedspectrophotometrically (Spirin, 1958) and DNA was determined by the diphenylamine reaction(Burton, 1956). RNA was then calculated by difference.

RESULTS

Effect of RNase on glutaraldehyde-fixed cells

When yeast cells previously fixed with fixative B were washed free of excess gluta-

raldehyde and incubated with RNase, ultraviolet-absorbing material was released

from the cells into the supernatant, reaching a maximum absorbance at 260 nm of

Table 1. Release of ultraviolet-absorbing material by RNase from yeastcells fixed with glutar aldehyde

Absorbance per cell (x io8) at 260 nm releasedinto the supernatant after incubation with

RNase for

concentration,fig/ml

O I

i - o2O-O

ioo-oo-oI-C + O - I % D P *

5

oo88

min

• 2 1

'35• i c

•37—

—

40 min

o-554'499 2 59-10

—

—

• RNase was added to the cells containing o-i %

150 min

2 4 38-89——

0-300-30

1050 min

9 5 59-07

————

diethyl pyrocarbonate (DP).

about 9 x io"8 per diploid cell. Typical results are shown in Table 1. The release ofabsorbance depends both on enzyme concentration (for a given incubation time) andon time of incubation (for a given enzyme concentration). Relatively little releaseoccurs in the absence of enzyme (row 5). Furthermore, no absorbance is released inthe presence of diethyl pyrocarbonate (row 6). The latter is a known inhibitor ofRNase (Fedorcsak & Ehrenberg, 1966).

Chromatin behaviour in S. cerevisiae 85

Table 2 shows the nucleic acid content of unfixed cells, of fixed cells not treatedwith RNase, and of fixed cells treated exhaustively with RNase (ioo/tg/ml, 2 h).(In the latter case, the release of absorbance at 260 nm was monitored to verify thatthe RNase limit digest had been reached.) Table 2 shows that RNase removes about90% of the cellular RNA from glutaraldehyde-fixed cells but does not affect theDNA content of these cells. In addition, glutaraldehyde does not in itself affect theDNA or RNA content of cells.

Table 2. Nucleic acid content of yeast cells

Nucleic acid content (pg/cell) of

Glutaraldehyde-fixed cellsNucleic Unfixed

acid cells ( - ) RNase ( + ) RNase

RNA 258, 261 253, 256 0-23, 023DNA 0-035, 0036 0-037, 0038 0-037, 0-037

The results of duplicate determinations are shown.

Aldehyde fixation is known to affect the permeability properties of membranes,permitting access to the cell by molecules whose size or other properties wouldnormally result in their exclusion (Hayat, 1973). The results of Tables 1 and 2 showthat the RNase molecule (molecular weight, 13 800) can pass the cell wall and mem-brane, gaining entrance to the cytoplasm. Because most of the cellular RNA of yeastis cytoplasmic, these data do not prove that RNase entered the nucleus and digestednuclear RNA. However, evidence that nuclear RNA has been affected derives fromprior studies by Molenaar et al. (1970). These workers purified yeast nuclei, fixedthem with glutaraldehyde and digested the fixed nuclei with 30 /tg/ml RNase for 30min; 84% of the nuclear RNA was released by this treatment. This value is similar tothat obtained in this work (Table 2) and indicates that the nuclear envelope of gluta-raldehyde-fixed cells does not act as a barrier to RNase.

Ultrastructure of the untreated nucleus

Cells prepared for electron microscopy by procedure A have not been treated withRNase. Fig. 1 shows the nucleus of such a cell with a small bud in early interphase.Prominent ultrastructural features are the nucleolus, which occupies much of thenuclear volume and straddles one side of the nuclear envelope (Robinow & Marak,1966; Sillevis Smitt, Vlak, Molenaar & Rozijn, 1973), and ribonucleoprotein par-ticles, which are slightly but significantly smaller than the ribosomes in the surround-ing cytoplasm (Mundkur, 1961 ; Gordon, 1977). Previous light-microscopic studieshave indicated that the non-nucleolar region is Feulgen positive and hence is pre-sumably the locus of most of the chromatin (Robinow & Marak, 1966).

Later in the cell cycle the nucleus migrates into the bud and is partitioned betweenparent and progeny cells. Untreated cells at successive stages of the cell cycle werecarefully observed. No internal changes in the chromatin-containing (non-nucleolar)

C. N. Gordon

Fig. i. Control cell (not digested with ribonuclease) in early interphase. The dashedline demarcates the nucleolus (no) from the chromatin-containing Feulgen-positiveregion of the nucleus, x 40000.Fig. 2. Control cell in the process of nuclear division. The nucleolus (no) remainsintact and is partitioned between parent and daughter cells, x 40000.

Chromatin behaviour in S. cerevisiae

spb

Fig. 3. Ribonuclease-digested cell in interphase. Compare to Fig. 1. The dashed linedemarcates the chromatin-containing region (ch) from a region which is largely devoidof electron density and which was the site of the nucleolus before ribonucleasedigestion. X 40000.Fig. 4. Ribonuclease-digested cell with the nucleus just prior to its entrance into thebud. The section cuts through most of the spindle and shows one spindle pole body(spb). (The SPB on the opposite side of the nucleus is not contained in this section.)x 40000.

88 C. N. Gordon

region could be observed before, during or after nuclear migration. Fig. 2 is anexample of a migrating nucleus with the persistent nucleolus partitioning itselfbetween parent and daughter cells (Robinow & Marak, 1966; McCully & Robinow,1973) and the ribonucleoprotein particles dispersed throughout the nucleus. Thenon-nucleolar region is diffuse, with no indication of chromatin condensation.

Effect of RNase on nuclear ultrastructure

Interphase. Robinow & Marak (1966) have shown by light microscopy that theFeulgen-positive region in the nucleus of interphase cells is closest to the bud,whereas the Feulgen-negative nucleolus is usually in an opposite or lateral position.Fig. 3 shows the nucleus of an interphase cell prepared by procedure B. The dashedline delineates electron-dense material (ch) in a region adjacent to the bud from aregion of lower electron density. Based on a comparison of randomly selected sectionsof RNase-digested and of undigested interphase cells, it was concluded that theposition of enhanced contrast in the nuclei of RNase-digested cells could be correlatedwith that of the Feulgen-positive region and that the electron-dense material in thisregion was chromatin (Gordon, 1977).

Comparing Figs. 1 and 3 it is evident that the ribosomes, nuclear particles andnucleolus are lost as cytological entities. Chromatin acquires distinct stainability andhas a diffuse character, showing neither condensation nor preferential accumulationin a particular region of the nucleus. (The latter was verified by observing a series ofconsecutive sections through most of the nucleus.)

Nuclear division. Wintersberger et al. (1975) have reported that yeast chromatin iscondensed into discrete chromosome-like structures. Peterson & Ris (1976), whilemaintaining that such condensation does not occur, describe the aggregation ofchromatin during nuclear division into narrow regions between the spindle polestermed 'metaphase' and 'anaphase plates'. Cells processed for electron microscopyby procedure B were examined for these or other possible changes in chromatinmorphology during nuclear division.

Fig. 4 shows the nucleus of a cell which has formed a complete spindle (Moens &Rapport, 1971). The section cuts through one spindle pole body or SPB (term pro-posed at the First International Mycological Congress [Aist & Williams, 1972 ;Kubai, 1975]), whereas the SPB on the opposite side of the nucleus is not containedin this section. (The microtubules which emanate from the SPBs are not visualizedclearly by this preparative procedure.) Note that the chromatin is uniformly dispersedthroughout the nucleus, with much the same character as interphase chromatin (Fig.3). Neither condensation into chromatids nor aggregation into a 'metaphase plate'midway between the spindle poles is evident. Note also the lack of direct attachmentof chromatin to the SPB.

Fig. 5 shows a stage later in nuclear division after the nucleus has entered the bud.A region of reduced electron density surrounds the SPB (dashed line), suggestinglack of direct attachment. The chromatin remains dispersed throughout the nucleuswith no indication of aggregation into an 'anaphase plate'.

Fig. 6 shows a cell in which partition of the nucleus is nearly complete. The

Chromatin behaviour in S. cerevisiae

1

Fig. 5. Ribonuclease-digested cell shortly after entrance of the nucleus into the bud.The dashed line indicates a chromatin-free region surrounding the SPB. x 40000.Fig. 6. Ribonuclease-digested cell at an advanced stage of nuclear division. Thearrows indicate the neck of the dividing nucleus which is free of chromatin and hasthe same electron density as the cytoplasm, x 40000.

9° C. N. Gordon

8

Fig. 7. Ribonuclease-digested cell in which partition of the nucleus is complete andcytokinesis has begun, x 30000.Fig. 8. Ribonuclease-digested cell in which cytokinesis is complete and cell wallseparation has begun, x 30000.

Chromatin behaviour in S. cerevisiae 91

chromatin is still dispersed but is now localized at opposite poles of the dividingnucleus, while the narrow intervening neck is devoid of chromatin (arrows, Fig. 6).Figs. 7 and 8 show the terminal stages of cell division. The chromatin remains dis-persed and has essentially the same cytological characteristics as in previous stages.

DISCUSSION

The formation and behaviour of the spindle apparatus in S. cerevisiae is welldocumented. The SPBs and associated microtubules are readily visualized in cellstreated with glutaraldehyde-osmium and this has made possible a detailed analysis ofspindle behaviour in serial thin sections (Moens & Rapport, 1971 ; Byers & Goetsch,1975). Additional features of chromatin segregation in yeast have recently emergedfrom studies of Peterson & Ris (1976). Chromatin fibrils become attached directly tospindle microtubules without recognizable kinetochores. As nuclear division proceeds,uncondensed chromatin fibrils are drawn to opposite poles of the spindle apparatus.

Peterson & Ris (1976) observed a region of enhanced electron density near the endsof the non-continuous microtubules and considered this to be an aggregation ofchromatin into metaphase and anaphase plates. While the overall aspects of nucleardivision which emerge from this work are compatible with their model, I fail toobserve any preferential aggregation of chromatin into discrete regions. Their con-clusion that such aggregation occurs is, in my opinion, not justified by the micrographsused to support this contention (figs. 12-14 °f Peterson & Ris, 1976). Only a smallregion of the nucleus between the spindle poles is shown and their identification ofelectron-dense material in this region as chromatin is unconvincing.

Chromatin in yeast nuclei which have been lysed by osmotic shock has the ap-pearance of knobby fibrils about 20 run in diameter (Peterson & Ris, 1976). Occasionaldense granules of roughly this size could be seen in some of my micrographs ofRNA-depleted cells and these may possibly represent individual chromatin fibrilscut in cross-section. In general, however, the cross-sections observed in RNA-depleted cells were considerably larger than 20 nm, suggesting that in situ, higherorders of folding occur.

Two aspects of chromatin behaviour emerging from this work seem quite definite.First, condensation into the discrete chromatids with staining characteristics typicalof higher eukaryotes does not occur in S. cerevisiae. Second, while some higher-orderfolding of the basic 20-nm fibril may occur in situ, the general topological structure ofchromatin as visualized in the electron microscope, using these procedures, is cytolo-gically indistinguishable throughout most of cell division. Only when cell division hasreached a relatively advanced stage is a separation into 2 distinct groups evident.

The results described in this work are in sharp conflict with the findings of Winters-berger et al. (1975) that condensed chromosomes are seen at some stages of the mitoticcycle of S. cerevisiae. The preparative procedure used by these authors was severe :cells were divested of their cell wall with snail-gut enzyme and placed on a glass slide.After heating with a bunsen flame, they were allowed to dry before fixation withethanol-acetic acid. It is possible that artifactual aggregation into Giemsa-positive,

92 C. N. Gordon

electron-dense structures occurred on the slide as a consequence of heating anddrying, since during this process the cells were unfixed and lacked the protection of acell wall.

Among lower eukaryotic organisms it is now clear that there are a number ofdeviations from the classically orthodox picture of mitosis in which chromosomescondense and align themselves between the poles of the spindle (reviewed by Kubai,1975). Among the fungal species there is apparently a broad range of mitotic behaviour,including several species in which typically orthodox mitosis seems evident (Kubai,1975). While a number of important details of nuclear division in S. cerevisiae remainto be worked out, at this point the available evidence favours the following scheme.At some stage in the cell cycle, chromatin fibrils become attached to non-continuousspindle microtubules (Peterson & Ris, 1976). The elongation of the spindle concomi-tant with movement of the nucleus into the bud provides the mechanism by whichparental and progeny chromatin separate. Before and during the separation processthe chromatin fibrils have the same topological structure as in interphase and do notundergo additional condensation or supercoiling. Nor does preferential aggregationinto discrete regions of the nucleus occur. Whether mitotic behaviour of this type canbe aptly characterized as 'orthodox' (Peterson & Ris, 1976) is semantical.

I thank Dr Hans Ris for supplying me with a copy of the manuscript by himself and DrPeterson prior to publication. This study was supported by grant BMS 73-06847Aoi from theNational Science Foundation.

REFERENCESAIST, J. R. & WILLIAMS, P. H. (1972). Ultrastructure and time course of mitosis in the fungus

Fusarium oxysporum. J. Cell Biol. 55, 368-389.BURTON, R. K. (1956). A study of the conditions and mechanism of the diphenylamine reaction

for the estimation of deoxyribonucleic acid. Biochem.J. 62, 315-323.BYERS, B. & GOETSCH, L. (1974). Duplication of spindle plaques and integration of the yeast

cell cycle. Cold Spring Harb. Symp. quant. Biol. 38, 123-131.BYERS, B. & GOETSCH, L. (1975). Behaviour of spindles and spindle plaques in the cell cycle

and conjugation of Saccharomyces cerevisiae. J. Bact. 124, 511-523.FEDOHCSAK, 1. & EHRENBERG, L. (1966). Effects of diethyl pyrocarbonate and methyl methansul-

fonate on nucleic acids and nucleases. Acta chem. scand. 20, 107-112.GORDON, C. N. (1977). Ribonucleoprotein particles in the yeast nucleus. (Submitted,?. Cell

Biol.)HARTWELL, L. H. (1967). Macromolecule synthesis in temperature-sensitive mutants of yeast.

J. Bact. 93, 1662-1670.HAYAT, M. A. (1970). Fixation. In Principles and Techniques of Electron Microscopy, vol. 1,

pp. 61-62. New York : Van Nostrand Reinhold.HAYAT, M. A. (1973). Specimen preparation. In Electron Microscopy of Enzymes, vol. 1 (ed.

M. A. Hayat), pp. 1-43. New York : Van Nostrand Reinhold.KUBAI, D. F. (1975). The evolution of the mitotic spindle. Int. Rev. Cytol. 43, 167-227.MCCULLY, E. K. & ROBLNOW, C. F. (1973). Mitosis in Mucor hiemalis. A comparative light and

electron microscopical study. Arch. Mikrobiol. 94, 133-148.MOENS, P. B. & RAPPORT, E. (1971). Spindles, spindle plaques, and meiosis in the yeast

Saccharomyces cerevisiae (Hansen). jf. Cell Biol. 50, 344-361.MOLENAAR, I., SILLEVIS-SMITH, W. W., ROZIJN, TH., H. & TONINO, G. J. M. (1970). Bio-

chemical and electron microscopic study of isolated yeast nuclei. Expl Cell Res. 60, 148-156.

Chromatin behaviour in S. cerevisiae 93

MUNDKUR, B. (1961). Electron microscopical studies of frozen-dried yeast cells. II. The natureof the basophile particles and vesicular nuclei in Saccharomyces. Expl Cell Res. 25, 1-23.

PETERSON, J. B. & Ris, H. (1976). Electron-microscopic study of the spindle and chromosomemovement in the yeast Saccharomyces cerevisiae. J. Cell Sci. 23, 219-242.

REYNOLDS, E. S. (1963). The use of lead citrate at high pH as an electron-opaque stain inelectron microscopy. J. Cell Biol. 17, 208-212.

ROBINOW, C. F. & MARAK, J. (1966). A fiber apparatus in the nucleus of the yeast cell. J. CellBiol. 29, 129-151.

SILLEVIS-SMITT, W. W., VLAK, J. M., MOLENAAR, I. & ROZIJN, TH. H. (1973). Nucleolarfunction of the dense crescent in the yeast nucleus. Expl Cell Res. 80, 313-321.

SPIRIN, A. S. (1958). Spectrophotometric determination of total nucleic acid content. Biok-himiya 23, 617-622.

SPURR, A. R. (1969). A low-viscosity epoxy resin embedding medium for electron microscopy.J. Ultrastruct. Res. 26, 31-43.

WILLIAMSON, D. H. (1966). Nuclear events in synchronously dividing yeast cultures. In CellSynchrony (ed. I. L. Cameron & G. M. Padilla), pp. 81-101. New York : Academic Press.

WINTERSBERGER, U., BINDER, M. & FISCHER, P. (1975). Cytogenic demonstration of mitoticchromosomes in the yeast Saccharomyces cerevisiae. Molec. gen. Genet. 142, 13-17.

YOTSUYANAGI, Y. (i960). Mise en Evidence au microscope electronique les chromosomes de lalevure par une coloration sp^cifique. C. r. Hebd. Se"anc. Acad. Sci., Paris 250, 1522-1524.

(Received 19 July 1976)