Hum Genet (1998) 102 : 241–251

© Springer-Verlag 1998

R A P I D C O M M U N I C AT I O N

Daniele Zink · Thomas Cremer · Rainer Saffrich · Roger Fischer · Michael F. Trendelenburg · Wilhelm Ansorge · Ernst H. K. Stelzer

Structure and dynamics of human interphase chromosome territories in vivo

Received: 10 November 1997 / Accepted: 24 November 1997

Abstract A new approach is presented which allows the in vivo visualization of individual chromosome territories in the nuclei of living human cells. The fluorescent thymidine analog Cy3-AP3-dUTP was microinjected into the nuclei of cultured human cells, such as human diploid fibroblasts, HeLa cells and neuroblastoma cells. The fluorescent analog was incorporated during S-phase into the replicating genomic DNA. Labelled cells were further cultivated for several cell cycles in normal medium. This well-known scheme yielded sister chromatid labelling. Random segregation of labelled and unlabelled chromatids into daughter nuclei resulted in nuclei exhibiting individual in vivo detectable chromatid territories. The territories were composed of subcompartments with diameters ranging between approximately 400 and 800 nm which we refer to as subchromosomal foci. Time-resolved in vivo studies demonstrated changes of positioning and shape of territories and subchromosomal foci. The hypothesis that subchromosomal foci persist as functionally distinct entities was supported by double labelling of chromatin with CldU and IdU, respectively, at early and late S-phase and subsequent cultivation of corresponding cells for 5–10 cell cycles before fixation and immunocytochemical detection. This scheme yielded segregated chromatid territories with distinctly separated subchromosomal foci composed of either early- or late-replicating chromatin. The size range of subchromosomal foci was similar after shorter (2 h) and longer (16 h) labelling periods and was observed in nuclei of both living and fixed cells, suggesting their structural

D. Zink (Y) · T. Cremer (Y) Institut für Anthropologie und Humangenetik, LMU München, Richard-Wagner-Strasse 10/I, D-80333 Munich, Germany Tel.: +49-89-5203-381; Fax: +49-89-5203-389 R. Saffrich · W. Ansorge · E. H. K. Stelzer European Molecular Biology Laboratory, Meyerhofstrasse 1, D-69117 Heidelberg, Germany R. Fischer · M. F. Trendelenburg Deutsches Krebsforschungszentrum, Abteilung 0195, Im Neuenheimer Feld 280, D-69120 Heidelberg, Germany

identity. A possible functional relevance of chromosome territory compartmentalization into subchromosomal foci is discussed in the context of present models of interphase chromosome and nuclear architecture.

Introduction In recent years the nucleus has emerged as a highly compartmentalized structure. It has been demonstrated that chromosomes form distinct territories in both animal and plant nuclei (Manuelidis 1985; Schardin et al. 1985; Cremer et al. 1988; Lichter et al. 1988; Pinkel et al. 1988; Leitch et al. 1990; for review see Cremer et al. 1993). Chromosome territories seem to be further partitioned into discrete chromosomal compartments, such as chromosome arm and band-like domains, centromeric domains and telomeric domains (Manuelidis 1990; Lawrence et al. 1993; Schedl and Grosveld 1995; Zhao et al. 1995; Kurz et al. 1996; Dietzel et al. 1998). Single active or inactive genes were visualized as apparently discrete dot-like domains (e.g. Kurz et al. 1996). The 3D-positioning of genes within or at the surface of chromosome territories has become a topic of active research (Cremer et al. 1993; Eils et al. 1996; Kurz et al. 1996). Nuclear proteins which are involved in nuclear functions, such as transcription, splicing, DNA replication and repair, participate in higher-order macromolecular domains of a size that can often be recognized in immunofluorescent labelling experiments as typical foci, speckles or punctate distributions of nuclear antigens (Spector 1993; Roth 1995; van Driel et al. 1995). During terminal differentiation of cells complex movements of chromatin were noted (for reviews see Manuelidis 1990; De Boni 1994). In spite of these advances we do not understand the functional meaning of topological relationships between higher-order nuclear structures and their dynamics and lack a coherent and widely accepted theory of the functional chromosome territory and nuclear architecture. Studies of chromosome territory architecture using fluorescence in situ hybridization (FISH) require fixed cells. It is not clear to what extent experimental artifacts have

242

affected the conclusions derived to date from these studies with regard to territory size, shape, surface structure and the extent of subcompartimentalization. Depending on the type of fixation and subsequent treatments necessary for permeabilizing cell nuclei for FISH probes or antibodies, shrinkage artifacts on the one hand and artificial dispersion of chromatin and other nuclear constituents on the other hand have to be taken into consideration. An alternative approach to the study of chromosome territories has been recently developed. It has been amply demonstrated that the incorporation of thymidine analogs, such as bromodeoxyuridine (BrdU), results in sister chromatid labelling at the second mitosis after the labelling event (Latt 1973). Labelled and unlabelled chromatids are randomly segregated to daughter cells during the second and all subsequent mitotic events. After five to ten cell cycles nuclei can be observed that contain few or even single chromatid territories (Ferreira et al. 1997; A. Lamond, personal communication). The immunocytochemical detection of incorporated halogenated thymidine analogs requires cell fixation and a denaturation step. Accordingly, results derived from labelling experiments with halogenated thymidine analogs are open to concerns of artifacts as noted above for FISH experiments. To overcome these limitations we set ourselves the goal of visualizing individual chromosome territories in the nucleus of the living human cell. In contrast to mitotic chromosomes (e.g. Hiraoka et al. 1989) the visualization of individual chromosome territories within cell nuclei has not been compatible with living cell microscopy except for a few specialized experimental systems (Hochstrasser et al. 1986; Hochstrasser and Sedat 1987). In this report we demonstrate that the fluorescent thymidine analog Cy3AP3-dUTP can be stably integrated during S-phase in the genomic DNA of living cells. Labelled cells continued to grow yielding descendants with replication-labelled and segregated chromatid territories that could be observed directly in the living cell.

(CldU) (Sigma), were added to the culture medium, each for 2 h, at a final concentration of 10 µM (Aten et al. 1992; 1993). The first pulse was followed by a chase of 4 h. After the second pulse fibroblasts were washed twice and subsequently grown in normal culture medium for up to 2 weeks. During this growth period the pulse-labelled chromatids segregated to the daughter cells. Cells with segregated chromatids were fixed at mitosis or at G1/G0. Chromosome spreads were prepared following standard protocols. Details of procedures for obtaining pre-S-phase cells will be given elsewhere (Zink et al., manuscript in preparation). Alternatively, in order to obtain chromatid territories labelled throughout an entire S-phase, cells were synchronized and blocked at the G1/S boundary with Aphidicolin (Boehringer Mannheim, 5 µg/ml medium) according to the following scheme: 12 h block, 12 h chase, 24 h block. After the second Aphidicolin block cells were directly released in bromodeoxyuridine (BrdU)-containing medium (10 µg/ml) and kept there for 16 h. Subsequently the completely BrdU-labelled chromatids were segregated in normal medium for 5 days.

Materials and methods

Chromosome painting

Cell culture

A quantity of 200 ng DNA from a PCR-amplified 15q microdissection probe (Guan et al. 1996) was nick-translated with biotin-16dUTP (Boehringer Mannheim) and precipitated together with 30 µg Cot-1 DNA (BRL) and 50 µg salmon testis DNA (Sigma). The precipitate was dissolved in 10 µl of a hybridization mixture (50% formamide, 10% dextran sulfate, 1 × SSC), denatured for 5 min at 75° C and preannealed for 20 min at 37° C. The preannealed probe was applied to the pretreated and denatured nuclei (3 min at 75° C in 70% formamide/0.6 × SSC). Preparations were sealed with rubber cement and incubated for 3 days at 37° C.

Diploid human oral cavity fibroblasts (GF 032, kindly provided by P. Tomakidi, Kopfklinik Heidelberg) and HeLa cells (subclone 6, kindly provided by W.W. Franke, DKFZ, Heidelberg) were cultured with Dulbecco’s modified Eagle’s medium (DMEM). The neuroblastoma line SH-EP N14 (derivative of SH-EP cells (Ross et al. 1983), stably transfected with a CMV N-Myc expression vector (Wenzel et al. 1991), kindly provided by J. Schürmann, DKFZ, Heidelberg) was grown in RPMI medium. Both media were supplemented with 2 mM glutamine, antibiotics (penicillin 100 µg/ml and streptomycin 100 µg/ml) and 10% FCS. The cells were kept at 37° C in an atmosphere of 5% CO2. The doubling time is about 10 h for cell line SH-EP N14 and about 24 h for the other two cell lines. Replication labelling combined with segregation and immunodetection For replication labelling with immunodetection (fibroblasts), two thymidine analogs, iododeoxyuridine (IdU) and chlorodeoxyuridine

Replication labelling with Cy3-AP3-dUTP, segregation and control of incorporation Cell nuclei were microinjected using a Zeiss AIS (automated injection system) provided with a capillary puller (P87, Sutter). For each coverslip approximately 100 nuclei were microinjected during a period of 45 min with a solution of Cy3-AP3-dUTP (Amersham) diluted in PBS to a final concentration of 10 mM (fibroblasts) or 50 mM (HeLa and SH-EP N14 cells). After microinjection segregated chromatid territories were obtained by growing the cells for 50 h (SH-EP N14 cells), 5 days (HeLa) or 1 week (fibroblasts) in normal culture medium. To check for proper incorporation of the label, cells were fixed (3.7% formaldehyde in PBS for 10 min) after in vivo microscopy, permeabilized with 0.5% Triton in PBS and washed with vigorous shaking for 5 h in this solution. Light microscopy revealed no obvious difference to the structures observed in vivo. Pretreatment of interphase nuclei for immunodetection combined with chromosome painting IdU/CldU-labelled fibroblasts were washed twice in PBS and fixed for 10 min in PBS/3.7% formaldehyde. After an additional wash in PBS the fibroblasts were permeabilized as described (Eils et al. 1996), except that cells were incubated for 3 nights in 50% formamide/1 × SSC at 37° C prior to chromosome painting.

Detection of replication label and hybridized DNA probes Coverslips with IdU/CldU-labelled and hybridized nuclei were washed for 3 × 5 min with 50% formamide/1 × SSC and for 3 × 5 min with 0.05 × SSC at 37° C. After rinsing in PBS, cells were blocked with a solution of PBS/5% BSA/0.2% Tween 20/0.2% NP-40. For detection of biotinylated sequences the nuclei were incubated for 2 h with avidin-Cy5 (Dianova) dissolved (1/300) in the blocking buffer. Thereafter, nuclei were washed for 3 × 10 min in PBS/0.2%

243 Tween 20/0.2% NP-40 and incubated overnight at 4° C with the following antibodies: biotinylated goat anti-avidin (Dianova, 1/200), rat anti-BrdU [recognizes BrdU and CldU (Aten et al. 1992, 1993); Seralab, 1/100] and mouse anti-BrdU [recognizes BrdU and IdU (Aten et al. 1992, 1993); Becton and Dickinson, 1/6] diluted in PBS/ 5% BSA/0.2% Tween 20/0.2% NP-40/10% normal goat serum. The coverslips were then washed for 2 × 5 min in PBS, 10 min in PBS/400 mM NaCl/0.2% Tween 20/0.2% NP-40 and rinsed in PBS. Subsequently the nuclei were blocked again for 30 min with PBS/5% BSA/10% normal goat serum and incubated for 2 h with the following reagents: FITC-conjugated goat anti-rat IgG antibody (Dianova, 1/100), TRITC-conjugated goat anti-mouse IgG antibody (Dianova, 1/100) and avidin-Cy5 (Dianova, 1/300) dissolved in the blocking solution. Subsequently the coverslips were washed for 3 × 10 min in PBS/0.2% Tween 20/0.2% NP-40 and rinsed in PBS. Nuclei were stained for 5 min in PBS/DAPI, washed for 5 min in PBS and embedded in Vectashield (Vector) for microscopy. Immunodetection of whole BrdU-labelled chromosome territories Fibroblasts were fixed with formaldehyde as described above and permeabilized for 3 × 5 min in PBS/0.1% Triton/0.1% Tween. The DNA was denatured for 30 min in 2 M HCl. Subsequently the slides were rinsed several times in PBS and blocked with PBS/5% BSA. BrdU was detected by incubation with a rat anti-BrdU (Seralab, 1/100) antibody in the same solution. After washing for 3 × 5 min with PBS the slides were incubated with a FITC-conjugated anti-rat IgG antibody (Dianova, 1/100) and washed again (3 × 5 min PBS) before counterstaining and embedding (see above). Microscopy and image acquisition Epifluorescence microscopy was performed with either a Leica TCS or a Zeiss Axiophot microscope. Nuclear images were taken with a video camera (Kappa) or a color slide film (Kodak Ektachrome 400). Optical sections were sampled with a three-channel Leica TCS 4D confocal microscope as described by Eils et al. (1996) (fixed cells) and an early prototype of a Carl Zeiss LSM5 (built at EMBL) confocal microscope (living cells). For living cell microscopy HeLa and SH-EP N14 cells were kept in an FCS2 chamber (Bioptechs) at 37° C. Depending on cell thickness a stack of 15 (SH-EP N14) to 29 (HeLa) sections at an axial distance of 0.5 µm was recorded in less than 2 min. Images were transferred to a Power Macintosh computer (color slides were scanned with the Agfa StudioScan IIsi scanner). Standard software tools (Adobe Photoshop 4.0; NIH Image 1.60) were used for image analysis.

Results Replication labelling and segregation experiments with BrdU, CldU and IdU thymidine analogs: studies of chromosomal subdomains in fixed cell nuclei Cultures of human diploid fibroblasts (HDF) were labelled with halogenated thymidine analogs (for details see Materials and methods). One to two weeks later (i.e. 5–10 cell cycles after the labelling event) we prepared metaphase spreads showing a small number of typically non-homologous chromosomes with one uniformly labelled and one unlabelled chromatid (Fig. 1A) and occasional sister chromatid exchanges (not shown). Division of such cells yielded daughter nuclei showing a few replication-labelled territories (or a partial territory in case of a sister chromatid exchange) within a background of unlabelled chromatin

(Fig. 1B, C). We refer to these replication-labelled territories as chromatid territories, since they represent individual chromatids throughout the entire cell cycle. (Note that in pre-S-phase nuclei a chromatid territory is equal to the entire chromosome territory, while in G2 nuclei it represents half of a chromosome territory). Chromosome painting was applied to identify individual replication-labelled chromatid territories in fixed pre-S-phase HDF nuclei. For example, Fig. 1C–E presents a cell nucleus demonstrating the colocalization of a replication-labelled, segregated chromatid territory with a chromosome 15q paint probe. In this nucleus the paint probe also identified the homologous territory, which does not show any replication label. Replication-labelled and segregated chromatid territories in fixed cell nuclei revealed subchromosomal foci with diameters ranging approximately from 400 nm to 800 nm (Figs. 1F–H, 2D). Labelling periods with halogenated thymidine analogs that varied from 2 h throughout an entire S-phase did not noticeably affect this size range (Fig. 1F– H, 2D). In order to test whether the 400- to 800-nm subchromosomal foci observed in nuclei with segregated chromatid territories represented functionally distinct entities a double-pulse labelling scheme was applied. IdU was added to the culture medium for 2 h, followed by a chase of 4 h and a second pulse with CldU for 2 h. Since the duration of the entire S-phase of human diploid fibroblasts was approximately 10 h (data not shown), all nuclei that had incorporated both nucleotides had incorporated IdU in chromatin replicated during the first half of S-phase (“early”-replicating chromatin), while CldU labelled chromatin represented chromatin replicated during the second half of S-phase (“late”-replicating chromatin). Labelled cells were fixed at G1/G0 after an additional cultivation period of up to 2 weeks. Antibodies specific for either IdU or CldU allowed the differential staining of IdU- or CldUlabelled chromatin in fixed nuclei (Aten et al. 1992, 1993). Figure 1F–H shows a typical result [> 60 nuclei were analyzed in detail; a comprehensive description will be given elsewhere (Zink et al. manuscript in preparation)]. The subchromosomal foci observed in segregated chromatid territories had retained their distinct separation of greenlabelled early-replicating and red-labelled late-replicating chromatin. This result demonstrates that subchromosomal foci are functionally distinct structures composed of either early- or late-replicating chromatin which persist from cell cycle to cell cycle. Replication labelling and segregation experiments with the thymidine analog Cy3-AP3-dUTP: studies of subchromosomal foci in living cell nuclei Definitive proof of the in vivo existence of subchromosomal foci required an approach compatible with living cell microscopy. To this end nuclei of HDF, of a human neuroblastoma cell line and of HeLa cells were microinjected with the fluorescent thymidine analog Cy3-AP3-dUTP (Amersham). One day and 1 week after microinjection, liv-

244

245 F Fig. 1A–H Replication-labelled and segregated human chromatids. Cultures of human diploid fibroblasts were grown after initial replication labelling for several cell cycles. All panels show chromatids of fixed cells labelled with IdU/CldU according to the scheme: 2 h IdU pulse/4 h chase/2 h CldU pulse. Both the IdU and the CldU label are depicted in green in panels A–E. In panels F–H the IdU label is shown in green and the CldU label is depicted in red. A Metaphase spread with two differently sized (non-homologous), replication-labelled chromosomes. Note that only one of the two sister chromatids contains the label. Chromosomes were counterstained with DAPI (blue). B Interphase nuclei counterstained with DAPI from the same slide as the metaphase shown in A depict replication-labelled and segregated chromatid territories. Small patches of replication-labelled chromatin (arrow) most likely result from sister chromatid exchanges. C Light optical section through a pre-S-phase nucleus with replication-labelled chromatid territories (green). D Subsequent light optical section from the same nuclear plane showing the two painted 15q territories (red; arrow and arrowhead). E Overlay of the images shown in C and D. The colocalization (yellow; arrow) of one of the two painted 15q territories with a replication-labelled territory unequivocally identifies this territory. (Note that the very small unpainted p arm of the acrocentric chromosome 15 is not present in this section). The arrowhead points to the homologous 15q territory identified by chromosome painting (D, E) that is not replication labelled in this case (C, E) reflecting the random segregation events of labelled and unlabelled sister chromatids (see text). F Confocal image of the IdU label within a 15q territory (in situ hybridization not shown). G Confocal image of CldU label obtained in the same nuclear plane as shown in F. H Overlay of the images depicted in F and G. Apparently, the IdU and the CldU label within the 15q territory localizes to distinct domains of approximately 400–800 nm in diameter that show only very few overlaps (indicated by the nearly complete absence of yellow-colored regions in H)

ing HDF cells were inspected by fluorescence microscopy. After 1 day, microinjected nuclei exhibited fluorescence that was distributed over the whole nucleus and concentrated in subchromosomal foci with diameters similar to those noted above for fixed cell nuclei (data not shown). No fluorescence was detectable within the cytoplasm. The fraction of nuclei with Cy3-specific fluorescence was the same as detected by immunocytochemical staining of S-phase cells fixed immediately after pulse labelling with BrdU and other thymidine analogs, suggesting that after 1 day only those nuclei still exhibited fluorescence that had stably incorporated Cy3-labelled nucleotides into their DNA, while non-incorporated nucleoFig. 2A–D Replication-labelled and segregated chromatids of human neuroblastoma cells (SH-EP N14; A, B) and diploid fibroblasts (C, D). Panel D shows BrdU labelled chromatid territories; the other panels depict Cy3-AP3-dUTP-labelled chromatids. Cy3AP3-dUTP-microinjected cells or BrdU-labelled cells were grown for several cell cycles. Panels A and D show fixed chromatids, while panels B and C display in vivo images. A Anaphase depicting two completely replication-labelled chromatids. The inset shows the DAPI counterstain. B Optical section (Cy3 detection) of an interphase nucleus from the same slide as the anaphase shown in A. The chromatid territories are composed of subdomains within a size range of approximately 400–800 nm. The inset shows an enlargement of one territory (arrowhead). C Video image (non-confocal) of several typical replication-labelled chromatid territories observed in a nucleus 1 week after microinjection of Cy3-AP3dUTP. D Optical section of a nucleus (only the upper right quadrant is shown) containing whole BrdU-labelled chromatid territories. The chromatid territories are compartmentalized (some typical subdomains are indicated by arrowheads)

246

Fig. 3A–G In vivo images of HeLa cells grown for several cell cycles after Cy3-AP3-dUTP microinjection. A–C Each panel shows a projection of confocal stacks (Cy3 detection) from one HeLa nucleus with replication-labelled chromatid territories. The nuclei were imaged every 15–20 min over a period of 240 min, and for each panel the same nucleus is shown at the beginning (top, 0 min) and the end (bottom, 240 min) of the time series. The framed territory shown in A kept its overall shape and subdomain structure although it moved for approximately 3 µm through the nucleus. Shape and position of the other territories within this nucleus show only slight changes, in common with the territories depicted in B. The arrow in C points to a territory that is shown in more detail in D–G. D Single optical section of the nucleus shown in C, 0 min. The subdomain structure of the two labelled territories is visible (arrowheads). E–G Enlarged projections of the right territory shown in C and D (arrow) at different indicated time points during the observation period. This territory exhibited a major change of shape that was mainly due to positional movements of the subdomains (arrowheads)

tides were lost. After several cell cycles, neuroblastoma cell nuclei (Figs. 2B, 4), HDF nuclei (Fig. 2C) and HeLa cell nuclei (Fig. 3) were observed in vivo with typical replication-labelled and segregated chromatid territories, indicating that cells with stably incorporated Cy3-labelled nucleotides continued to grow and divide. Figure 2A shows a fixed anaphase cell with two Cy3-AP3-dUTPlabelled chromatids directed to one nuclear pole. Light optical sections revealed that chromatid territories were composed of subchromosomal foci with diameters of approximately 400–800 nm (Figs. 2B, 3D). Neuroblastoma cell nuclei were rather flat compared to the considerably higher HeLa cell nuclei. Correspondingly, chromatid territories of the latter were more expanded in the axial direction. Cy3-AP3-dUTP-labelled and segregated chromatid territories in HeLa and neuroblastoma cell nuclei were subject to 4D (3D over time) analyses. In order to keep the possible interference of laser light microscopy with the dynamics of chromatid territories and movements of subchromosomal foci at a minimum, we used highly sensitive

optical detection systems and applied the lowest light levels compatible with a reasonable recording of light optical sections. Under these conditions we noted movements of the entire cell and of its nucleus, as well as dynamic changes of the position of chromatid territories and subchromosomal foci, respectively, during the whole observation period. Figures 3 and 4 show the dynamics of interphase chromatid territories during several hours of observation. We noted changes over distances of several micrometers in the positioning of chromatid territories relative to each other in some nuclei (Fig. 3A), while their position was more stable in other nuclei (Fig. 3B). We also noted pronounced changes in the relative positioning of subchromosomal foci within some territories (Fig. 3C, E–G), while other territories within the same nucleus revealed a more stable positioning of subchromosomal foci during the observation time. Individual subchromosomal foci revealed changing patterns of foldings and extensions (Fig. 4).

Discussion Evidence for the existence of chromosome territories and subchromosomal foci in the nucleus of living cells Exponentially growing human cell cultures were replication labelled with thymidine analogs. Halogenated thymidine analogs, such as bromodeoxyuridine (BrdU), chlorodeoxyuridine (CldU) and iododeoxyuridine (IdU) were added to the culture medium for studies of fixed cells, while the fluorescent thymidine analog Cy3-AP3-dUTP was microinjected into the cell nucleus for studies of living cells. Labelled cells were further grown in normal medium for up to 2 weeks (i.e. up to some 10 cell cycles). This scheme yielded sister chromatid labelling. During the second and subsequent mitotic events labelled and unlabelled chromatids were randomly segregated to daughter nuclei. Nu-

247

Fig. 4 Time-lapse recording of one nucleus from the neuroblastoma cell line. Cy3-AP3-dUTP-microinjected cells were grown for several cell cycles. Each panel shows a projection of optical sections. 3D stacks were recorded every 20 min for 5 h. The nucleus performed rotational and translational movements during the entire period. The framed segregated territory is clearly separated from the rest of the labelled territories. After a major initial rearrangement (indicated by large arrowheads) this territory maintained its overall shape. The inset (upper right) within the first six panels shows an enlargement (factor 2.3) of subdomains (arrow). The subdomains display changing patterns of foldings and extensions (some examples are indicated by arrowheads)

clei with individual labelled chromatid territories composed of subchromosomal foci with a diameter of ca. 400–800 nm could be demonstrated in nuclei of both fixed and living cells. While the immunocytochemical detection of chromatin labelled with halogenated thymidine analogs requires cell fixation and DNA denaturation, labelling with Cy3-AP3-dUTP is compatible with living cell microscopy.

Compartmentalization of territories generally appeared rather more distinct in nuclei of living cells than in nuclei of fixed and denatured cells, indicating some artificial dispersion of chromatin in the latter. While we expect that overall arrangements of chromosome territories, arm domains and possibly band domains can be successfully studied in fixed cell nuclei using multicolor FISH, this approach may turn out to be artifact prone and insufficient to study the 3D organization of individual chromosomal subcompartments (Robinett et al., 1996) and their topological relationships with other macromolecular domains. While studies of fluorescently labelled chromatid territories in vivo rule out artifacts of fixation and staining procedure and provide essential tools to explore the dynamics of chromatin movements, this new approach requires the consideration of other potential damage. We do not know at present to what extent the incorporation of fluorescently labelled nucleotides may impair the over-

248

all transcriptional activity of the in vivo-labelled chromatin. Since changes of the transcriptional activity of genes in labelled chromosome territories may correspond with changes of chromatin structure, it is important to address this point. Our observations provide circumstantial evidence that the cells containing the segregated, replication-labelled chromatid territories were functionally intact at the time we started our in vivo observations. When the progeny of microinjected diploid and aneuploid cells showing replication labeled and segregated chromatid territories (see below) were observed after 2–14 days (depending on the proliferation kinetics of these cells), their morphology was not detectably different from that of non-microinjected control cells growing in adjacent areas of the same chamber. The impairment of nuclear functions usually inhibits cell growth or leads to cell death. Our finding that in vivo-labelled cells not only survived for at least 2 weeks but kept cycling (a prerequisite for obtaining nuclei with segregated, labelled territories) suggests that these cells had retained their structural and functional integrity. Damaging effects of the absorbed laser light are a serious concern when laser confocal microscopy is used to study in vivo-labelled cells. Being aware of this problem, we tried to keep the possible exposure to laser light as low as possible using a highly sensitive optical detection system. Except for the brief periods (each less than 2 min) required to obtain stacks of light optical sections with the CLSM, the cells were strictly kept in darkness. Cells and nuclei continued to perform rotational and translational movements during the whole observation period (compare Fig. 4). Continued movements and changes in shape were also noted for subchromosomal foci even in cells where the observation period was extended to 5 h and optical sectioning was repeated 16 times. Selectivity and speed of territory and subdomain positional changes are in agreement with other recent findings concerning the dynamics of centromeric heterochromatin in interphase nuclei in vivo. Shelby et al. (1996) demonstrated that centromeric domains in living cells are positionally mostly stable and show only slow and selective movements in some cases. A quantitative analysis of the dynamics of chromatid territories and subchromosomal domains is presently under way in an attempt to discriminate brownian movements from non-random movements of chromatin. State of evidence for the structural persistence of subchromosomal foci Studies involving thymidine analogs have demonstrated the presence of replication foci (also called replication sites or replication granules) during S-phase with a diameter of approximately 400–800 nm (Nakayasu and Berezney 1989; Meng and Berezney 1991; Sparvoli et al. 1994; Berezney et al. 1995a, 1995b). It is well known that chromatin of the gene-dense R/T bands replicates first during S-phase, while chromatin of the gene-poor G/C bands replicates thereafter (Dutrillaux et al. 1976; Vogel et al. 1986; Craig and Bickmore 1993). Depending on the condensa-

tion state of mitotic chromosomes, chromatin with a distinct replication timing aggregates into subbands and bands (Drouin et al. 1991). In agreement with these findings, immunostaining of metaphase chromosome spreads prepared from HDF cultures which were double labelled with CldU and IdU according to the scheme described here yielded mitotic chromosomes with R/G banding patterns (our unpublished data). Several studies have demonstrated that pulse-labelled replication foci persist during subsequent cell cycle stages and cell cycles after labelling (Meng and Berezney 1991; Sparvoli et al. 1994; Berezney et al. 1995a, 1995b; present study). Berezney et al. (1995a, b) concluded “that each RS [i.e. replication site] contains an average of 5–10 replicons which are replicated in a relatively synchronous wave. The labeled DNA remains organized in ‘replication-like sites’ throughout the cell cycle and subsequent cell generations.” Although there is as yet no unequivocal proof, the experimental evidence described above suggests that subchromosomal foci observed in vivo are identical to replication-labelled persistent foci observed by Berezney and coworkers (1995a, b) and Sparvoli et al. (1994) in fixed cell nuclei. Accordingly, subchromosomal foci can be considered as persistent higher-order structural units. Our work demonstrates that these persistent structural units organize the in vivo substructure of chromosomes. It still has to be proven whether each individual subchromosomal focus is composed of the same DNA sequence during a series of subsequent cell cycles. Nevertheless, our double pulse labelling experiments demonstrate that at least the subdivision into subchromosomal foci containing either early- or late-replicating chromatin is stable. Possible functional significance of subchromosomal foci As discussed so far, the in vivo observations together with the double pulse labelling experiments indicated that subchromosomal foci are stably organized and are units of replicational regulation (indicated by their exclusive composition of chromatin replicating early or late during Sphase). Their persistence during cell cycle stages other than S-phase (Fig. 1F–H shows pre S-phase territories) suggests a functional significance of subchromosomal foci in addition to being specifically regulated units of replication. A relationship between transcriptional activity and replication timing of chromatin has been demonstrated (Goldman et al. 1984; Hatton et al. 1988; Riggs and Pfeiffer 1992). Accordingly, subchromosomal foci with a specific replication timing may also serve as spatially organized units of chromatin displaying a specific transcriptional regulation. We propose the following hypothesis: Chromosomes are built up by stable chromatin units, visible as 400- to 800-nm subchromosomal foci during interphase and bands or subbands during mitosis. The subchromosomal foci are structural/functional entities and units of replicational and transcriptional regulation. So far, the evidence for such a general functional significance of subchromosomal foci is only indirect.

249

Models of a functional nuclear architecture

Random walk/giant loop model

Numerous models have been proposed for a functional nuclear architecture (e.g. Blobel 1985; Chai and Sandberg 1988; Manuelidis 1990; Spector 1990; Cremer et al. 1993; Kramer et al. 1994; Wansink et al. 1993, 1994; Berezney 1995a). The territorial organization of interphase chromosomes requires certain topological constraints in the folding of the respective chromatin fibers. Otherwise each fiber would simply extend more or less throughout the entire nuclear volume and strongly intermingle with the fibers from other interphase chromosomes (G. Kreth, C. Münkel, C. Cremer, unpublished modelling data). The following discussion is restricted to models which are compatible with the evidence of a territorial chromosome organization. These models were chosen to exemplify the range of present ideas about interphase chromosome organization and are discussed in relation to the proposed structural/functional suborganization of chromosomes.

The random walk/giant loop (RW/GL) model (Sachs et al. 1995; Liu and Sachs 1997) predicts a highly flexible backbone to which giant loops, each comprising several Mbp of DNA, are attached. As emphasized by its name, the model assumes that both the backbone and the giant loops are folded according to a random-walk path with loop sizes adapted to fit measured interphase distances (Yokota et al. 1995, 1997). Several parameters in the random-walk model could be responsible for the regional differences in large-scale interphase chromosome structure observed by Yokota et al. (1997). The RW/GL model does not predict a highly ordered, functionally relevant organization of a chromosome territory.

Chromatin fiber models of interphase chromosome architecture based on electron-microscopic evidence Numerous studies have been carried out to explore the structure of chromatin at the electron-microscopic level (e.g. Woodcock and Horowitz 1995, 1997; for a review of the older but still relevant literature see Stubblefield 1973). Manuelidis and Chen (1990) and Belmont and Bruce (1994) have proposed models which try to explain the interphase and metaphase chromosome structure by a hierarchy of chromatin fibers. There appears to be general agreement regarding the existence of 10-nm and 25- to 30-nm chromatin fibers. The evidence for possible higherorder arrangements of 25- to 30-nm fibers into large scale chromatin structures, such as the 100- to 130-nm chromonema fibers described by Belmont and Bruce (1994) or the 240-nm fibers described by Manuelidis and Chen (1990), seems much less certain. Recently, Horowitz et al. (1994) have provided evidence that a asymmetric 3D zigzag of nucleosomes and linker DNA represents a basic chromatin folding in contrast to helical models. Manuelidis (1990) has stressed a functional correlation between interphase and metaphase chromosome structure. In her view different metaphase bands give rise to distinct functional domains in interphase chromosomes. Further studies are necessary to explore the relationship between the subchromosomal foci observed in the present in vivo study and these large-scale chromatin structures. One difficulty is that structures considerably smaller than 400 nm cannot be resolved by conventional and confocal light microscopy. Recent developments in 3D light microscopy which yield improved resolution (Stelzer et al. 1997) and allow high-precision distance measurements below 100 nm (Bornfleth et al. 1997), as well as electron-microscopic studies, will help to overcome this limitation.

Channel model of nuclear matrix structure In this model (Razin and Gromova 1995) the nuclear matrix is viewed as a system of channels which start in the interior of interphase chromosomes and extend to the nuclear pores. Active genes or replicon clusters are attached to these nuclear matrix channels, which provide a means for the export of mRNAs and the import of precursors for DNA and RNA synthesis, enzymes and regulatory proteins. While it is assumed that within interphase chromosome territories DNA loops are regularly attached to the matrix channels, no prediction is made with regard to a stable suborganization of interphase chromosome territories in the size range of light microscopically observable structures. Interchromosomal domain compartment model The interchromosomal domain compartment (ICD) model has been developed by one of the authors (T.C.) together with C. Cremer and P. Lichter as a working hypothesis for a functional nuclear architecture. This model predicts that nuclear functions, such as transcription, splicing, mRNA transport, DNA replication and repair, occur within a threedimensional ICD space connected to the nuclear pores. In a first version (Zirbel et al. 1993) it has been assumed that the ICD space extends only between the the exterior surfaces of chromosome territories, while a modified version of the ICD compartment model implies that branches of the channel network lead from the chromosome territory periphery into the chromosome territory interior and end between the surfaces of chomosomal subcompartments (Cremer et al. 1995, 1996). Future studies will show whether subchromosomal foci reflect the compartmentalized structure of chromosome territories predicted by the ICD model. Outlook The precise three-dimensional and (taking time as fourth dimension) four-dimensional topology between the various

250

structural components of the cell nucleus, such as chromosome territories and subchromosomal foci, nuclear matrix constituents and protein complexes involved in nuclear functions, is still unknown (for review see Cremer et al. 1995). Such knowledge is essential to test the topological predictions made by the channel model of Razin and Gromova (1995), the ICD compartment model or other models. To this end, in vivo studies of chromosome territories and subchromosomal foci can now be carried out during defined stages of the cell cycle or after the induction of terminal cellular differentiation events and combined with other approaches allowing the in vivo visualization of specific DNA sequences (Robinett et al. 1996; Shelby et al. 1996) and nuclear proteins (Htun et al. 1996; Misteli et al. 1997). Finally, it is of great interest in this context to investigate whether procedures for the detection of nascent RNA in fixed cell nuclei (Wansink et. al. 1993, 1994) can be modified to allow in vivo studies as well. Acknowledgements This work was supported by grants from the Deutsche Forschungsgemeinschaft (DFG; Cr 59/18-1) and the European Union (Biomed 2 program, BMH-4-CT95-1139) to T. Cremer and a DFG travelling grant (ZI 560/1-1) to D. Zink. We are grateful to Astrid Visser (University of Amsterdam) for her support in the development of combined replication labelling and chromosome painting protocols and to Jamie White (EMBL, Heidelberg) for help with living cell microscopy. We thank Jeffrey Trent (National Institute of Human Genome Research, NIH, Bethesda) for providing the 15q microdissection probe, W. W. Franke, Jörg Schürmann (DKFZ, Heidelberg) and Pascal Tomakidi (Kopfklinik, Universität Heidelberg) for providing cell lines, and Isabell Jentsch for help with graphic illustrations. We gratefully acknowledge a gift of Cy3-AP3-dUTP from Amersham International, Bucks., UK.

References Aten JA, Bakker PJM, Stap J, Boschmann GA, Veenhof CHN (1992) DNA double labelling with IdUrd and ClUrd for spatial and temporal analysis of cell proliferation and DNA replication. Histochem J 24:251–259 Aten JA, Stap J, Manders EMM, Bakker PJM (1993) Progression of DNA-replication in cell nuclei and changes in cell proliferation investigated by DNA double-labelling with IdUrd and CldUrd. Eur J Histochem 37:65–71 Belmont AS, Bruce K (1994) Visualization of G1 chromosomes: a folded, twisted, supercoiled chromonema model of interphase chromatid structure. J Cell Biol 127:287–302 Berezney R, Ma H, Meng C, Samarabandu J, Cheng P-C (1995a) Connecting genomic architecture and DNA replication in three dimensions. Zool Stud 34 [Suppl I]:29–32 Berezney R, Mortillaro M J, Ma H, Wei X, Samarabandu J (1995b) The nuclear matrix: a structural milieu for genomic function. Int Rev Cytol 162A:1–65 Blobel G (1985) Gene gating: a hypothesis. Proc Natl Acad Sci USA 82:8527–8529 Bornfleth H, Sätzler K, Eils R, Cremer C (1997) High precision distance measurements and volume-conserving segmentation of objects near and below the resolution limit in threedimensional confocal fluorescence microscopy. J Microsc (in press) Chai LS, Sandberg AA (1988) Chromosomes and their relationship to nuclear components during the cell cycle in Chinese hamster cells. Cell Tissue Res 25:197–204

Craig JM, Bickmore WA (1993) Chromosome bands – flavours to savour. Bioessays 15:349–354 Cremer C, Muenkel C, Granzow M, Jauch A, Dietzel S, Eils R, Guan X Y, Meltzer P S, Trent J M, Langowski J, Cremer T (1996) Nuclear architecture and the induction of chromosomal aberrations. Mutat Res 366:97–116 Cremer T, Lichter P, Borden J, Ward DC, Manuelidis L (1988) Detection of chromosome aberrations in metaphase and interphase tumor cells by in situ hybridization using chromosomespecific library probes. Hum Genet 80:235–246 Cremer T, Kurz A, Zirbel R, Dietzel S, Rinke B, Schröck E, Speicher MR, Mathieu U, Jauch A, Emmerich P, Scherthan H, Ried T, Cremer C, Lichter P (1993) The role of chromosome territories in the functional compartmentalization of the cell nucleus. Cold Spring Harbor Symp Quant Biol 58:777–792 Cremer T, Dietzel S, Eils R, Lichter P, Cremer C (1995) Chromosome territories, nuclear matrix filaments and inter-chromatin channels: a topological view on nuclear architecture and function. In: Bradham PE, Bennett MD (eds) Kew Chromosome Conference IV. Royal Botanic Gardens, Kew, London, pp 63– 81 De Boni U (1994) The interphase nucleus as a dynamic structure. Int Rev Cytol 150:149–171 Dietzel S, Jauch A, Kienle D, Qu G, Holtgreve-Grez H, Eils R, Muenkel C, Bittner M, Meltzer PS, Trent JM, Cremer T (1998) Separate and variably shaped chromosome arm domains are disclosed by chromosome arm painting in human cell nuclei. Chromosome Res (in press) Drouin R, Lemieux N, Richer CL (1991) Chromosome condensation from prophase to late metaphase: relationship to chromosome bands and their replication time. Cytogenet Cell Genet 57:91–99 Dutrillaux B, Couturier J, Richer C-L, Viegas-Peguinot E (1976) Sequence of DNA replication in 277 R- and Q-bands of human chromosomes using a BrdU treatment. Chromosoma 58:51–61 Eils R, Dietzel S, Bertin E, Granzow M, Schröck E, Speicher MR, Volm T, Ried T, Robert-Nicoud M, Cremer C, Cremer T (1996) Three-dimensional reconstruction of painted human interphase chromosomes: active and inactive X chromosome territories have similar volumes but differ in shape and surface structure. J Cell Biol 135:1427–1440 Ferreira J, Paolella G, Ramos C, Lamond AI (1997) Spatial organization of large-scale chromatin domains in the nucleus: a magnified view of single chromosome territories. J Cell Biol (in press) Goldman MA, Holmquist GP, Gray MC, Caston LA, Nag A (1984) Replication timing of genes and middle repetitive sequences. Science 224:686–692 Guan X-Y, Zhang H, Bittner M, Jiang Y, Meltzer P, Trent J (1996) Chromosome arm painting probes. Nat Genet 12:11–12 Hatton KS, Dhar V, Brown EH, Iqbal MA, Stuart S, Didamo VT, Schildkraut CL (1988) Replication program of active and inactive multigene families in mammalian cells. Mol Cell Biol 8:2149–2158 Hiraoka Y, Minden JS, Swedlow JR, Sedat JW, Agard DA (1989) Focal points for chromosome condensation and decondensation revealed by three-dimensional in vivo time-lapse microscopy. Nature 342:293–296 Hochstrasser M, Sedat JW (1987) Three-dimensional organization of Drosophila melanogaster interphase nuclei. I. Tissue-specific aspects of polytene nuclear architecture. J Cell Biol 104: 1455–1470 Hochstrasser M, Mathog D, Gruenbaum Y, Saumweber H, Sedat JW (1986) Spatial organization of chromosomes in the salivary gland nuclei of Drosophila melanogaster. J Cell Biol 102:112– 123 Horowitz RA, Agard DA, Sedat JW, Woodcock CL (1994) The three-dimensional architecture of chromatin in situ: electron tomography reveals fibers composed of a continuously variable zig-zag nucleosomal ribbon. J Cell Biol 125:1–10

251 Htun H, Barsony J, Renyi I, Gould DL, Hager GL (1996) Visualization of glucocorticoid receptor translocation and intranuclear organization in living cells with a green fluorescent protein chimera. Proc Natl Acad Sci USA 93:4845–4850 Kramer J, Zachar Z, Bingham PM (1994) Nuclear pre-mRNA metabolism: channels and tracks. Trends Cell Biol 4:35–37 Kurz A, Lampel S, Nickolenko JE, Bradl J, Benner A, Zirbel RM, Cremer T, Lichter P (1996) Active and inactive genes localize preferentially in the periphery of chromosome territories. J Cell Biol 135:1195–1202 Latt SA (1973) Microfluorometric detection of deoxyribonucleic acid replication in human metaphase chromosomes. Proc Natl Acad Sci USA 70 : 3395–3399 Leitch AR, Mosgoller W, Schwarzacher T, Bennett MD, HeslopHarrison JS (1990) Genomic in situ hybridization to sectioned nuclei shows chromosome domains in grass hybrids. J Cell Sci 95:335–341 Lichter P, Cremer T, Borden J, Manuelidis L, Ward DC (1988) Delineation of individual human chromosomes in metaphase and interphase cells by in situ suppression hybridization using recombinant DNA libraries. Hum Genet 80:224–234 Liu B, Sachs RK (1997) A two-backbone polymer model for interphase chromosome geometry. Bull Math Biol 59:325–337 Manuelidis L (1985) Individual interphase chromosome domains revealed by in situ hybridization. Hum Genet 71:288–293 Manuelidis L (1990) A view of interphase chromosomes. Science 250:1533–1540 Manuelidis L, Chen T L (1990) A Unified Model of Eukaryotic Chromosomes. Cytometry 11:8–25 Meng C, Berezney R (1991) Replication cluster domains persist throughout the cell cycle of mouse 3T3 cells. J Cell Biol 115: 95a Misteli T, Caceres JF, Spector DL (1997) The dynamics of a premRNA splicing factor in living cells. Nature 387:523–527 Nakayasu H, Berezney R (1989) Mapping Replicational Sites in the Eucaryotic Cell Nucleus. J Cell Biol 108:1–11 Pinkel D, Landegent J, Collins C, Fuscoe J, Segraves R, Lucas J, Gray JW (1988) Fluorescence in situ hybridization with chromosome-specific libraries: detection of trisomy 21 and translocations of chromosome 4. Proc Natl Acad Sci USA 85:9138– 9142 Razin SV, Gromova II (1995) The channels model of nuclear matrix structure. Bioessays 17:443–450 Riggs AD, Pfeiffer GP (1992) X-chromosome inactivation and cell memory. Trends Genet 8:169–174 Robinett CC, Straight A, Li G, Willhelm C, Sudlow G, Murray A, Belmont AS (1996) In vivo localization of DNA sequences and visualization of large-scale chromatin organization using Lac operator/repressor recognition. J Cell Biol 135:1685–1700 Ross RA, Spengler BA, Biedler JL (1983) Coordinate morphological and biochemical interconversion of human neuroblastoma cells. J Natl Cancer Inst 71:741–747 Roth MB (1995) Spheres, coiled bodies and nuclear bodies. Curr Opin Cell Biol 7:325–328 Sachs RK, van den Engh G, Trask B, Yokota H, Hearst JE (1995) A random-walk/giant-loop model for interphase chromosomes. Proc Natl Acad Sci USA 92:2710–2714 Schardin M, Cremer T, Hager HD, Lang M (1985) Specific staining of human chromosome position in Chinese hamster × man hybrid cell lines demonstrates interphase chromosome territories. Hum Genet 71:281–287

Schedl P, Grosveld (1995) Domains and Boundaries. In: Elgin CR (ed) Chromatin structure and gene expression. Oxford University Press, Oxford, pp 172–196 Shelby RD, Hahn KM, Sullivan KF (1996) Dynamic elastic behavior of alpha-satellite DNA domains visualized in situ in living human cells. J Cell Biol 135:545–557 Sparvoli E, Levi M, Rossi E (1994) Replication clusters may form structurally stable complexes of chromatin and chromosomes. J Cell Science 107:3097–3103 Spector DL (1990) Higher order nuclear organization: three-dimensional distribution of small nuclear ribonucleoprotein particles. Proc Natl Acad Sci USA 87:147–151 Spector DL (1993) Macromolecular domains within the cell nucleus. Annu Rev Cell Biol 9:265–315 Stelzer EHK, Lindek S (1994) Fundamental reduction of the observation volume in far-field light microscopy by detection orthogonal to the illumination axis: confocal theta microscopy. Optics Comm 111:536–547 Stubblefield E (1973) The structure of mammalian chromosomes. Int Rev Cytol 35:1–60 van Driel R, Wansink DG, van Steensel B, Grande MA, Schul W, de Jong L (1995) Nuclear domains and the nuclear matrix. Int Rev Cytol 162A:151–189 Vogel W, Authenried M, Speit G (1986) Detection of bromodeoxyuridine-incorporation in mammalian chromosomes by a bromodeoxyuridine-antibody. I. Demonstration of replication patterns. Hum Genet 72:129–132 Wansink DG, Schul W, van der Kraan I, van Steensel B, van Driel R, de Jong L (1993) Fluorescent labeling of nascent RNA reveals transcription by RNA polymerase II in domains scattered throughout the nucleus. J Cell Biol 122:283–293 Wansink DG, van Driel R, de Jong L (1994) Organization of (pre-) mRNA metabolism in the cell nucleus. Mol Biol Rep 20:45–55 Wenzel A, Cziepluch C, Hamann U, Schuermann J, Schwab M (1991) The N-Myc oncoprotein is associated in vivo with the phosphoprotein Max(p20/22) in human neuroblastoma cells. EMBO J 10:3703–3712 Woodcock CL, Horowitz RA (1995) Trends Cell Biol 5:272–277 Woodcock CL, Horowitz RA (1997) Electron microscopy of chromatin. Methods 12:84–95 Xing Y, Johnson C V, Moen PT Jr, McNeil JA, Lawrence J (1995) Nonrandom gene organization: structural arrangements of specific pre-mRNA transcription and splicing with SC-35 domains. J Cell Biol 131:1635–1647 Yokota H, van den Engh G, Hearst JE, Sachs RK, Trask B (1995) Evidence for the organization of chromatin in megabase pairsized loops arranged along a random walk path in the human G0/G1 interphase nucleus. J Cell Biol 130:1239–1249 Zhao K, Hart CM, Laemmli UK (1995) Visualization of chromosomal domains with boundary element associated factor beaf32. Cell 81:879–889 Zirbel RM, Mathieu UR, Kurz A, Cremer T, Lichter P (1993) Evidence for a nuclear compartment of transcription and splicing located at chromosome domain boundaries. Chromosome Res 1:93–106