Reorganization of chromatin in Xenopus egg extracts: electron microscopic studies

Original article

Reorganization of chromatin in Xenopus egg extracts:electron microscopic studies

Chandra K. Pyne*

Groupe biologie expérimentale, UMR 7622, Biologie moléculaire et cellulaire du développement,CNRS–Université Pierre-et-Marie-Curie, 9, quai Saint-Bernard, 75005 Paris, France

Received 23 May 2000; accepted 31 January 2001

The structural basis of mitotic condensation of chromosomes is one of the problems of cell biology yet to beelucidated. A variety of approaches have been used to study this problem and a large number of hypotheseshave been proposed to explain the different levels of compaction of chromatin. Xenopus egg extracts, nowwidely used to study various aspects of cell biology, provide a valuable tool to study mitotic condensationof chromosomes. No detailed study has however yet been reported on the submicroscopic organization ofcondensed chromosomes in vitro in egg extracts. We present here the results of our electron microscopicstudies on the organization of condensed chromosomes in vitro, using demembranated sperm nuclei andmitotic (CSF-arrested) extracts of Xenopus laevis eggs, clarified by high speed centrifugation. Uponintroduction of sperm nuclei in egg extracts, the nuclei swell and the chromatin undergoes a rapiddecondensation; at this stage the chromatin is formed of 10 nm fibrils. After longer incubation, the chromatincondenses, and by 2 h chromosomal structures can be visualized by staining with DAPI or Hoechst 33258.Our results on the organization of chromosomes in different stages of condensation are discussed in relationto the different hypotheses proposed to explain the process of mitotic condensation of chromosomes. Finally,this study demonstrates the feasibility of high-resolution analysis of the process of chromosome condensa-tion. © 2001 Éditions scientifiques et médicales Elsevier SAS

chromatin / chromosomes / egg extracts / electron microscopy

1. INTRODUCTION

The organization of chromatin in its different statesand in different types of nuclei still remains one of theunsolved problems in cell biology. Though there nowseems to be an unanimity that the chromatin is basi-cally organized as a chain of nucleosomes in a fibrillarstructure, about 10 nm in diameter, the higher orderorganization is still a subject of controversy. Electron

microscopic studies have revealed the presence of fi-bers around 30 nm in diameter in mitotic chromosomesand in different types of nuclei (e.g. Adolph, 1980;Rattner and Lin 1985; Gianasca et al 1993). But themechanism of compaction of the 10 nm nucleosomalfibers into 30 nm fibers, as well as the compaction ofthe latter in mitotic chromosomes and in condensedchromatin in different types of nuclei, still remain to beelucidated. A variety of approaches have been used tostudy this problem, the most common of which hasbeen the treatment of chromatin or chromosomes withhypotonic solutions of different salt concentrations(e.g., Thoma and Koller, 1977; Thoma et al. 1979;

* Correspondence and reprints: 43, rue Vasco-de-Gama, 75015Paris, France.

Biology of the Cell 93 (2001) 309−320© 2001 Éditions scientifiques et médicales Elsevier SAS. All rights reservedS0248490001011170/FLA

Mitotic condensation of chromosomes Pyne

Adolph 1980; Woodcock and Horowitz, 1995). Anotherapproach has been the removal of histones from mi-totic chromosomes (Paulson and Laemmli, 1977; Laem-mli et al., 1978). On the basis of these and various otherapproaches, a large number of hypotheses have beenproposed to explain the different levels of compactionof chromatin (see Discussion).

The introduction of amphibian egg extracts as anexperimental medium for studying chromatin organi-zation by Lokha and Masui (1983) has provided anovel approach to study chromosome dynamics aswell as other cellular processes. These authors showedthat when demembranated sperm nuclei of Xenopuslaevis are introduced into extracts of Rana pipiens eggs,the chromatin undergoes decondensation, DNA repli-cation and then recondensation into chromosomes.Since then, Xenopus egg extracts have become awidely used medium for studying various cellularprocesses, such as chromatin decondensation, DNAreplication, transcription, nuclear envelope and spindledynamics, as well as mitotic condensation of chromo-somes etc. (see Almouzni and Wolffe, 1993, for areview). Using this system, Hirano and his collabora-tors have isolated several SMC (structural maintenanceof chromosomes) proteins of Xenopus, and studied therole of these proteins in the dynamics of chromosomes(Hirano and Mitchison, 1993, 1994; Hirano et al., 1997;Losada et al., 1998, 2000; see Hirano, 2000 for a review).Some of the authors utilising Xenopus egg extracts fortheir studies have used electron microscopy, but themain interest of these authors being elsewhere, theydid not study the process of chromosomal condensa-tion nor the submicroscopic organization of chromo-somes condensed in vitro in such extracts (Sheehan etal., 1988, Sawin and Mitchison, 1991; Shamu and Mur-ray, 1992; Desai et al., 1997). We present here the resultsof our electron microscopic studies on the reorganiza-tion of chromatin of sperm nuclei, and particularly theprocess of chromosomal condensation, in clarified mi-totic (CSF-arrested) egg extracts.

2. MATERIAL AND METHODS

Xenopus laevis, originally obtained from AnimalBreeding Station of CNRS, Rennes, France, were bredin our laboratory. Demembranated sperm nuclei wereobtained as described by Newmeyer and Wilson(1991), and stored in liquid nitrogen at a concentrationof around 10 000 sperm heads µL–1. Mitotic (CSF-arrested) egg extracts were obtained as described byHirano and Mitchison (1993, 1994) in the followingmedium : Sodium �-glycerophosphate 80 mm, MgCl215 mM, EGTA 20 mM, Hepes 10 mM (pH 7.4), DTT1 mM, cytochalasin B 10 µg mL–1. Anti-proteases, leu-peptin, chymostatin and pepstatin, were added at at10 µg mL–1 each. Eggs were crushed by centrifugation

at 10 000 rpm in a HB-4 rotor in Sorval centrifuge for15 min at 16°C. The crude extract was centrifuged at50 000 rpm for 30 min at 4°C in a TLS rotor andBeckman TL-100 ultracentrifuge. The clarifed extractwas further purified by centrifugation at 50 000 rpmfor 30 min. Energy regenerating system (ATP, phospho-creatine and creatine phosphokinase) was added to thesystem; 50 µL aliquots were frozen in liquid nitrogenand stored at –80°C.

For studying the reorganization of chromatin, thesperm nuclei were diluted in the sperm dilution buffer(Murray 1991) and added to the egg extracts at aconcentration of 200 µL–1 and incubated at 23° C fordifferent times. For driving the nuclei into interphase,CaCl2 was added to a concentration of 0.4 mM; after60 min equal volume of fresh egg extract was added toinduce condensation of chromatin (Hirano and Mitchi-son, 1993, 1994). For examination under the light mi-croscope, 3 µL samples were removed at differentintervals and added to an equal volume of a solution ofHoechst 33258 at 1 µg mL–1, and examined underepifluorescence. For electron microscopy, the sampleswere prepared by flat-embedding as described else-where (Pyne et al 1989). Briefly stated, isolation cham-bers were prepared by fixing coverslips onto boredglass slides with paraffin; the samples were put in suchchambers and covered with another cover slip. Thesechambers were centrifuged in a refrigerated Jouancentrifuge equiped with buckets with flat bottomsdesigned specially for centrifuging slides. Centrifuga-tion was done for 5 min at 50 × g and then for 25 min at3000 × g (the temperature of the centrifuge varied from4° to 8°C). After centrifugation, the slides were floodedwith the fixative solution, the cover slip gently pushedaway, and a few drops of fixatives was gently added tothe isolation chambers. For fixation, we used 2% glut-araldehyde in 0.1 M phosphate or cacodylate buffer, for30 min at room temperature. No difference was ob-served between the two buffers. The preparation werethen rinsed three times using the same buffer. DAPI,1 µg mL–1 in the same buffer, was added to the thirdrinsing solution, and the preparations thus stained for15 min. After another brief rinse, the prepartions werecovered with another cover slip, gently wiped with atissue paper sheet (Kimwipe), and the edges of thecover slips sealed with nail polish. The slides couldthus be inverted for examination and photographed,using a Leitz Dialux 22 microscope equiped for epif-luorescence. Following micrography, the coverslipbearing the centrifuged material was carefully re-moved, carried through ethanol series for dehydrationand embedded in Epon as described (Pyne et al 1989).Thin sections were cut using a diamond knife and aReichert ultramicrotome, picked on uncoated grids andstained with a half saturated solution of uranyl acetatein 50% ethanol, with or without post-staining in lead

310 Biology of the Cell 93 (2001) 309–320


citrate. (For examination of decondensed chromatin, itwas found necessary to stain the grids with leadcitrate; for condensed chromosomes, this post-stainingwas not necessary). The grids were floated on a sus-pension of 10 nm colloïdal gold particles (actuallyStreptavidin-gold conjugate; British Biocell Interna-tional) for 1 min and rinsed briefly, for internal calibra-tion. Grids were examined with a Philips 201 electronmicroscope, at 80 kV. Stereo micrographs were taken atan angle of –6° and +6°.

3. RESULTS

Isolated sperm nuclei, fixed in glutaraldehyde beforecentrifugation onto coverslips are shown in figures 1aand 2a. In electron micrographs of thin sections, a densehomogenous organization is observed, in which nofibrillar structures can be differentiated (figure 2a).When the sperm nuclei are suspended in cacodylatebuffer (0.1 M, pH 7.4), centrifuged for 30 min at 4–8°C,and then fixed, fibrillar structures about 20 nm indiameter can be observed (figure 2b). There has appar-ently been some swelling and/or extraction of material

surrounding these fibers. These fibers can be comparedto those observed by Giannasca et al (1993) in thesperm nucleus of the starfish Patiria minata fixed inartificial sea water.

The swelling of sperm nuclei in egg extracts isknown to be a rapid process (Philpott et al 1991). Weobserved that when the sperm nuclei were introducedinto egg extracts and put to centrifugation immediatelyfor 30 min at 4–8°C, the sperm chromatin underwentcomplete decondensation (figures 1b, 2c, e). (Such aprocedure, i.e., centrifugation before fixation, besidesproviding a better adhesion of chromatin to the coverslips also avoided the problem of dilution of spermnuclei in isolation chambers and thus provided higherdensity of sperm nuclei in the final preparations). Finefibrils, around 10 nm or less in diameter could beobserved, comparable to those observed by Sheehan etal (1988). No nucleosomes could be observed in suchdecondensed chromatin, as is always the case whenroutine electron microscopy procedures are used. Thechromatin continued to be in the decondensed stateuntil 60 min incubation (figures 1c, 2d, f). In these latterpreparations, amongst the thin 10 nm fibers, we regu-larly observed a few dense structures (arrows in figure2d, f). Their nature remains to be determined; they areunlikely to be initiation foci for chromatin condensa-tion, as their organization is not comparable to that ofcondensing chromosomes described below.

In our experiments, we have added fresh mitoticextracts after 60 min incubation of sperm nuclei in eggextracts containing calcium, to induce condensation ofchromatin. We may mention here that the chromatindoes not undergo replication in clarified egg extracts(Hirano and Mitchison 1993); as such what we aredescribing below are unreplicated condensed chroma-tids. For simplicity, however, we shall use the termchromosomes for these structures, just as earlier au-thors (Hirano and Mitchison 1993, 1994, Hirano et al1997, Losada et al 1998, Ohsumi et al 1993, Cubizolleset al 1998).

In sperm nuclei prepared for electron microscopy20 min after addition of fresh extract, dense aggregatesof chromatin fibers can be observed in thin sections(figures 3a, 4a). These aggregates can be interpreted asinitiation foci for chromosomal condensation, and thuscompared to chromomeres of classical litterature (seeLima de Faria 1975, and earlier references cited therein)These foci could not be discerned under epifluores-cence. They appear as strands of varying lengths andthickness, from 50 to 100 nm or more in diameter; thisvariation is more likely to be due to different degrees ofcondensation of chromatin fibers rather than to theplane of section. In some cases these strands show ahelical twist (figure 4a, arrowhead). These condensationfoci or chromomeres are basically formed of 10 nmfibrils, just as those around them. The latter evidently

Figure 1. Immunofluorescence micrographs of isolated spermheads incubated in Xenopus egg extracts for different times. Allthe micrographs are from material centrifuged in isolation cham-bers, fixed in buffered 2% glutaraldehyde, rinsed and stained inDAPI; these preparations were then used for electron microscopy.a. Sperm heads fixed before centrifugation. b. A sperm headincubated in egg extract during centrifugation, 30 min at 4–8°C,before fixation. c. A sperm head after 60 min incubation.d. 120 min incubation. ×1000; the bar represents 10 µm for allmicrographs.

Biology of the Cell 93 (2001) 309–320 311


Figure 2. a. Electron micrograph of a thin section of a sperm head fixed in 2% glutaraldehyde in 0.1 M Na-cacodylate buffer beforecentrifugation (from the same preparation that is shown in figure 1a. b. Thin section of a sperm head centrifuged in 0.1 M Na-cacodylatebefore fixation; note the difference compared to figure 2a. c. A sperm head centrifuged in egg extract, for 30 min at 4–8°C, beforefixation. d. A sperm head incubated for 60 min in egg extract before centrifugation and fixation; note the dense structure in the center(arrow). e and f. Higher magnifications of parts of c and d respectively. Magnification : a–d : ×23 000, e–f : ×46 000; the bar represents1 µm in all micrographs. Note that the small black dots on these and all subsequent electron micrographs represent 10 nm colloidal goldparticles.



Figure 3. a. Electron micrograph of a thin section of a nucleus after 80 min incubation in egg extract; the arrowheads indicate chromatinstrands showing a helical twist. b. Electron micrograph of a nucleus after 100 min incubation. c, d. Electron micrographs of nuclei after120 min incubation in egg extracts from two different experiments; note that we observed some variability in the degree of condensationof chromosomes in different experiments. Magnification : a, b, and d : ×24 000; c : ×16 000; the bar represents 1 µm in all micrographs.



represent chromatin fibers which have not yet con-densed. The possible existence of loops radiating outfrom these foci should also be taken into account,particularly as such loops can be observed in chromo-somes incubated for longer time (figure 5, arrows). Thisinitial stage of condensation of chromatin can be con-sidered as comparable to that of very early prophase individing somatic cells. Such an organization suggeststhat the chromosomal condensation in egg extracts isnot a synchronous process. 40 min after addition offresh extract, the proportion of decondensed chromatinis highly reduced, and most of the chromatin appearscondensed in structures about 100 to 200 nm in diam-

eter (figure 3b). Such electron micrographs suggest thatas the chromatin fibers condense, they produce morechromomeres and/or become progressively incorpo-rated into neighbouring pre-formed chromomeres.This process thus produces an increase in the numberand/or the volume (i.e., in length and thickness) ofchromomeres, until the latter join up to form thechromosome axis.

After 120 min incubation of sperm nuclei (i.e. 60 minafter addition of fresh extract), the chromosomes at-tained their maximal condensation in our preparations;further incubation upto 180 min did not produce fur-ther condensation. In preparations examined under



epifluorescence after a slight squeezing in between thecover slip and the slide, isolated chromosomes couldsometimes be observed (not shown here). In spermnuclei centrifuged onto cover slips and prepared forelectron microscopy, however, the chromosomes of thesperm nuclei remained aggregated (figure 1d for epif-luorescence, figures 3 c, d, 4b, c and 5 for electronmicroscopy). We observed some variation in the degreeof condensation of chromosomes in our preparations;figures 4b, c, and 5 represent thin sections of spermnuceli after 120 min incubation in different experi-ments and show chromosomes in different degrees ofcondensation. It may be mentioned that this differencewas not apparent under epifluorescence. These chro-mosomes also appear to be formed basically of 10 nmfibrils, though in some cases these fibrils are observedin a loose helical organization, which by further pack-ing down the coils could produce thicker fibers (figure4b, arrow). The compaction of the chromatin fibers andtheir incorporation into chromomeres does not appearto be a simple coiled coil packing, even if in certainregions strands showing a helical twist can be ob-served (figure 4a, arrowhead); other processes, such assuper coiling, irregular folding and twisting may alsoplay some role at this level of compaction. Loops canbe observed at the periphery of condensing chromo-somes (figure 5, arrows), but not in chromosomes inmore advanced stages of condensation (figure 4c). Noaxial or core structure distinct from the chromatinfibers, which could be interpreted as a possible chro-

mosomal scaffold, has been observed in these condens-ing chromosomes. It is evident that in ourpreparations, the chromosomes do not condense to thesame degree as in metaphase of somatic cells; as suchthe structures described here should be considered as‘partially condensed chromosomes. The possible rea-sons for this observation are discussed below (seediscussion).

4. DISCUSSION

The major problem in the study of chromatin orga-nization is the extreme sensibility of chromatin todifferent preparative procedures (Belmont et al., 1987)It is normally assumed that the protocol generally usedfor routine electron microscopy – glutaraldehyde fixa-tion, ethanol dehydration and epoxy embedding(which we have used for the present study) – intro-duces artifacts. Thus for exemple, Horowitz et al.(1990) have shown that the chromatin fibers are ob-served to have different diameters using different pre-parative procedures : with glutaraldehyde fixation andepoxy embedding, chromatin fibers were observed tohave a diameter of 20 nm, while with low temperatureembedding in the hydromiscible resins Lowicryl K4Mor K11M, these fibres had a diameter around 30 nm. Assuch, the values of thickness of chromatin fiber givenhere and in other studies using epoxy embeddingshould be considered only as approximative values.

Figure 4. a–c. Stereo electron micrographs of condensed chromosomes, at –6° and +6°. a is a part of figure 3a at higher magnification;the arrowhead indicates a chromatin strand showing a helical twist. b is a part of figure 3d at higher magnification; arrows indicatechromatin fibers showing a loose helical packing ; arrowhead indicates a condensing chromatin strand. Magnification : ×58 000 for allmicrographs; the width of each micrograph represents 1 mm ; the bar in a represents 100 nm for all micrographs.



The chromatin of sperm nuclei rapidly decondenseswhen introduced into egg extracts. Under the electronmicroscope, 10 nm fibrils are seen, in which no nucleo-somal structures are observed (Sheehan et al., 1988,present studies), as is the case for nuclei examined byroutine procedures for electron microscopy. Duringthis swelling, the chromatin undergoes changes in itsbasic protein composition (see Ohsumi and Katagiri,1991; Philpott and Leno, 1992 for details). It remainshowever to be determined how far the decondensedchromatin observed in swollen sperm nuclei incubatedin egg extracts corresponds to decondensed chromatinin somatic nuclei.

We have mentioned above that the structures de-scribed here (figures 3–5) represent partially condensed

chromosomes. A casual examination of the publishedimmunofluorescence micrographs of in vitro con-densed chromosomes shows that the chromosomes donot condense to the same degree in high speed eggextacts as in metaphase of somatic cells (e.g., Cubi-zolles et al., 1998; compare their figure 4 with figure 3j-o ; see also Hirano and Mitchison 1993, 1994). Thedegree of chromosomal condensation seems to be inde-pendent of replication of chromosomes (see Losada etal 1998, their figures 6a–c, showing unreplicated andreplicated chromosomes) but seems to depend essen-tially on whether the egg extracts are submitted to highspeed centrifugation or not. In their studies on spindleand centromere dynamics, Sawin and Mitchison (1991)and Shamu and Murray (1992), using low speed ex

Figure 5. A higher magnification of a part of figure 3c. A few of the loops are indicated by arrows. Magnification ×58 000, Bar represents1 µm.



tracts, have published electron micrographs of chromo-somes condensed to a higher degree than those ob-served in high speed extracts. Their publishedmicrographs however led us to the conclusion thatthese do not shed any light on the submicroscopicorganization of chromosomes, and as such, followingtheir procedure would serve no purpose for studyingthe process of chromsomal condensation.

Several factors may be invoked to explain this differ-ence in the degree of condensation of chromosomes invitro in egg extracts and in vivo at metaphase insomatic cells. Hirano and Mitchison (1994) have notedthat chromosomes condensed in vitro have a simplerprotein composition than that reported for the chromo-somes isolated from somatic cells. Histone H1 is appar-ently absent in Xenopus eggs and is replaced byhistone B4, which shows a 30% amino acid sequenceidentity with the histone H1 (Smith et al., 1988; Hock etal., 1993). Chromatin isolated from early cleavage-stageXenopus embryos have been found to be deficient inhistone H1 (Wolffe, 1981, 1989; Ohsumi and Katagiri1991). Hock et al (1993) confirmed the absence ofhistone H1 in chromosomes of such embryos throughimmunolabeling studies. It would be interesting tostudy the fine structure of chromosomes in earlycleavage-stage of Xenopus embryos (i.e., in stageswhen the chromosomes do not contain histone H1). Inyeast, which also apparently lack histone H1 (Grun-stein, 1990), the chromosomes do not condense to thesame degree as in higher eukaryotes (Guacci et al.,1994).

While considering the condensation of chromatininto mitotic chromosomes, two distinct levels of com-paction have to be taken into account – the transforma-tion of the 10 nm nucleosomal fiber into 30 nm fiberand the compaction of the latter into mitotic chromo-somes. A detailed discussion of all the hypotheses thathave been advanced to explain these different levels ofcompaction of chromatin into mitotic chromosomeswould be beyond the scope of this article. The variousmodels proposed for the compaction of 10 nm nucleo-somal fiber into 30 nm fiber can be grouped into twobroad categories : (i) the solenoid model or its variants(Finch and Klug, 1976; Thoma et al., 1979; Daban andBermudez 1998); (ii) a more open conformation, eitherin a flattened zigzag structure (Bednar et al., 1998;Woodcock and Horowitz, 1995) or in an irregularquasi-helical conformation (van Holde and Zlatanova1995). Most of these studies have been done on isolatedchromatin or on isolated polynucleosomes, and thequestion remains – how far isolated chromatin corre-sponds to chromatin in vivo? The study of chromatinorganization in egg extacts provides an alternativeapproach, in that the conditions are more near to invivo conditions than in simple salt solutions, even ifthey are not quite identical to those in the nucleoplasm

Figure 6. A schematic drawing of the proposed model of mitoticcondensation of chromosomes. (Not drawn to scale). a. Chroma-tin in fully decondensed state is formed of 10 nm fibrils. In eggextracts, the totality of chromatin becomes simultaneously decon-densed, which may not be the case in vivo in somatic cell nuclei.b. In the earliest stage of condensation, certain regions of chro-matin fibers condense into thicker (30 nm) fibers and aggregateinside initiation foci of condensation or chromomeres. Initiallythese chromomeres are separated by dencondensed chromatinfibers; some chromatin fibers may also radiate out from thesechromomeres as loops. c. As the chromatin fibers, between thechromomeres and those radiating out from them, progressivelycondense and are incorporated into chromomeres, the latterbecome more numerous and voluminous. d. The chromomeresfinally join up to form a continuous chromosomal structure. Inearly stage of chromosomal condensation, some chromatin fibersloop out from the chromosome. e. As the condensation of chro-mosomes proceeds, all the chromatin loops retract, condenseand are incorporated into the mass of the chromosome. Theterminal stage of chromosomal condensation (i.e., prophase tometaphase transition) most likely involves a helical coiling of thechromatid axis (figure 6d), followed by a packing down the coils,so that in the fully condensed metaphase chromosome the helicalorganization is no more apparent.



of somatic cells. Our results suggest that at the firstlevel of chromatin condensation, there is a loose helicalcoiling of the 10 nm fibers, which by further packingdown the coils could produce thicker 30 nm fibers.

Amongst the numerous hypotheses advanced toexplain higher order condensation of chromatin intomitotic chromosomes, we may mention the earliermodels based on successive levels of helical foldings,proposed by Sedat and Manuelidis (1978). More re-cently, Belmont and his collaborators have proposed analternative model based on 100 nm chromonema fiberformed by compaction of 10 and 30 nm chromatinfibers; the chromonema fiber is supposed to undergoan irregular folding into the 200–300 nm prophasechromatid which coils to form the metaphase chromo-some (Belmont and Bruce, 1994; Belmont et al., 1999; Liet al., 1998). In the model proposed by Laemmli andhis collaborators, SARs (AT-rich scaffold-associated re-gions of DNA, also known as MARs or matrix-attachment regions) form the bases of chromatin loopsand are brought into juxtaposition by the scaffoldproteins, resulting in a lining up of SARs to form an ATqueue in the mitotic chromosomes. This AT queue issupposed to be helically folded, with a relatively tightfolding in the gene-poor Q bands, and a looser foldingin the gene-rich R bands (Saitoh and Laemmli, 1994;Hart and Laemmli, 1998).

Our present observations on the reorganization ofchromatin in Xenopus egg extracts lead us to proposethe following model for mitotic condensation of chro-mosomes. The condensation of chromatin in vitro inXenopus egg extracts is not a synchronous process,even if all the chromatin is in a decondensed state tostart with (figure 6a). In dividing somatic cell nuclei invivo, where some chromatin is always condensed un-der normal conditions, the chromosomal condensationis also likely to be asynchronus. The condensationstarts by aggregation of chromatin fibers at certaininitiation foci, which can be compared to chromomeresof classical litterature (Lima de Faria, 1975). In theearliest stages of condensation, these chromomeres areseparated by decondensed chromatin fibers; somechromatin fibers may also extend out as radial loops ofvariable lengths from these chromomeres (figure 6b). Asthese chromatin fibers condense progressively, theyproduce more chromomeres and/or become incorpo-rated into neighbouring chromomeres. In this process,the chromomeres become more numerous and/ormore voluminous ( i.e., they become thicker andlonger) until they finally join up to form the chromo-some (or chromatid) axis (figure 6c, d). At the first levelof compaction, the 10 nm chromatin fiber appears toundergo a loose helical coiling which by further com-paction would produce thicker (30 nm) fibers. As forthe compaction of 30 nm chromatin fibers into chro-momeres, a helical packing of these fibers may play

some role, but other processes such as super coiling,irregular folding and twisting could also intervene atthis level of condensation. Evidently this problemneeds further study. The condensation of 10 nm fibersinto thicker 30 nm fibers and the compaction of thelatter into chromomeres may be temporally relatedprocesses, i.e., the formation of 30 nm fiber in a par-ticular region of chromatin is immediately followed byincorporation of that region into a chromomere. As thecondensation of the chromosomes proceeds, the loopsprogressively retract, condense and are incorporatedinto the chromosomal mass. In our opinion, the loopsdo not exist as such in fully condensed metaphasechromosomes in vivo, and can be visualized only afterhypotonic or other harsh treatments. All along theprocess of chromosomal condensation, the chromatinfibers inside the chromomeres and later in the chromo-somes become more and more tightly packed, so thatthe path of individual 30 nm chromatin fibers can notbe discerned in the fully condensed metaphase chro-mosomes. The transition from prophase to metaphasestage most likely involves a helical coiling, as is sug-gested by several earlier studies (e.g., Ohnuki, 1968;Rattner and Lin, 1985; Boy de la Tour et al., 1988). Inthe final stage of chromosomal condensation, there is apacking down the coils, so that in the metaphasechromosomes the helical organization can not be dis-cerned (figure 6d, e). It must be emphasized here thatthe model presented here is based on condensation ofchromatin in vitro in high speed Xenopus egg extracts;further studies would be necessary to check its validityfor in vivo condensation of chromosomes.

In the past few years, the importance of severalnon-histone proteins in the chromosome dynamics hasbeen revealed. DNA topoisomerase II was the first ofthese proteins to be recognized as having an importantrole in the mitotic condensation of chromosomes (Earn-shaw and Heck, 1985; Wood and Earnshaw, 1990;Adachi et al., 1991; Hirano and Mitchison, 1993, 1994).Since then, several other proteins, now collectivelyknown as SMC proteins, as well as a few othersassociated with them, have been described and theirimportance in chromosomal condensation demon-strated (see Koshland and Strunnikov 1996, Hirano2000 for reviews). Condensation of chromatin in vitroin Xenopus egg extracts has been valuable for thesestudies. Submicroscopic localization of all these pro-teins in in vitro condensing chromosomoes, throughelectron microscopic immunocytochemistry, could pro-vide more precise informations on the respective rolesof these proteins and their interactions. The presentwork demonstrates the feasibility of such cytochemicalstudies as well as further high resolution electronmicroscopical analyses of in vitro condensing chromo-somes, which might lead to a better comprehension ofthe process of mitotic condensation of chromosomes.



Acknowledgments. This work was financially sup-ported by grants from the Centre National de la Re-cherche Scientifique and from the Université Pierre-et-Marie-Curie. The author expresses his sincere gratitudeto Prof. J. C. Boucaut and Dr. D. Shi for providing himwith all the facilities, for going through the manuscriptand for their suggestions, and to Dr. E. Puvion and Dr.F. Puvion-Dutilleul for a critical reading of the manu-script and for important suggestions. Thanks are alsodue to all my colleagues in this laboratory for fruitfuldiscussions. Electron microscopy was carried out at theCentre Interuniversitaire de Microscopie Electronique,Jussieu, Paris; the help of D. Touret, F. Devienne, V.Hosansky and J. Desrosiers for photographic illustra-tions and to J. Parent for the maintainance of theelectron microscopes is gratefully acknowledged.

REFERENCES

Adachi, Y., Luke, M., Laemmli, U.K., 1991. Chromosome assembly invitro: topoisomerase II is required for condensation. Cell 64, 137–148.

Adolph, K.W., 1980. Organization of chromosomes in mitotic HeLacells. Exp. Cell Res. 125, 95–103.

Almouzni, G., Wolffe, A.P., 1993. Nuclear assembly, structure andfunction : the use of Xenopus in vitro system. Exp. Cell Res. 205,1–15.

Bednar, J., Horowitz, R.A., Grigoryev, S.A., Carruthers, L.M., Hansen,J.C., Koster, A.J., Woodcock, C.L., 1998. Nucleosomes, linker DNA,and linker histone form a unique structural motif that directs thehigher-order folding and compaction of chromatin. Proc. Natl. Acad.Sci. USA 95, 14173–14178.

Belmont, A.S., Bruce, K., 1994. Visualization of G1 chromosomes - afolded, twisted, supercoiled chromonema model of interphase chro-matid structure. J. Cell Biol. 127, 287–302.

Belmont, A.S., Dietzel, S., Nye, A.C., Strukov, Y.G., Tumbar, T., 1999.Large-scale chromatin structure and function. Curr Opin Cell Biol11, 307–311.

Belmont, A.S., Sedat, J.W., Agard, D.A., 1987. A three dimensionalapproach to mitotic chromosome structure : evidence for a complexheirarchical organization. J. Cell Biol. 195, 77–92.

Boy de la Tour, E., Laemmli, U.K., 1988. The metaphase scaffold ishelically folded : sister chromatids have predominantly oppositehelical handedness. Cell 55, 937–944.

Cubizolles, F., Legagneux, V., Le Guellec, R., Chartrain, I., Uzbekov, R.,Ford, C., Le Guellec, K., 1998. pEg7, a new Xenopus protein requiredfor mitotic chromosome condensation in egg extracts. J. Cell Biol.143, 1437–1446.

Daban, J.R., Bermudez, A., 1998. Interdigitated solenoid model forcompact chromatin fibers. Biochemistry 37, 4299–4304.

Desai, A., Deacon, H.W., Walczak, C.E., Mitchison, T.J., 1997. A methodthat allows the assembly of kinetochore components on chromo-some condensed in clarified Xenopus egg extracts. Proc. Natl. Acad.Sci. USA 94, 12378–12383.

Earnshaw, W.C., Heck, M.M.S., 1985. Localization of topoisomerase II inmitotic chromosomes. J. Cell Biol. 100, 1716–1725.

Finch, J.T., Klug, A., 1976. Solenoid model for superstructure inchromatin. Proc. Natl. Acad. Sci. USA 73, 1897–1901.

Giannasa, P.J., Horowitz, R.A., Woodcock, C.L., 1993. Transition be-tween in situ and isolated chromatin. J. Cell Sci. 105, 551–561.

Grunstein, M., 1990. Histone function in transcription. Ann. Rev. CellBiol. 6, 643–678.

Guacci, V., Hogan, E., Koshland, D., 1994. Chromosome condensationand sister chromatid pairing in budding yeast. J. Cell Biol. 125,517–530.

Hart, C.M., Laemmli, U.K., 1998. Facilitation of chromatin dynamics bySARs. Curr. Opin. Gen. Dev. 8, 519–525.

Hirano, T., 2000. Chromosome cohesion, condensation, and separation.Annu. Rev. Biochem. 69, 115–144.

Hirano, T., Kobayashi, R., Hirano, M., 1997. Condensins, Chromosomecondensation protein complexes containing XCAP-C, XCAP-E and aXenopus homolog of the Drosophila Barren protein. Cell 89, 511–521.

Hirano, T., Mitchison, T.J., 1993. Topoisomerase II does not play ascaffolding role in the oganization of mitotic chromosomes as-sembled in Xenopus egg extracts. J. Cell Biol. 120, 601–612.

Hirano, T., Mitchison, T.J., 1994. A heterodimeric coiled-coil proteinrequired for mitotic chromosome condensation in vitro. Cell 79,449–458.

Hock, R., Moorman, A., Fischer, D., Scheer, U., 1993. Absence of somatichistone H1 in oocytes and preblastula embryos of Xenopus laevis.Dev. Biol. 158, 510–522.

Horowitz, R.A., Giannasca, P.J., Woodcock, C.L., 1990. Ultrastructuralpreservation of nuclei and chromatin : improvement with lowtemperature methods. J. Micros. 157, 205–224.

Koshland, D., Strunnikov, A., 1996. Mitotic chromosome condensation.Ann. Rev. Cell Dev. Biol. 12, 305–333.

Laemmli, U.K., Cheng, S.M., Adolph, K.W., Paulson, J.R., Brown, J.A.,Baumwach, W.R., 1978. Metaphase chromosome structure : the roleof non-histone proteins. Cold Spring Harb. Symp. Quant. Biol. 42,351–360.

Li, G., Sudlow, G., Belmont, A.S., 1998. Interphase cell cycle dynamics ofa late replicating, heterochromatic homogenously staining region :Precise choreography of condensation/decondensation and nuclearpositioning. J. Cell Biol. 140, 975–989.

Lima de Faria, A., 1975. The relation between chromomeres, replicons,operons, transcription units, genes, viruses and palindromes.Hereditas 81, 249–284.

Lokha, M.J., Masui, Y., 1983. Formation in vitro of sperm pronuclei andmitotic chromosomes by amphibian ooplasmic contents. Science 220,719–721.

Losada, A., Hirano, M., Hirano, T., 1998. Identification of Xenopus SMCprotein complexes required for sister chromatid cohesion. GenesDev. 12, 1986–1997.

Losada, A., Yokochi, T., Kobayashi, R., Hirano, T., 2000. Identificationand characterization of SA/Scc3p subunits in the Xenopus andhuman cohesin complexes. J. Cell Biol. 150, 405–416.

Murray, A.W., 1991. Cell cycle extracts. In: Kay, B.K., Peng, H.B. (Eds.),Xenopus laevis : Practical uses in Cell and Molecular Biology. Aca-demic Press, San Diego, CA, pp. 581–605.

Newmeyer, D.D., Wilson, K.L., 1991. Egg extracts for nuclear importand nuclear assembly reactions. In: Kay, B.K., Peng, H.B. (Eds.),Xenopus laevis : Practical uses in Cell and Molecular Biology. Aca-demic Press, San Diego, CA, pp. 607–634.

Ohnuki, Y., 1968. Structure of chromosomes: 1. Morphological studiesof the spiral structure of human somatic chromosomes. Chromo-soma (Berl) 25, 402–428.

Ohsumi, K., Katagiri, C., 1991. Occurrence of H1 subtypes specific topronuclei and cleavage stage cell nuclei of anuran amphibians. Dev.Biol. 147, 110–120.

Ohsumi, K., Katagiri, C., Kishimoto, T., 1993. Mitotic chromosomecondensation without histone H1. Science 262, 2033–2035.

Paulson, J.R., Laemmli, U.K., 1977. The structure of histone depletedchromosomes. Cell 13, 817–828.

Philpott, A., Leno, G.H., Laskey, R.A., 1991. Sperm decondensation inXenopus egg cytoplasm is mediated by nucleoplasmin. Cell 65,569–578.

Philpott, A., Leno, G.H., 1992. Nucleoplasmin remodels sperm chroma-tin in Xenopus egg extracts. Cell 69, 759–767.

Pyne, C.K., Loones, M.T., Lacroix, J.C., 1989. Correlation betweenisolated lampbrush chromosomes and nuclear structures of Pleurode-les waltl oocytes : an electron microscopic study. Chromosoma (Berl)98, 181–193.

Rattner, J.B., Lin, C.C., 1985. Radial loops and helical coils coexist inmetaphase chromosomes. Cell 42, 291–296.



Saitoh, Y., Laemmli, U.K., 1994. Metaphase chromosomùe structure :Bands arise from a differential folding path of the highly AT-richscaffold. Cell 76, 609–622.

Sawin, K.E., Mitchison, T.J., 1991. Mitotic spindle assembly by twodifferent pathways in vitro. J. Cell Biol. 112, 925–940.

Sedat, J.W., Manuelidis, L., 1978. A direct approach to the structure ofeukaryotic chromosomes. Cold Spring Harb. Symp. Quant. Biol. 42,331–350.

Shamu, C.E., Murray, A.W., 1992. Sister chromatid separation in frogegg extracts requires DNA topoisomerase II activity duringanaphase. J. Cell Biol. 117, 921–934.

Sheehan, M.A., Mills, A.D., Sleeman, A.M., Laskey, R.A., Blow, J.J., 1988.Steps in the assembly of replication-competent nuclei in a cell-freesystem from Xenopus eggs. J. Cell Biol. 106, 1–12.

Smith, R.C., Dworkin-Rastl, E., Dworkin, M.B., 1988. Expression of ahistone H1-like protein is restricted to early Xenopus development.Genes Dev. 2, 1284–1295.

Thoma, F., Koller, T.H., 1977. Influence of histone H1 on chromatinstructure. Cell 12, 101–107.

Thoma, F., Koller, T.H., Klug, A., 1979. Involvement of histone H1 in theorganization of the nucleosome and of the salt-dependent super-structures of chromatin. J. Cell Biol. 83, 403–427.

van Holde, K., Zlatanova, J., 1995. Chromatin higher order structure:chasing a mirage?. J. Biol. Chem. 270, 8373–8376.

Wolffe, A.P., 1989. Dominant and specific repression of Xenopus oocyte5S RNA genes and satellite I DNA by histone H1. EMBO J. 8,527–537.

Wolffe, A.P., 1981. Developmental regulation of chromatin structure andfunction. Trends Cell Biol. 1, 61–66.

Wood, E.R., Earnshaw, W.C., 1990. Mitotic chromatin condensation invitro using somatic cell extracts and nuclei with variable levels ofendogenous topoisomerase II. J. Cell Biol. 111, 2839–2850.

Woodcock, C.L., Horowitz, R.A., 1995. Chromatin organization re-viewed. Trends Cell Biol. 5, 272–277.



Documents

Reorganization of chromatin in Xenopus egg extracts: electron microscopic studies