Text práce (2.470Mb)

(1)

First Faculty of Medicine, Charles University in Prague

DNA REPLICATION AND CHROMATIN: FROM 3D TO FUNCTION

PhD Thesis

Anna Ligasová

Prague 2009

(2)

2

ACKNOWLEDGEMENT

I thank my supervisor RNDr. Karel Koberna, CSc. for his support, experience and opportunity for scientific progress.

I would also like to thank all of my former colleagues from the Institute of Cellular Biology and Pathology and all of my present colleagues from the Laboratory of Cell Biology for their assistance.

Last but not least, I thank my family and friends for their support, patience and tolerance.

(3)

3

ABBREVIATIONS

RS Replication sites

RF Replication foci

dT 2-deoxythymidine

DMEM Dulbecco‘s modified Eagle‘s medium

BrdU 5-bromo-5-deoxyuridine

Biotin-dUTP, biotin-16-dUTP biotin-16-deoxyuridine triphosphate

Biotin-dU biotin-16-2`deoxy-uridine

LM light microscopy

EM electron microscopy

ChIP chromatin immunoprecipitation

UBF upstream binding factor

IC initiation complex

Pre-RC pre-replicative complex

ORC origin recognition complex

MCM proteins minichromosome maintenance proteins

CDK cyclin-dependent kinase

DDK Dbf4-dependent kinase

RPA replication protein A

RFC replication factor C

PCNA proliferating cell nuclear antigen

NORs nucleolar organizing regions

DFC dense fibrillar component

FC fibrillar center

GC granular component

CB Cajal body

PML promyelocytic leukemia body

snRNP small nuclear ribonucleoprotein particle

snoRNPs small nucleolar ribonucleoprotein particles

Pol I RNA polymerase I

Pre-rRNA pre-ribosomal RNA

rDNA ribosomal DNA

(7)

7

CHAPTER 1 OUTLINE OF THE THESIS

The PhD. project was focused on the organization of replicated chromatin in human cells and the in situ organization of replication complexes ensuring proper DNA replication.

In this respect, the most relevant results of my project are contained in Koberna et al. (2005), Ligasová et al. (2009) and Cvačková et al. (2008). In addition, I am a co-author of a further study dealing with chromosome positioning (Kalmárová et al., 2007).

Here is a brief overview of the content of my thesis:

Chapter 2 presents the basic information on the cell nucleus, chromatin, DNA replication and nuclear architecture. The main aims of the thesis are defined in Chapter 3.

Chapter 4 briefly describes the material and methods used. The methods are characterized in further detail in the next three chapters (5, 6, 7), corresponding to the three published articles.

Chapter 5 provides a study dealing with the organization of the replication sites in the various stages of the S phase at the electron microscopy level. This study was published in the Journal of Cellular Biochemistry (Koberna et al., 2005). Some aspects of the organization of the replisome complexes during and after replication are described in Chapter 6. This work will be published in the Journal of Structural Biology (Ligasová et al., 2009). The last study (Chapter 7) is devoted to the localization of a putative DNA helicase, Pontin in the cell nucleus. In this study, I performed electron microscopy localization of this protein. The study was published in Chromosoma (Cvačková et al., 2008). The rest of the thesis includes a general discussion (Chapter 8), conclusions (Chapter 9), summary (Chapter 10) and the list of references (Chapter 11).

(8)

8

CHAPTER 2 GENERAL INTRODUCTION

2.1 CELL NUCLEUS

The cell nucleus, a highly dynamic organelle, was discovered and originally described by Franz Bauer in 1802 and later popularized by Robert Brown (Dundr and Misteli, 2001).

For a long time, it has attracted the attention of scientists as it contains most of a cell‘s genetic information, which occupies a large portion of the nucleus in the form of linear molecules of DNA consisting of about 3.2 x 10⁹ base pairs. Such an amount of DNA has to be extensively compacted in order for it to intrude into the nucleus with a volume of ~ 1000 µm³(Schneider and Grosschedl, 2008). The packing of the DNA is achieved by its association with various components. The dominant role in the packing is played by basic proteins designated as histones. The complex of DNA with these components is commonly called chromatin (Trinkle-Mulcahy and Lamond, 2008). Besides chromatin, the nucleus contains additional compartments. The most common are: nucleoli, Cajal bodies (CB), promyelocytic leukemia (PML) bodies, speckles and paraspeckles (reviewed in e.g. Spector, 2001; Trinkle-Mulcahy and Lamond, 2008).

2.2 CHROMATIN

Two basic morphological forms of the chromatin are commonly distinguished in the cell nucleus in interphase. The most highly condensed regions are called heterochromatin; the less condensed parts of chromatin euchromatin (Schneider and Grosschedl, 2008).

Heterochromatin has a low gene density but can contain active genes. On the other hand, euchromatin is presumed to have a high gene density but can contain genes which are not expressed (Gilbert et al., 2004). In this respect, it is supposed that the gene expression is regulated by the covalent modification of histones or DNA rather than by their localization in heterochromatin or euchromatin (Spector, 2004).

The genome is divided into particular chromosomes. In the interphase, chromosomes occupy defined so-called chromosome territories instead of being extensively intermingled (e.g. Cremer et al., 1993; Visser et al., 2000; Cremer and Cremer, 2001, 2006). These territories are spatially arranged according to their gene density, replication timing, the size of the chromosome territory and the activity of their genes (e.g. Zink et al., 1998; Manders et al., 1999; Mahy et al., 2002; Walter et al., 2003; Gilbert and Ramsahoye,2005).

(9)

9

As mentioned above, DNA in the nucleus is extensively compacted. In the interphase, DNA is compacted at least through two levels of condensation. The first level is nucleosomal fiber, consisting of nucleosomes, the basic repeating units. In nucleosomes, 145–147 bases of DNA are wrapped around the histone octamer. Such nucleosomes are linked together with the linker DNA and form a nucleosomal fiber (Luger et al., 1997; Kornberg and Lorch, 1999).

This fiber can be visualized by means of electron microscopy (EM) as a 10 nm fiber of ―beads on string‖ (Kornberg and Lorch, 1999; Alberts et al., 2002; Muthurajan et al., 2003). It seems that the nucleosomal fiber is further packed into a 30 nm fiber, which is stabilized by a linker histone H1 (Alberts et al., 2002). Although additional levels of chromatin organization have been proposed, the evidence for them is very scarce and contradictory. Three basic models of interphase chromatin arrangements above the level of 30 nm fiber have been suggested: a model of a chromonema fiber (Belmont and Bruce, 1994; Belmont et al., 1999), a model of giant loops (Yokota et al., 1995; Münkel et al., 1999) and a model of condensed radial loops (Paulson and Laemmli, 1977; Marsden and Laemmli, 1979; Adolph, 1980). Belmont and Bruce (1994) observed chromatin fibers in the early G1 phase and the late G1/early S phase and labeled them chromonema fibers. In this model, a chromosome fiber of a diameter of ca 100 nm folds into prophase chromatids of a diameter of 200–300 nm, which in turn coil into a metaphase chromosome structure (Belmont et al., 1999). The giant-loop model was suggested by a statistical analysis of the mean separation between two chromosomal sites as a function of genomic distance (Yokota et al., 1995; Münkel et al., 1999). The model of radial loops represents the extrapolation of the model suggested for mitotic chromosomes (Paulson and Laemmli, 1977; Marsden and Laemmli, 1979; Adolph, 1980). A completely different view of the organization of interphase chromatin has been provided by a model based on the EM observation of the lattices, comprising exclusively 10 and 30 nm fibers connecting inter- and intra-chromosomal space to form an almost contiguous nucleoplasmic space (Bazett-Jones and Hendzel, 1999; Dehghani et al., 2005).

2.3 DNA REPLICATION

Eukaryotic chromosomal DNA is divided into hundreds to thousands of independent subunits of replication, termed replicons. Replicon is a part of DNA replicated from one replication origin. Only a small portion of replicons is active at any time of the S phase. It is supposed that about 10–15 % of replicons are active at any time of the S phase (Jackson, 1995; Jackson and Pombo, 1998). As the human genome consists of about 40,000 replicons (Singer et al., 1996), around 4,000 replicons should be active at any time of the S phase. The

(10)

10

replication of replicons begins from multiple origins, proceeds bidirectionally through the agency of two replication forks and terminates when the forks of neighboring replicons converge (Heintz, 1996; Blow and Dutta, 2005). Unlike yeasts, which use origins with a conserved DNA sequence, the mammalian genome does not contain specific sequences for the initiation of replication. It is supposed that the initiation of replication origins occurs in

―initiation zones‖. Such zones are several kilobases long and contain many potential origins.

Only one of these origins per zone is used in the S phase. In addition, different origins can be used in different cell cycles. Particular origins differ in their efficiency, and some of them are used more frequently than others (Costa and Blow, 2007).

2.3.1 The course of DNA replication

The initiation of DNA replication proceeds in two time-separated steps during the cell cycle: the origin licensing and the formation of the replication fork (Moldovan et al., 2007).

Already at the M/G1 transition and during the G1 phase, the assembly of a multi- protein complex, a pre-replicative complex (pre-RC), is initiated at the replication origins in a highly regulated and ordered fashion (Bell and Dutta, 2002; Diffley, 2004). First of all, a heteromeric complex of six proteins, an origin recognition complex (ORC), assembles on the sites of DNA where the replication is to start (reviewed in Bell and Dutta, 2002; DePamphilis, 2003; Chesnokov, 2007; Sasaki and Gilbert, 2007). Further, the Cdc6 protein directly binds to the ORC complex. Further, the Cdt1 protein binds to this protein complex. The whole complex, consisting of ORC, Cdc6 and Cdt1 proteins, enables the binding of the minichromosome maintenance protein complex (MCM2-7 proteins) to DNA (e.g. Lei and Tye, 2001; DePamphilis, 2003; Stillman, 2005; Blow and Dutta, 2005). It is supposed that MCM not only has the direct role in the ―licensing‖ of the origins for replication but also forms the catalytic core of the replicative helicase, acting during the elongation phase of DNA duplication (Takahashi et al., 2005).

In the next step, the initiation complex (IC) forms, which requires the activity of cyclin-dependent kinase (CDK) and Dbf4-dependent kinase/Cdc7 (DDK) as well as the presence of the Cdc45 protein, a GINS complex (a complex of four proteins: Sld5 and Psf1-3) and the MCM10 protein (Diffley and Labib, 2002; Bauerschmidt et al., 2007). The activity of the above-mentioned kinases together with the formation of the complex of Cdc45, GINS and MCM2-7 proteins is likely to trigger the helicase activity of MCM proteins (Aparicio et al., 2006). The subsequent factors participating in the actual DNA replication form a complex

(11)

11

called replisome (Baker and Bell, 1998; Waga and Stillman, 1998; Johnson and O‘Donnel, 2005). After the assembly of the replisomes at the replication forks, DNA synthesis starts in a semi-discontinuous and bidirectional manner. The leading strand is continuously replicated because of the 5‘-3‘- polymerase activity of DNA polymerases. The lagging strand is synthesized in a discontinuous fashion via Okazaki fragments.

The essential factors ensuring the DNA synthesis are: DNA helicase, the replication protein A (RPA), replication factor C (RFC), proliferating cell nuclear antigen (PCNA), complex of DNA polymerase α/primase, DNA polymerases δ and ε, ribonuclease H1 and FEN1, DNA ligase I and topoisomerase I and II.

At first, the DNA is unwound by the helicase complex and the single-stranded DNA is stabilized by the interaction with the RPA (Wobbe et al., 1987; Fairman and Stillman, 1988).

Topoisomerase play an important role during this unwinding step as they unlink the parental strands to relax the supercoiled DNA (Wang, 1985). DNA polymerase α/primase ensures the synthesis of the initial, short RNA primer (Brush and Kelly, 1996; Moldovan et al., 2007) and extends the primer by about 20 bases of DNA. Next, the RFC is bound to the primer and catalyses the loading of the PCNA that encircles the DNA while preventing the dissociation of the polymerase δ and ε from DNA (Brush and Kelly, 1996; Johnson and O‘Donnell, 2005;

Moldovan et al., 2007). This leads to the replacement of DNA polymerase α/primase by polymerases δ and ε with proofreading activity enabling the high fidelity of the replication (Johnson and O‘Donnell, 2005; Moldovan et al., 2007). In the case of the lagging strand, when the replicative polymerase reaches the end of a previous Okazaki fragment, it partially displaces this fragment by an ongoing DNA synthesis. Primers are removed by means of the ribonuclease H1 and FEN1. The resulting gap is filled by DNA polymerase, probably δ or ε, and the nick is sealed by DNA ligase I (Brush and Kelly, 1996; Moldovan et al., 2007).

2.3.2 Spatio-temporal organization of DNA replication

As mentioned above, the basic unit of DNA replication is the replicon. The size of the individual replicons usually varies from 30 to 450 kbp, but replicons below 10 kbp and above 1 Mbp have also been described (Edenberg and Huberman, 1975; Yurov and Liapunova, 1977; Hand, 1978; Hyrien and Mechali, 1993; Jackson and Pombo, 1998; Berezney et al., 2000). Based on the studies of stretched DNA fibers, it is assumed that small clusters of adjacent replicons are activated synchronously (Edenberg and Huberman, 1975; Hand 1978).

The number of replicons in one replicon cluster is usually below ten, and the replicon cluster

(12)

12

is usually shorter than 1 Mb (Nakamura et al., 1986; Jackson and Pombo, 1998; Ma et al., 1998). It is assumed that the replication of one such replicon cluster takes about 45 to 60 minutes to complete (Edenberg and Huberman, 1975; Nakamura et al., 1986; Manders et al., 1992; Ma et al., 1998).

In situ, active replicon clusters were located to discrete foci, also called replication foci or sites, structures which can be observed after the immunocytochemical detection of DNA synthetic activity (Nakamura et al., 1986; Nakayasu and Berezney, 1989; Fox et al., 1991; O‘Keefe et al., 1992; Ma et al., 1998; Dimitrova and Gilbert, 1999; Leonhardt et al., 2000). The size, number and position of replication foci change during the S-phase progression. In this respect, several distinct replication patterns have been described, with some groups resolving three different replication patterns (e.g. Nakayasu and Berezney, 1989;

Manders et al., 1992; Jackson, 1995) and other groups distinguishing between five distinct patterns (e.g. van Dierendonck et al., 1989; O‘Keefe et al., 1992; Dimitrova and Gilbert, 1999). In general, in the early S phase, many small fluorescent foci scattered throughout the nucleoplasm, except for nucleoli, are observed. In the course of the S phase, less numerous and larger foci appear in the perinuclear and perinucleolar heterochromatin regions. In the late S phase, heterochromatin replicates only in a few, usually bright, large foci. Cell cycle studies have also revealed that replication foci persist in the cell‘s nucleus throughout the cell cycle and also in the next cell generations (Sparvoli et al., 1994; Ferreira et al., 1997; Ma et al., 1998; Jackson and Pombo, 1998; Zink et al., 1998; Dimitrova and Gilbert 1999). The number and size of the replication foci have been measured with the help of light microscopy (LM) techniques such as fluorescent microscopy (Nakayasu and Berezney, 1989; Tomilin et al., 1995) or laser scanning confocal microscopy (Ma et al, 1998). A great variability in the size (0.1–0.5 µm) and number (120–1500) of foci has been described for the early-S-phase cells (Nakamura et al., 1986; Mills et al., 1989; Nakayasu and Berezney, 1989; Ma et al., 1998;

Tomilin et al., 1995; Jackson, 1995).

2.4 INTERCHROMATIN SPACE

A number of different structural compartments have been discerned in the interchromatin space. The best known include: the nucleolus, Cajal bodies, PML bodies, nuclear speckles and paraspeckles.

The most prominent nuclear compartment is the nucleolus formed at the end of mitosis around the chromosomal loci, also called the nucleolar organizing regions (NORs), containing

(13)

13

the tandemly-repeated rDNA gene clusters (Spector, 2001; Boisvert et al., 2007; Trinkle- Mulcahy and Lamond, 2008). In humans, approximately 400 copies of 43 kb repeat units are distributed along all the acrocentric chromosomes (chromosomes 13, 14, 15, 21 and 22) to form NORs (Boisvert et al., 2007). The nucleolus is a place of rRNA synthesis, rRNA processing and the assembly of ribosomal subunits (Hadjiolov, 1985; Spector, 2001). The size and the number of nucleoli depend on the cell‘s activity, proliferation and differentiation.

Three distinct components can be identified in the nucleus by electron microscopy: fibrillar centers (FCs), dense fibrillar components (DFCs), surrounding the FCs and granular components (GCs), in which FCs and DFCs are embedded. Besides these components, more or less condensed chromatin can enter into the nucleolus from the surrounding perinucleolar chromatin (Hadjiolov, 1985; Hernandez-Verdun, 2006).

Cajal bodies are dynamic nuclear bodies present in the nuclei of most cells which also associate with specific genes (U snRNA genes, Spector, 2001; Trinkle-Mulcahy and Lamond, 2008). CBs can move, split and again rejoin in the nucleoplasm. Similarly, like in the case of nucleolus, the size and the number of CBs depend on the cell‘s activity and cycle (Carmo- Fonseca et al., 1993; Carvalho et al., 1999). The function of CBs seems to be playing a role in small nuclear ribonucleoprotein particle‘s (snRNP) biogenesis as well as in the trafficking of snRNPs and small nucleolar ribonucleoprotein particles (snoRNPs) (Spector, 2001).

Promyelocytic leukemia bodies are discrete nuclear bodies present in most mammalian nuclei, which associate with gene dense, transcriptionally active chromatin (Bernardi and Pandolfi, 2007). It was assumed that PML bodies do not directly regulate transcription of genes in this region (Wang et al., 2004), but Block et al. (2006) have recently shown a promoter-specific regulation of transcription within the PML body region.

Nuclear speckles are usually found next to sites of active transcription. They contain pre-messenger RNA splicing factors but they are not the place of the splicing. They are also rich in other transcription and splicing-related proteins (Lamond and Spector, 2003). They are believed to serve as the sites of the storage and recycling of the splicing factors (Fox et al., 2002).

Paraspeckles are usually found adjacent to speckles and contain at least three RNA- binding proteins: PSP1, PSP2 and p54/nrb. All these proteins interact with the nucleolus in a transcription-dependent manner (Fox et al., 2002). Paraspeckles disperse at the beginning of mitosis and re-form in the G1 phase (Trinkle-Mulcahy and Lamond, 2008).

(14)

14

CHAPTER 3 SPECIFIC AIMS OF THE THESIS

The specific aims of the thesis are:

1) the description of the spatio-temporal organization of the active replicons in HeLa cells;

2) the mutual organization of the sister replisomes during the S phase; and 3) the nuclear localization of the putative DNA helicase, Pontin.

(15)

15

CHAPTER 4 MATERIALS AND METHODS

This chapter surveys the material and methods used in the thesis. A more detailed description can be found in Chapters 5, 6 and 7.

4.1 CELL CULTURE AND SYNCHRONIZATION

Human HeLa cells, derived from adenocarcinoma cervix cells, and HepG2, human hepatocyte carcinoma cells, were cultured and synchronized as described in Chapters 5, 6 and 7.

4.2 LABELING OF NEWLY SYNTHESISED DNA

5-bromo-5‘-deoxyuridine (BrdU) and biotin-16-2'-deoxy-uridine-5'-triphosphate (biotin-dUTP) were used as the markers of newly synthesized DNA. BrdU was added directly to the culture medium for a specific time and the cells were subsequently processed for LM or EM (see Chapters 5, 6). Biotin-dUTP was delivered into the cell nuclei by means of hypotonic shift and the cells were further processed for LM or EM microscopy (see Chapters 5, 6).

4.3 TRANSCRIPTION LABELING

5-fluorouridine was directly added to the culture medium for 10 minutes, after which the cells were processed for LM (see Chapter 7).

4.4 siRNA

The cells were transfected with two siRNA duplexes as described in Watkins et al.

(2004) and further processed as described in Elbashir et al. (2002). See Chapter 7 for details.

4.5 ANTIBODIES

The anti-bromodeoxyuridine antibody and anti-biotin antibody were used as the primary antibodies in the studies concerning the organization of replication sites and replisomes. In the study dealing with the localization of the Pontin in the nucleolus, the primary antibodies against Pontin, fibrillarin, Upstream Binding Factor (UBF), RNA polymerase I, fibrillarin 17C12, vimentin V1-01, c-myc, bromodeoxyuridine and SART3 were used. For LM, we used secondary antibodies conjugated with FITC or Cy3. For EM

(16)

16

studies, antibodies conjugated with ultra-small or 6 nm gold particles were used. For more details, see Chapters 5, 6 and 7.

4.6 ELECTRON MICROSCOPY

Cells grown on coverslips or embedded in 10% gelatin were dehydrated and embedded in Epon or Lowicryl. Ultra-thin (70 nm thick) sections were cut on a Leica UltraCut S microtome with a diamond knife and stained with uranyl acetate. For EM tomography, 200 nm thick sections were used. Concerning the Pontin localization, the cells were embedded in gelatin and cryo-preserved according to Raska et al. (1995). Thin cryosections were incubated with primary and secondary antibodies and stained with uranyl acetate. For a detailed description, see Chapters 5, 6 and 7.

4.7 WESTERN BLOT ANALYSIS AND CHROMATIN IMMUNOPRECIPITATION Nucleoplasmic and nucleolar fractions from whole-cell lysate were prepared according to Andersen et al., 2002. Whole-cell lysates were separated on SDS/PAGE and transferred to a nitrocellulose membrane. The primary anti-pontin antibodies were detected by means of a secondary anti-mouse antibody conjugated with horseradish peroxidase (Chapter 7).

For chromatin immunoprecipitation (ChIP), the cells were fixed with formaldehyde and scraped from the culture dish and the nuclei were pelleted by centrifugation and disrupted by sonication. The DNA fragments were incubated with primary anti-pontin or anti-myc antibodies and precipitated with protein A-sepharose (see the description in Chapter 7).

4.8 MICROSCOPES USED

The coverslips mounted in Mowiol were viewed using the Olympus AX 70 Provis microscope equipped with the Photometrics CCD camera, Leica TCS NT confocal microscope and Zeiss LSM 5 DUO confocal microscope. The ultrathin sections were viewed by Jeol 1200 EX, Zeiss 900 electron microscope equipped with a MegaView II camera, or Morgagni 268 electron microscope. The EM tomography was performed on a Tecnai G2 Sphera tomography microscope equipped with a Gatan Ultrascan 894 US1000 camera.

4.9 STATISTICAL EVALUATION

Stereological analysis was used to estimate the size of the replication domains and is described in Chapter 5. For the evaluation of the tomographic pictures, IMOD and Amira software were used (Chapter 6), whereas SigmaPlot software was used for the statistical

(17)

17

evaluation of the distances between the doublets of the labeled domains, SigmaPlot software was used (Chapter 6).

(18)

18

CHAPTER 5 ELECTRON MICROSCOPY OF DNA REPLICATION IN 3-D:

EVIDENCE FOR SIMILAR-SIZED REPLICATION FOCI THROUGHOUT S-PHASE

5.1 INTRODUCTION

Considerable progress has been made in studying replication sites (RS) or foci in mammalian cells using ﬂuorescence microscopy and computer imaging approaches.

Characteristic patterns of RS are observed in early-, mid-, and late-S-phase that serve as indicators for each of these S-phase periods (Nakamura et al., 1986; Nakayasu and Berezney, 1989; van Dierendonck et al., 1989; Mazzotti et al., 1990; Fox et al., 1991; Kill et al., 1991;

Manders et al., 1992; Neri et al., 1992; O‘Keefe et al., 1992; Sparvoli et al., 1994; Ferreira et al., 1997; Somanathan et al., 2001; Dimitrova and Berezney, 2002). The changes in the spatial and temporal organization of these RS are believed to reﬂect the choreography for the DNA replication program that proceeds from copying transcriptionally active euchromatic DNA in early-S to relatively inactive heterochromatic DNA later in S-phase (Ma et al., 1998).

These studies have led to models in which the individual RS of early-S-phase are proposed to be composed of multiple replicons (chromatin loops) with an average DNA content of 1Mbp (Nakamura et al., 1986; Jackson and Pombo, 1998; Ma et al., 1998; Berezney et al., 2000;

Berezney, 2002). The persistence of these labeled sites throughout the cell cycle and into subsequent cell generations has led to the further view that the early-S RS are fundamental higher order domains of chromatin organization and function (Jackson and Pombo, 1998; Ma et al., 1998; Zink et al., 1998; Berezney, 2002). In contrast to the RS of early-S-phase, our understanding of the organization of chromatin in RS of mid- and late-S-phases is much more limited. A major difﬁculty is that many of these RS are much larger in size as they correspond to heterochromatin regions replicated in these later periods of S-phase. While it is commonly assumed that these larger foci are composed of multiple units or chromatin domains of replication of similar size to the RS of early-S-phase, data supporting this possibility is limited (Nakayasu and Berezney, 1989; Raska et al., 1989, 1991; Leonhardt et al., 2000).

Electron microscopic analysis provides an enormous increase in resolution to overcome imaging limitations of the light microscope and to enable a more accurate and detailed understanding of replication site organization in the cell nucleus. Surprisingly, however, high resolution electron microscopic studies have to date not contributed signiﬁcantly (with the exception of the earlier observations of Raska et al., 1989, 1991) to our

(19)

19

understanding of the numerous replication foci detected in early-S-phase and the much larger but more limited number of foci characteristic of late-S-phase (Mazzotti et al., 1990; Rizzoli et al., 1992; Hozak et al., 1993, 1994; Jaunin et al., 2000). Since a major difficulty with this approach is the relatively weak labeling patterns typically obtained following short pulses with bromodeoxyuridine, we have taken steps in this study to improve the signal intensity for thin sectioning electron microscopic analysis. Intense labeling for DNA replication is observed that is concentrated at discrete sites throughout S-phase. The typically larger RS or foci observed by fluorescence microscopy in mid- and late-S-phase are found to be composed of smaller discrete replication foci virtually identical in size to the RS of early-S-phase. Pulse- chase studies reveal that the smaller foci of early-S are maintained throughout the S-phase and into the subsequent cell generation. Moreover, the relative number of smaller RS present at a given time does not drastically change during S-phase progression. 3-D reconstruction of serial sections demonstrates a higher order arrangement of the early-S-phase replication foci into network-like arrays and confirms that the small replication foci of late-S-phase are tightly compacted into the characteristic larger RS or foci.

5.2 MATERIALS AND METHODS 5.2.1 Cell Culture and Synchronization

HeLa cells were grown either in flasks or on circular coverslips in Petri dishes and cultured in Dulbecco‘s modified Eagle‘s medium (DMEM) supplemented with 10% fetal calf serum (Sigma-Aldrich, Steinheim, Germany), 1% glutamine, 1% penicillin, 1% streptomycin, and 0.85 g/L NaHCO3 at 37°C in a humidified atmosphere containing 5% CO2. Synchronization at the G1/S border was achieved by a double 2-deoxythymidine block (dT;

Sigma, St. Louis, MO). The cells were incubated in DMEM containing 3 mM dT for 16 h, then in a fresh medium for 12 h and again for16 h in a medium with 3 mM dT. After 100 min incubation in a normal medium, more than 90% of the cell population started DNA synthesis as judged by immunocytochemistry. Nine hours later cells exited S-phase as more than 95%

of them did not incorporate the 5-bromo-5-deoxyuridine (BrdU).

5.2.2 Labeling of Newly Synthesized DNA

BrdU (Sigma Chemicals Co.) or biotin-16-deoxyuridine triphosphate (biotin-16- dUTP) (Roche Diagnostics GmbH, Penzberg, Germany) were used as markers of newly synthesized DNA. If BrdU was used, cells were incubated in medium containing 20 µM BrdU in 5% CO₂ at 37°C for 10 min and processed for light or electron microscopy. Biotin-16-

(20)

20

dUTP was delivered into cells with a hypotonic shift procedure (Koberna et al., 1999). HeLa cells were quickly rinsed in pre-warmed KHB buffer (30 mM KCl, 10 mM HEPES, pH7.4), then overlaid with KHB containing 0.2 mM biotin-16-dUTP, incubated in a humidified chamber at 37°C for 5 min, washed, incubated in medium in 5% CO2 at 37°C for 10 or 3 min and processed for light or electron microscopy. Importantly, the biotin-16-dUTP is efficiently incorporated into DNA of cells for 15–20 min as determined from fluorescence intensity measurements (data not shown). In the case of pulse-chase-pulse experiments, BrdU and biotin-16-dUTP were used as replication markers. After the first pulse, the remaining BrdU was washed out by medium containing 100 µM thymidine. Cells were viewed using the Olympus Provis or Leica TCS NT microscope. The images were captured by a charge- coupled device camera (PXL with KAF 1400 chip; Photometrics, Ottobrunn, Germany) running on IPLab Spectrum or AnalySIS softwares. We avoided multinuclear cells as well as cells with large nuclei possessing apparently highly elevated genome copies, which were occasionally seen in the culture.

5.2.3 Antibodies

Mouse anti-bromodeoxyuridine antibody (Roche Diagnostics GmbH) and rabbit anti-biotin antibody (Enzo, New York, NY) were used as primary antibodies. The secondary anti-mouse and anti-rabbit antibodies conjugated with FITC, Cy3, or 6 nm or ultra-small grade gold particles, were kindly donated by Jackson ImmunoResearch Laboratories.

5.2.4 Electron Microscopy

If not stated otherwise, synchronized cells were used in all the electron microscopic studies. HeLa cells with incorporated BrdU were fixed in 8% formaldehyde in 0.2 M PIPES, pH 6.95 for 12 h, washed in PBS, dehydrated in methanol and propyleneoxide and embedded in Epon (Fluka Chemie GmbH). Ultrathin sections were cut, incubated with anti- bromodeoxyuridine antibody, washed in PBS, incubated with 6 nm gold anti-mouse adduct, washed in water and air-dried.

The ultrastructural mapping of newly synthesized biotin-16-dUTP labeled DNA was achieved using a pre-embedding approach. The cells were fixed in 2% formaldehyde in PBS for 10 min, washed in PBS, treated for two min each in 30, 50, 70, 90, 70, 50, and 30% ice- cold methanol and washed in PBS. Prior to immunolabeling, the cells were treated with 0.05 M glycine in PBS and then in 0.5% BSA for the blocking of unspecific binding. Biotinylated DNA was visualized with an anti-biotin antibody followed by an anti-rabbit ultrasmall gold

(21)

21

adduct. The silver intensification was performed according to Danscher (1981). Finally, the cells were post-fixed in 8% formaldehyde and dehydrated in gradually increasing methanol concentrations and propyleneoxide, and embedded in epon or lowicryl. Ultrathin epon and lowicryl sections were cut on a Reichert UltraCut E microtome with a Diatome diamond knife (Diatome Ltd.). The sections were stained with uranyl acetate and viewed using a Jeol 1200 EX or Zeiss EM 900 electron microscope equipped with MegaView II camera. The evaluation of gold labeling was performed using analySIS software (Soft Imaging System GmbH, Münster, Germany).

5.2.5 3-D Reconstruction and Stereological Analysis

Prints of electron micrographs from serially sectioned nuclei were aligned into a stack of images by registering contiguous features from adjacent sections. After alignment, a region of interest was cut from the stack and each section was scanned. Labeled sites were selected either by manual contour mapping or computer segmentation using IPLab software (Scanalytics, Fairfax, VA). For 3D reconstruction, the stacks of either pseudocolored segments or contours were projected in 3-D with IPLab software. Anaglyphs were prepared using two images from the 3-D projection series which were rotated 10º relative to one another.

The size of replication domains was estimated as follows: the most external gold (silver) particles were joined by straight lines given rise to a polygon. The maximum diameter and area of these polygons were measured. 300 RS were analyzed for each period of S-phase for three separate experiments. Only sites containing four or more gold particles were taken into account.

The ratio (RE:R_M:R_L)of relative nuclear volumes occupied by RS in early-, mid-, and late-S-phase cells was calculated as the ratio of relative nuclear areas occupied by RS. The square lattice of regularly spaced points was used for area estimation (Gundersen et al., 1988).

Thirty images of nuclear profiles from three independent experiments were evaluated for every S-phase stage. The value of REwas set to 1.00.

The ratio of nuclear volumes of cells in early-, mid-, and late-S-phase was estimated from stacks of confocal images. Cells were labeled with the DNA-specific dye YOYO-1 or with anti lamin A/C antibody and 3-D stacks of confocal images with the sampling 50 x 50 x 300 nm were acquired. The total area of all sections was calculated for 20 nuclei in every S- phase stage. If not stated otherwise, cells incorporating biotin-16-dUTP were used for all the stereological evaluations.

(22)

22 5.3 RESULTS

5.3.1 Kinetic Studies of Fluorescent Replication Patterns during S-Phase

Numerous studies using BrdU or biotinylated deoxyuridine incorporation and fluorescence microscopy have demonstrated three major types of replication site or foci patterns in a variety of mammalian cell lines (Fig. 1) corresponding to early (type I), mid (type II), and late (type III) stages of S-phase (Nakayasu and Berezney, 1989; van Dierendonck et al., 1989; Mazzotti et al., 1990; Fox et al., 1991; Kill et al., 1991; Manders et al., 1992; Neri et al., 1992; O‘Keefe et al., 1992; Sparvoli et al., 1994; Ferreira et al., 1997;

Somanathan et al., 2001; Dimitrova and Berezney, 2002). While some studies have identified five different patterns of foci, these correspond to further subdivision of the three basic types (Dimitrova and Berezney, 2002). As background for a comprehensive study of these replication patterns and the foci that compose them at the electron microscopic level, an analysis of the S-phase and the corresponding replication site patterns was performed in HeLa cells grown in monolayer. In an initial control, we demonstrated that a 10 min in vivo pulse of BrdU or biotin-16-dUTP had no effect on the doubling time of the HeLa cells (~24 h). The average length of S-phase in this cell line was estimated by means of a dual labeling pulse- chase-pulse experiment as described in Materials and Methods. An asynchronous cell population was labeled with BrdU for 10 min, chased in cultured medium without BrdU for 5, 6, 7, 8, 9, or 10 h followed by in vivo labeling with biotin-16-dUTP for 10 min after hypotonic shift (see Materials and Methods). Data from 900 cells from three independent experiments demonstrated that 16, 10, 7, 3, and 0% of cells exhibited two color signals following 5, 6, 7, 8, and 9 or 10 h chase periods, respectively. We conclude that the average S-phase duration is approximately 9 h. This was further supported by single labeling experiments with BrdU in asynchronously dividing HeLa cells. Approximately 38% of cells exhibited a replication signal after 10 min incubation with BrdU (results not shown).

Next, we determined the temporal order and length of the replication site patterns that appear during the progression of S-phase. Following a single 10 min pulse with BrdU, the typical five patterns of RS were identified (Dimitrova and Berezney, 2002) corresponding to a large number of small foci in early-S (types IA and IB), foci concentrated along the periphery of the nucleus and nucleoli characteristic of mid-S (type II) and larger foci over heterochromatin domains in late-S-phase (types IIIA and IIIB). Using a double labeling pulse-chase-pulse experiment (see Materials and Methods), we estimated the approximate

(23)

23

time for each replication pattern as 3.9 h for type I (early-S), 3.4 h for type II (mid-S), and 1.7 h for type III (late-S).

To confirm the temporal order of these replication patterns, we performed experiments with HeLa cells synchronized at the G1/S border with a double thymidine block (see Materials and Methods). Previous studies have demonstrated that this synchronization method did not affect the timing and organization of the replication patterns (Malinsky et al., 2001).

We found that 90% of the cells exhibited the early-S replication pattern 100 min after release from the thymidine block. After an additional 4 and 7 h, most of the cells exhibited mid- and late-S-phase patterns, respectively. This timing of replication patterns matched well our results in asynchronous cells and led us to conclude that we could use this synchronization procedure to accurately examine, by electron microscopy, the replication foci in early-, mid-, and late-S-phases.

5.3.2 Identification of DNA Replication Sites (RS) Following High Resolution Electron Microscopic Immunogold Labeling

As a first step in examining the organization of RS at the electron microscopic level, HeLa cells synchronized by the double thymidine block procedure were processed for post- embedding immunogold labeling during early-S-phase (100 min after release), mid-S-phase (340 min after release), and late-S-phase (520 min after release). Typical results are shown in Figure 2A–C. While the gold particles observed during early- and mid-S-phase are typically clustered into small regions that could correspond to discrete RS or foci (Fig. 2A, B, insets), the relatively weak labeling resulted in significant distances between many of the gold particles. This makes it difficult to estimate the size and borders of these presumptive clusters.

Similarly, the gold particles observed in late-S-phase were often grouped together into small clusters, but the relatively large distances between many of the particles, made it difficult to determine how these small clusters might be arranged into the larger foci characteristic of late-S-phase (Fig. 2C, inset). Indeed, determining where one cluster ends and another one begins was a challenging task.

These findings led us to develop a more sensitive approach for labeling RS at the electron microscopic level. We reasoned that a significant increase in the number of gold particles decorating the RS would enable a more accurate identification and evaluation of these individual sites. Three major changes in conditions were adapted to enhance the sensitivity of immunogold labeling (see Materials and Methods). First, rather than incorporation of BrdU, a hypotonic shift method (Koberna et al., 1999) was used to

(24)

24

incorporate biotin-16-dUTP label directly into living cells. Secondly, the samples were processed for immunogold labeling before embedding (pre-embedding) rather than post- embedding. This enabled immunogold labeling to occur through the fixed cell sample rather than on a single thin section embedded in the plastic resin. Lastly, silver enhancement was used to increase the signal intensity of the gold particles.

Use of this improved pre-embedding protocol resulted in a striking enhancement in the signal intensity over individual RS. Discrete clusters of closely spaced gold particles of high intensity were now typically observed that resembled replication foci in early-, mid-, and late- S-phase (Fig. 2D–F). At higher magnification (see insets to Fig. 2D–F) these sites appear to be similar in size throughout S-phase. While the individually labeled gold clusters in early-S- phase are generally at a distance (>0.2 µm) that would enable their resolution as separate foci by fluorescence microscopy (Fig. 2D, inset), this is not the case for late-S (or in some instances mid-S patterns) where discrete clusters are observed within a 0.1 µm distance of each other (Fig. 2 E, F, insets). These findings suggest that the larger foci observed in late- and mid-S-phase by fluorescence microscopy, are composed of smaller foci. The smaller foci are resolvable at the electron microscopic level and appear to be similar in size to the replication foci observed in early-S-phase.

This prompted us to determine the size of the individual gold clusters following the 10 min labeling period. Three hundred measurements from 30 different nuclei and three separate pre-embedding experiments were performed on gold clusters from each of the three periods of S-phase. The average area occupied by individual clusters in early-, mid-, and late-S-phase was strikingly similar (4,515, 4,650, and 4,625 nm², respectively) as were the frequency distributions of area for each period of S-phase (Fig. 3A, C). Consistent with the area measurements, the average size of individual clusters based on the diameter was nearly identical (110, 120, and 110 nm) in early-, mid-, and late-S-phase, respectively, as were the corresponding frequency distributions of measured diameters (Fig. 3B, D). These data demonstrate that the size of the gold clusters, and the individual replication foci that they represent, do not change significantly during S-phase. Moreover, they are in accordance with our measurements of RS in non-synchronized cells. The average area of individual RS in non- synchronized cells was 4,605 nm² and thus matched the sizes found throughout S-phase in synchronized cells.

Previous fluorescent microscopic studies demonstrated that the replication foci in early-S-phase do not significantly change their size in pulse periods ranging from 2 to 30 min (Ma et al., 1998). Similarly, we compared the size of RS of early-S-phase labeled with biotin-

(25)

25

16-dUTP for 3 versus 10 min. The average area of replication clusters was strikingly similar (4,550 ±1,300 and 4,345 ±1,220 nm²) for 10 and 3 min labeling intervals, respectively. It is likely that this small difference in area size is a result of the much lower labeling density of RS following the 3 min versus the 10 min pulse (data not shown) rather than to a real change in the size of replication foci.

5.3.3 Replication Foci Labeled in Early-S-Phase Persist as Similarly Labeled Sites throughout S-Phase and Into the Next Cell Generation

Studies at the fluorescent microscopic level have demonstrated that the replication foci labeled in early-S-phase, are maintained as identically-looking labeled foci throughout the S- phase and into subsequent cell generations (Sparvoli et al., 1994; Jackson and Pombo, 1998;

Ma et al., 1998; Zink et al., 1998). These findings have contributed to the current view that the early-S-phase replication foci represent fundamental units of higher order chromatin domains (Jackson and Pombo, 1998; Ma et al., 1998; Zink et al., 1998; Berezney, 2002). If the gold clusters that we have visualized in our microscopic analysis correspond to these early-S-phase replication foci, they should also persist as similar clusters in subsequent cell generations. To address this issue, HeLa cells synchronized in early-S-phase were pulsed for 10 min with biotin-16-dUTP and chased for 4 h into a later stage of S-phase or for 18 h into the G1-phase of the next cell generation. In both cases, clusters of gold particles were observed scattered throughout the extranucleolar regions of the nucleus in patterns very similar to corresponding gold clusters observed during early-S-phase (Fig. 4).

5.3.4 Calculating the Relative Number of Small RS during S-Phase Progression

Using stereologic analysis and assuming that the average size of the small replication foci is virtually the same throughout S-phase, we estimated the relative number of small RS present at a given time in S-phase. First, we determined the ratio of relative nuclear volumes occupied by the total RS present on an average in early-, mid-, and late-S-phases (see Materials and Methods). For the pre-embedding approach, this ratio was approximately 1.00, 0.93, and 0.54 for early-, mid-, and late-S-phase cells, respectively. Importantly, data derived from BrdU labeled cells (post-embedding approach) provided similar results: 1.00, 0.92, and 0.62, respectively. As the pre-embedding approach enabled clearer resolution of replication foci (compare Fig. 2A—G), only data derived from this method were used for further analysis.

(26)

26

Since during S-phase progression, the nuclear volume increases (Fidorra et al., 1981), the relative volumes occupied by RS needed to be normalized to the relative nuclear volumes in early-, mid-, and late-S-phase cells, respectively. Using confocal microscopy (see Materials and Methods), we determined the ratios of total nuclear volumes to be 1.00 (early-S), 1.09 (mid-S), and 1.28 (late-S), respectively. Next, we calculated the relative number of RS as a product of the relative volume occupied by RS and the relative nuclear volume and obtained values of 1.00 (early-S), 1.01 (mid-S), and 0.79 (late-S). Our analysis, therefore, indicates that the relative number of active small RS at anytime during S-phase is approximately constant with a small decrease (~20%) during late-S-phase.

5.3.5 3-D Reconstruction of RS from Electron Microscopic Serial Sections

Serial sections from cells labeled at early- and late-S-phase were processed for 3-D reconstruction (see Materials and Methods). Briefly, the serial sections were aligned and projected in 3-D for direct observation as anaglyphs. In early-S-phase, the gold clusters corresponding to individual replication foci are arranged three-dimensionally into higher order arrays that form a network-like organization (Fig. 5E). Contour mapping of the gold clusters was then performed on the individual sections (Fig. 5A—D), followed by 3-D reconstruction of the contours at each section. Consistent with the analysis of the gold labeled sites (Fig. 5E), the contours form higher order arrays that resemble network-like structure (Fig. 5F).

In late-S-phase, the smaller clusters of gold particles that compose the much larger replication foci merge together in 3-D (Fig. 6E). Contour mapping of the individual gold clusters at each section (Fig. 6A—D) followed by 3-D reconstruction, demonstrates the large number of contours that typically compose the larger replication foci of late-S-phase (Fig.

6F). The number of contours at these large late-S-phase replication foci ranged—in correspondence to the size of the foci—from <10 to >75.

5.4 DISCUSSION

Fluorescent microscopic studies of DNA RS or foci have contributed strikingly to our understanding of the structural organization of DNA replication in mammalian cells. As the cell progresses through S-phase, distinct patterns of these RS are observed that are characteristic of early-, mid-, and late-S-phases (Fig. 1 and Nakayasu and Berezney, 1989;

van Dierendonck et al., 1989; Mazzotti et al., 1990; Fox et al., 1991; Kill et al., 1991;

Manders et al., 1992; Neri et al., 1992; 0'Keefe et al., 1992; Sparvoli et al., 1994; Ferreira et

(27)

27

al., 1997; Somanathan et al., 2001; Dimitrova and Berezney, 2002). More detailed studies including three-dimensional fluorescence microscopy and computer image analysis have led to current models of RS as discrete chromatin domains containing an average of ~1 Mbp of DNA (Nakamura et al., 1986; Jackson and Pombo, 1998; Ma et al., 1998; Berezney et al., 2000; Berezney, 2002). One assumption of these models is that the chromatin domains measured in early-S-phase are also the basic units for replication in the later stages of S- phase. The larger replication foci that are observed in mid- and late-S-phase are interpreted as being comprised of numerous RS of similar size as the early-S sites (see Fig. 1 and Nakayasu and Berezney, 1989). Since the larger foci typically correspond to heterochromatic region, it is a reasonable assumption that the much higher compaction of chromatin in these regions makes it difficult to visualize the individual smaller RS that might comprise the larger foci.

While studies involving GFP expression of the major replicational factor proliferating cell nuclear antigen (PCNA) provide evidence for this possibility (Leonhardt et al., 2000), these observations are limited in both scope and by the resolution of light microscopy.

A number of previous electron microscopic studies using colloidal gold labeling, have suggested that the RS are organized into clusters that might correspond to the replication foci observed in early-S-phase and the larger foci observed in late-S-phase by fluorescence microscopy (Raška et al., 1989, 1991; Mazzotti et al., 1990; Rizzoli et al., 1992). These previous studies, however, are typically limited by the relatively weak labeling obtained using traditional post-embedding protocols resulting in relatively large distances between individual gold particles and a corresponding difficulty in accurately assigning individual gold particles to higher order clusters (see Fig. 2A–C). While larger regions of gold particles are typically observed in late-S-phase that likely correspond to the larger foci identified by fluorescence microscopy, the limited labeling has made it difficult to resolve the smaller foci that might comprise the larger labeled regions. Immunogold labeling for PCNA, however, has provided some evidence for smaller clusters of replicative activity within the larger replication foci of late-S-phase (Raska et al., 1989).

To circumvent these technical difficulties, we have developed a highly sensitive electron microscopy (EM) localization procedure for RS involving direct in vivo incorporation of biotin-16-dUTP followed by labeling with colloidal gold particles and silver enhancement before embedding of the specimens (pre-embedding) for electron microscopy.

Under these conditions, we have observed the close association of numerous silver grains at numerous sites in early-S-phase. The labeled sites had discrete sizes as evaluated by both area and diameter determinations (Fig. 3). Moreover, the much larger heterochromatin domains

(28)

28

that replicate in late-S-phase, and to a lesser extend in mid-S-phase, are actually composed of closely associated labeled foci with virtually identical diameters and areas as the early-S labeled foci (Fig. 3). Pulse-chase experiments demonstrated that the RS labeled in early-S- phase are maintained as similarly labeled clusters of colloidal gold particles later in S-phase and following the next cell generation (Fig. 4).

These findings provide direct evidence at the electron microscopic level for a common size of replication foci throughout the S-phase and their maintenance as fundamental higher order chromatin domains throughout the cell cycle and into subsequent cell generations.

Stereologic analysis of labeled electron microscopic sections further demonstrates that the total relative volume occupied by the entire population of RS is virtually identical in early- and mid-S-phases with only a small decrease (approximately 20%) in late-S-phase. Since the average size of the small replication foci is identical throughout S-phase, this implies that the total number of RS at any given time in S-phase is relatively constant with a small decrease during late-S-phase. Calculation of the total number of RS based on data from early-S-phase and extrapolation to the other periods of S-phase, therefore, provides a reasonable accurate approximation and supports the model of fundamental chromatin domains of ~1 Mbp during DNA replication and throughout the cell cycle (Jackson and Pombo, 1998; Ma et al., 1998;

Zink et al., 1998; Berezney, 2002). Further stereologic studies at the electron microscopic level will be required to directly calculate the number of RS at each period of S-phase and to determine the average amount of DNA present at each chromatin domain.

3-D reconstruction of serial sections enabled a more detailed visualization of these labeled RS. Consistent with previous studies using 3-D confocal microscopy (Wei et al., 1998), we find that the early-S sites form an overall network of association throughout the cell nucleus (Fig. 5). Analysis of the closely associated RS of late-S heterochromatin domains enabled us to better visualize the many individual RS that compose these larger replication foci (Fig. 6).

In summary, the characteristic RS of early-S-phase is demonstrated for the first time at the electron microscopic level. Moreover, late-S heterochromatin is replicated in similar-sized chromatin domains. This leads to the conclusion that replication occurs in similarly-sized RS throughout the S-phase and supports the view that these chromatin domains are universal features of higher order organization and function in the mammalian cell nucleus. 3-D electron microscopic visualization further reveals that the RS during early-S-phase are arranged into higher order networks (Fig. 5). These networks, previously identified following 3-D confocal microscopy and computer image analysis (Wei et al., 1998), have been

(29)

29

proposed to form the structural basis for the coordination of DNA replication and transcription programming in the mammalian cell (Berezney and Wei, 1998; Berezney, 2002).

(30)

30 5.5 FIGURES

Fig. 1 The three main types of replication site patterns in the mammalian cell nucleus are composed of similarly sized foci.

Type I (early-S) patterns consist of numerous foci distributed throughout the nuclear interior.

In type II (mid-S) patterns, the replication foci are predominantly of similar size as in early-S- phase and are concentrated along the nuclear and nucleolar borders in association with heterochromatin. Some foci, however, appear larger than those in early-S-phase due to the tight packing of the individual foci over heterochromatic regions along the periphery of the nucleus and nucleolus. Type III (late-S) patterns are composed of much larger replication foci distributed over the late replicating heterochromatinic regions. As demonstrated in this study, these larger foci are comprised of numerous smaller replication foci that are identical in size to the replication sites (RS) observed in early-S-phase. This is illustrated for the replication site in the boxed area.

nu, nucleolus

(31)

31

Fig. 2 Electron microscopy mapping of RS on thin sections of HeLa cells

Cells were either labeled with 5-bromo-5^‘-deoxyuridine (BrdU) (A–C: post-embedding detection), or with biotin-16-deoxyuridine triphosphate (biotin-16-dUTP) (D–G: pre- embedding detection). Cells were synchronized in early (A, D, G), mid (B, E), and late (C, F) S-phase. Groups of gold particles decorating small foci were observed after 10 min of labeling (insets to D-F). Several such foci were sometimes clustered in mid-S and frequently in late-S-phase (E, F, and insets). Similarly sized foci were observed after 3 min labeling pulse (G and inset). Bar: 1 µm. Insets are 2x magnified

(32)

32

Fig. 3 Quantitative analysis of replication foci size during S-phase

The total distribution of replication foci in terms of area (A) and diameter (B) of individual foci are plotted for early-S (circles), mid-S (triangles), and late-S (diamonds). C and D show the average areas and diameters, respectively, for early-, mid-, and late-S-phase replication foci. Error bars correspond to 1 standard deviation.

(33)

33

Fig. 4 Electron microscopy mapping of chromatin domains previously labeled in early-S- phase

Cells were synchronized in early-S-phase, labeled for 10 min with biotin-16-dUTP and chased for 4 h (A) or 18 h (B). Groups of gold particles decorate small chromatin domains in both images that are similar to the replication foci labeled in early-S-phase (see Fig. 2). Bar: 500 nm.

(34)

34

Fig. 5 Three-dimensional visualization of RS in early-S-phase

Images of seven consecutive serial sections of early-S-phase labeled replication foci were aligned using ultrastructural markers common to adjacent sections (see Materials and Methods). Panels A–D show four representative sections in which the replication foci are outlined with yellow contours. Black field anaglyphs of the 3-D reconstructions are shown for the computer segmented replication foci (pseudo-colored in yellow, panel E) and for contours (yellow) outlining the replication foci (panel F). The images of panels E and F can be observed in 3-D with red/blue-green viewers.

Fig. 6 Three-dimensional visualization of RS in late-S-phase

Images of nine consecutive serial sections of late-S-phase labeled replication foci were aligned using ultrastructural markers common to adjacent sections (see Materials and Methods). Panels A–D show four representative sections in which the small replication foci within the larger-sized foci of late-S-phase are outlined with yellow contours. Black field anaglyphs of the 3-D reconstructions are shown for the computer segmented replication foci (pseudo-colored in yellow, panel E) and for contours (yellow) outlining the small replication foci (panel F). The images of panels E and F can be observed in 3-D with red/blue-green viewers.

(35)

35

CHAPTER 6 ORGANIZATION OF HUMAN REPLICON: SINGLES OR ZIPPING COUPLES?

6.1 INTRODUCTION

During the replication of both prokaryotic and eukaryotic genomes, two replication complexes, commonly called replisomes, are believed to be established at the origins of the replication. According to this widely accepted scheme, the synthesis of DNA in opposite directions from the replication origin is ensured by couples of ‗‗sister‖ replisomes (Baker and Bell, 1998; Waga and Stillman, 1998; Johnson and O‘Donnell, 2005). Two basic arrangements of a replisome couple during the DNA replication have been suggested earlier.

In the ﬁrst one, ‗‗sister‖ replisomes move along the DNA molecule in opposite directions. The second one is based on a tight association of replisomes in a replisome couple. This organization results into a transient formation of a loop consisting of newly synthesized DNA.

Although the model of tightly associated ‗‗sister‖ replisomes was suggested already in 1974 (Dingman, 1974), the ﬁrst convincing proof of such arrangement was not available until much later. The fact that prokaryotic cells use this type of replisome organization has been indicated by several recent ﬁndings (Lemon and Grossman, 2000; Jensen et al., 2001; Lau et al., 2003;

Migocki et al., 2004). However, the only supporting data for such organization in eukaryotic cells has been yet provided in budding yeast Saccharomyces cerevisiae (Kitamura et al., 2006, commented by Meister et al., 2006).

The DNA of vertebrates is replicated via a large number of DNA segments termed as replicons, which are continuously activated in the S phase (Edenberg and Huberman, 1975;

Hand, 1978). The size of the individual replicons usually varies from 30 to 450 kbp, with the most frequent size being 75–175 kbp, although replicons below 10 kbp and above 1 Mbp have also been described (Edenberg and Huberman, 1975; Yurov and Liapunova, 1977; Hand, 1978; Hyrien and Mechali, 1993; Jackson and Pombo, 1998; Berezney et al., 2000). Based on studies of stretched DNA ﬁbers, it is supposed that clusters of adjacent replicons are usually synchronously activated and jointly ensure the replication of several hundreds of kilobases of DNA (Edenberg and Huberman, 1975; Hand, 1978). The number of replicons in one such replicon cluster varies but is usually less than 10 (Jackson and Pombo, 1998; Ma et al., 1998).

In situ, replicon clusters are commonly identiﬁed with replication foci (light microscopy/LM entities) or replication factories (electron microscopy/EM entities), structures which can be

Text práce (2.470Mb)

First Faculty of Medicine, Charles University in Prague

DNA REPLICATION AND CHROMATIN: FROM 3D TO FUNCTION

PhD Thesis

Anna Ligasová

Prague 2009

ACKNOWLEDGEMENT

CONTENTS

ABBREVIATIONS

CHAPTER 1

OUTLINE OF THE THESIS

CHAPTER 2

GENERAL INTRODUCTION

CHAPTER 3

SPECIFIC AIMS OF THE THESIS

CHAPTER 4

MATERIALS AND METHODS

CHAPTER 5

ELECTRON MICROSCOPY OF DNA REPLICATION IN 3-D:

EVIDENCE FOR SIMILAR-SIZED REPLICATION FOCI THROUGHOUT S-PHASE

CHAPTER 6

ORGANIZATION OF HUMAN REPLICON: SINGLES OR ZIPPING COUPLES?