Published in Nature Reviews Genetics vol. 2, no. 4, pp. 292-301 (April, 2001):

"Chromosome Territories, Nuclear Architecture and Gene Regulation in Mammalian Cells"

Thomas Cremer 1 and Christoph Cremer 2.

1 Faculty of Biology, Ludwig Maximilians University, Munich, Germany,  and
2 Faculty of Physics, Ruprecht Karls University, Heidelberg, Germany


Conclusions:
Figures 1-7 with Legends:
1. Model of Functional Nuclear Architecture:
2. Chromosome Territories in the Chicken:
3. Features of Human Chromosome Territories:
4. Gene-Rich and Gene-Poor Chromosome Territories:
5. Early- and Mid-to-Late-Replicating Chromatin Domains:
6. The Interchromatin compartment:
7. The Multiloop Subcompartment Model:
Additional References:
Other Sites:
Information and Feedback:

Conclusions:
 
  • Chromosomes occupy discrete territories in the cell nucleus and contain distinct chromosome-arm and chromosome-band domains.

    •  
  • Chromosome territories (CTs) with different gene densities occupy distinct nuclear positions.

    •  
  • Gene-poor, mid-to-late-replicating chromatin is enriched in nuclear compartments that are located at the nuclear periphery and at the perinucleolar region.

    •  
  • A compartment for gene-dense, early-replicating chromatin is separated from the compartments for mid-to-late-replicating chromatin.

    •  
  • Chromatin domains with a DNA content of  ~1 Mb can be detected in nuclei during interphase and in non-cycling cells.

    •  
  • The interchromatin compartment (IC) contains various types of non-chromatin domains with factors for transcription, splicing, DNA replication and repair.

    •  
  • The CT-IC model predicts that a specific topological relationship between the IC and chromatin domains is essential for gene regulation.

    •  
  • The transcriptional status of genes correlates with gene positioning in CTs.

    •  
  • A dynamic repositioning of genes with respect to centromeric heterochromatin has a role in gene silencing and activation.

    •  
  • Various computer models of CTs and nuclear architecture make different predictions that can be validated by experimental tests.

    •  
  • Comprehensive understanding of gene regulation requires much more detailed knowledge of gene expression in the context of nuclear architecture and organization.

    • Conclusions:
    Figures 1-7 with Legends:
    1. Model of Functional Nuclear Architecture:
    2. Chromosome Territories in the Chicken:
    3. Features of Human Chromosome Territories:
    4. Gene-Rich and Gene-Poor Chromosome Territories:
    5. Early- and Mid-to-Late-Replicating Chromatin Domains:
    6. The Interchromatin compartment:
    7. The Multiloop Subcompartment Model:
    Additional References:
    Other Sites:
    Information and Feedback:



    Figure 1: Model of functional nuclear architecture.  Structural features that support the
    chromosome-territory–interchromatin-compartment (CT–IC) model are shown. These features are drawn roughly to scale on an optical section taken from the nucleus of a living HeLa cell. Although experimental evidence is available to support these features, the overall model of functional nuclear architecture is speculative (see text).

    a: CTs have complex folded surfaces. Inset: topological model of gene regulation23. A giant chromatin loop with several active genes (red) expands from the CT surface into the IC space.

    b: CTs contain separate arm domains for the short (p) and long chromosome arms (q), and a centromeric domain (asterisks). Inset: topological model of gene regulation78, 79. Top, actively transcribed genes (white) are located on a chromatin loop that is remote from centromeric heterochromatin. Bottom, recruitment of the same genes (black) to the centromeric heterochromatin leads to their silencing.

    c: CTs have variable chromatin density (dark brown, high density; light yellow, low density). Loose chromatin expands into the IC, whereas the most dense chromatin is remote from the IC.

    d: CT showing early-replicating chromatin domains (green) and mid-to-late-replicating chromatin domains (red). Each domain comprises 1 Mb. Gene-poor chromatin (red), is preferentially located at the nuclear periphery and in close contact with the nuclear lamina (yellow), as well as with infoldings of the lamina and around the nucleolus (nu). Gene-rich chromatin (green) is located between the gene-poor compartments.

    e: Higher-order chromatin structures built up from a hierarchy of chromatin fibres88. Inset: this topological view of gene regulation27, 68 indicates that active genes (white dots) are at the surface of convoluted chromatin fibres. Silenced genes (black dots) may be located towards the interior of the chromatin structure.

    f : The CT–IC model predicts that the IC (green) contains complexes (orange dots) and larger non-chromatin domains (aggregations of orange dots) for transcription, splicing, DNA replication and repair.

    g: CT with 1-Mb chromatin domains (red) and IC (green) expanding between these domains. Inset: the topological relationships between the IC, and active and inactive genes72. The finest branches of the IC end between 100-kb chromatin domains. Top: active genes (white dots) are located at the surface of these domains, whereas silenced genes (black dots) are located in the interior. Bottom: alternatively, closed 100-kb chromatin domains with silenced genes are transformed into an open configuration before transcriptional activation. 




    Figure 2: Chromosome territories in the chicken.

    a: 4,6-diamidino-2-phenylindole (DAPI)-stained, diploid, chicken metaphase spread with macro- and microchromosomes.

    b: The same metaphase spread after multicolour fluorescence in situ hybridization with pseudocoloured chromosomes. Chicken chromosome paint probes (image courtesy of Johannes Wienberg) were labelled by a combinatorial scheme with oestradiol (1, 4, 5, 6), digoxigenin (2, 4, 6, Z) and biotin (3, 5, 6, Z).

    c: Oestradiol- and digoxigenin-labelled probes were detected using secondary antibodies labelled with Cy3 and fluorescein isothiocyanate (FITC); biotinylated probes were detected with Cy5-conjugated streptavidin.

    d: Mid-plane light optical section through a chicken fibroblast nucleus shows mutually exclusive chromosome territories (CTs) with homologous chromosomes seen in separate locations. (Note that only one of the two CTs for each of 4 and 6 is displayed in this section.) (Image courtesy of F. Habermann.)




    Figure 3: Features of human chromosome territories.

    a: Two-colour painting of the p-arm (red) and the q-arm (green) of human chromosome 1 in a lymphocyte metaphase spread.

    b: Visualization of the two arms in a light optical section through a human diploid fibroblast nucleus (bottom) shows two distinct, mutually exclusive arm domains20. ( Image courtesy of Steffen Dietzel).

    c: Painting of the human X chromosome (red) and several distal bands of its p-arm and q-arm (green) using MICRODISSECTION PROBES20.

    d: Visualization of the active and inactive X-chromosome territories (Xa and Xi, respectively) together with the respective distal-band domains in a light optical section through a female human fibroblast nucleus. (Image courtesy of Joachim Karpf and Irina Solovei).

    e: Three-dimensional reconstructions of the Xa and Xi territories from a human female fibroblast nucleus (Reproduced with permission from Ref. 22). The three-dimensional positions of the ANT2 and ANT3 (adenosine nucleotide translocase) genes are noted as green and blue spheres, respectively. Note that active ANT genes can be seen at the territory surface (two on Xa and one on Xi). The white box provides a transparent view of the Xi territory (pink), indicating the location of the inactive ANT2 gene in the territory interior.

    f: Three-dimensonal reconstructions of two chromosome-17 territories, established from light optical serial sections through a human diploid fibroblast nucleus, show complex territory surfaces. (Image courtesy of Irina Solovei.)




    Figure 4: Gene-rich and gene-poor chromosome territories.  Three-dimensional reconstructions of chromosome 18 (red; gene-poor) and 19 (green; gene-rich) territories painted in the nucleus of  a non-stimulated human lymphocyte. (Image courtesy of Marion Cremer and Irina Solovei.) Chromosome 18 territories were typically found at the nuclear periphery, whereas chromosome 19 territories were located in the nuclear interior42.

    a: X,Y view: a mid-plane section of the nucleus is shown as a grey shade. Only the parts of the territories below this section can be seen.

    b: X,Z view: the arrow marks the side from which the section in part a is viewed.




    Figure 5: Early- and mid-to-late-replicating chromatin domains.

    a: Mid-plane light optical section through the nucleus of an SH-EP N14 neuroblastoma cell fixed 20 h after direct two-colour labelling of DNA with Cy3- and Cy5-conjugated nucleotides at early and mid-S-phase, respectively, shows typical early- (blue) and mid-replicated chromatin (red)115. The cell shown is a daughter of the labelled cell produced after one cell division. The experiment shows that the arrangement of early- and mid-replicating chromatin domains is maintained from one cell cycle to the next.

    b: Light optical mid-section through the nucleus of an SH-EP N14 cell fixed three days after two-colour DNA labelling at early- and mid-S-phase shows several chromosome territories (CTs) with typical early-(blue) and mid-(red) replicated 1-Mb chromatin domains115. Only a minority of CTs is labelled, which indicates that fixation was carried out after at least three post-labelling cell cycles. Note that the topology of mid-replicating chromatin (at the nuclear periphery and around the nucleoli) and early-replicating chromatin (in the interior nuclear compartment) was maintained through several post-labelling cell cycles.

    c, d: Mid-plane light optical section through the same cell nucleus shown in a immunostained with lamin B (green). A comparison of c and d shows that mid-replicating 1-Mb chromatin domains (c, red) are closely associated with the lamina, in contrast to the early-replicating domains (d, blue).
    (Images courtesy of Lothar Schermelleh.) (Adapted with permission from Ref. 115.)




    Figure 6: The interchromatin compartment.  LASER CONFOCAL sections through a HeLa cell nucleus with green-fluorescent-protein (GFP)-tagged H2B (cells kindly supplied by Ken Sullivan)56, and staining of speckles (image courtesy of Irina Solovei).

    a: Section showing GFP-tagged chromatin (high density, white; low density, grey), two nucleoli (nu) and the interchromatin compartment (IC) space (black). Note the variability in the width of this space with examples of IC lacunas (asterisks). The inset shows expansions of less-condensed chromatin into the IC space at higher magnification.

    b: Speckles visualized in the same section using antibodies to the non-snRNP splicing factor SC-35.

    c: Overlay of sections (chromatin, green; speckles, red) shows that speckles form clusters in IC lacunas. These lacunas are only partially filled by the speckles, leaving space for other non-chromatin domains.




    Figure 7: The multiloop subcompartment model.

    a: Two 1-Mb chromatin domains or 'subcompartments' are shown linked by a chromatin fibre (redrawn with permission from Ref. 72). Each 1-Mb chromatin domain is built up as a rosette of looped 100-kb chromatin fibres. At the centre, loops are held together by a magnified LOOP BASE SPRING, which simulates the function of CHROMOSOME TERRITORY ANCHOR PROTEINS 116.

    b, c: Two three-dimensional models of the internal ultrastructure of a 1-Mb chromatin domain (image courtesy of Gregor Kreth; models redrawn with permission from Ref. 12 © Begell House, Inc. (2000)).

    b: The nucleosome chain is compacted into a 30-nm chromatin fibre (visualized by cylinder segments) and folded into ten 100-kb-sized loop domains according to the multiloop subcompartment model. Occasionally, 30-nm fibres are interrupted by short regions of individual nucleosomes (small white dots). The arrow points to a red sphere, with a diameter of 30 nm, that represents a transcription factor complex.

    c: Each of the ten 100-kb chromatin domains was modelled under the assumption of a restricted random walk (zig-zag) nucleosome chain. Each dot represents an individual nucleosome. Nine 100-kb chromatin domains are shown in a closed configuration and one in an open chromatin configuration with a relaxed chain structure that expands at the periphery of the 1-Mb domain. The open domain will have enhanced accessibility to partial transcription complexes preformed in the interchromatin compartment. By contrast, most of the chromatin in the nine closed domains remains inaccessible to larger factor complexes (arrows).


    Additional References:

    1. "Ultrastructural Continuity between Active and Repressed Chromatin".

    2. "Ultrastructural Probes of DNA Templates within Human Bone Marrow and Lymph Node Cells".

    3. "Selective Control of DNA Helix Openings During Gene Regulation".


    Top of Page - Euchromatin Network - Current Research - Forums - Other Sites - Future Events -

    For Further Information and Feedback:
    E-mail:  frenster@euchromatin.net
    Phone:  +1 650 367 6483
    Fax:  +1 650 364 1773

    euchromatin: "the most active portion of the genome within the cell nucleus".