Chromosomes and Higher-Order Chromatin Structure

Chromatin becomes highly condensed during mitosis to form the compact metaphase chromosomes that are distributed to daughter nuclei (see Figure 4.12). During interphase, some of the chromatin (heterochromatin) remains highly condensed and is transcriptionally inactive; the remainder of the chromatin (euchromatin) is decondensed and distributed throughout the nucleus (Figure 8.15). Cells contain two types of heterochromatin. Constitutive heterochromatin contains DNA sequences that are never transcribed, such as the satellite sequences present at centromeres. Facultative heterochromatin contains sequences that are not transcribed in the cell being examined, but are transcribed in other cell types. Consequently, the amount of facultative heterochromatin varies depending on the transcriptional activity of the cell. Much of the heterochromatin is localized to the periphery of the nucleus, possibly because one of the principal proteins associated with heterochromatin binds to a protein of the inner nuclear membrane.

Figure 8.15

Heterochromatin in interphase nuclei. The euchromatin is distributed throughout the nucleus. The hetero-chromatin is indicated by arrowheads, and the nucleolus by an arrow. (Courtesy of Ada L. Olins and Donald E. Olins, Oak Ridge National Laboratory.) (more...)

The phenomenon of X chromosome inactivation provides an example of the role of heterochromatin in gene expression. In many animals, including humans, females have two X chromosomes, and males have one X and one Y chromosome. The X chromosome contains thousands of genes that are not present on the much smaller Y chromosome (see Figure 4.26). Thus, females have twice as many X chromosome genes as males have. Despite this difference, female and male cells contain equal amounts of the proteins encoded by X chromosome genes. This results from a dosage compensation mechanism in which one of the two X chromosomes in female cells is inactivated by being converted to heterochromatin early in development. Consequently, only one copy of the X chromosome is available for transcription in either female or male cells. The mechanism of X chromosome inactivation is fascinating though not yet fully understood; it appears to involve the action of a regulatory RNA that coats the inactive X chromosome and induces its conversion to heterochromatin.

Although interphase chromatin appears to be uniformly distributed, the chromosomes are actually arranged in an organized fashion and divided into discrete functional domains that play an important role in regulating gene expression. The nonrandom distribution of chromatin within the interphase nucleus was first suggested in 1885 by C. Rabl, who proposed that each chromosome occupies a distinct territory, with centromeres and telomeres attached to opposite sides of the nuclear envelope (Figure 8.16). This basic model of chromosome organization was confirmed nearly a hundred years later (in 1984) by detailed studies of polytene chromosomes in Drosophila salivary glands. Rather than randomly winding around one another, each chromosome was found to occupy a discrete region of the nucleus (Figure 8.17). The chromosomes are closely associated with the nuclear envelope at many sites, with their centromeres and telomeres clustered at opposite poles.

Figure 8.16

Chromosome organization. Reproduction of hand-drawn sketches of chromosomes in salamander cells. (A) Complete chromosomes. (B) Telomeres only (located at the nuclear membrane). (From C. Rabl, 1885. Morphologisches Jahrbuch 10: 214.)

Figure 8.17

Organization of Drosophila chromosomes. (A) A model of the nucleus, showing the five chromosome arms in different colors. The positions of telomeres and centromeres are indicated. (B) The two arms of chromosome 3 are shown to illustrate the topological (more...)

Individual chromosomes also occupy distinct territories within the nuclei of mammalian cells (Figure 8.18). Actively transcribed genes appear to be localized to the periphery of these territories, adjacent to channels separating the chromosomes. Newly transcribed RNAs are thought to be released into these channels between chromosomes, where RNA processing takes place.

Figure 8.18

Organization of chromosomes in the mammalian nucleus. (A) Probes to repeated sequences on chromosome 4 were hybridized to a human cell. The two copies of chromosome 4, identified by yellow fluorescence, occupy distinct territories in the nucleus. (B) (more...)

Like the DNA in metaphase chromosomes (see Figure 4.13), the chromatin in interphase nuclei appears to be organized into looped domains containing approximately 50 to 100 kb of DNA. A good example of this looped-domain organization is provided by the highly transcribed chromosomes of amphibian oocytes, in which actively transcribed regions of DNA can be visualized as extended loops of decondensed chromatin (Figure 8.19). These chromatin domains appear to represent discrete functional units, which independently regulate gene expression.

Figure 8.19

Looped chromatin domains. Light micrograph of a chromosome of amphibian oocytes, showing decondensed loops of actively transcribed chromatin extending from an axis of highly condensed nontranscribed chromatin. (Courtesy of Joseph Gall, Carnegie Institute.) (more...)

The effects of chromosome organization on gene expression have been demonstrated by a variety of experiments showing that the position of a gene in chromosomal DNA affects the level at which the gene is expressed. For example, the transcriptional activity of genes introduced into transgenic mice depends on their sites of integration in the mouse genome. This effect of chromosomal position on gene expression can be alleviated by sequences known as locus control regions, which result in a high level of expression of the introduced genes irrespective of their site of integration. In contrast to transcriptional enhancers (see Chapter 6), locus control regions stimulate only transfected genes that are integrated into chromosomal DNA; they do not affect the expression of unintegrated plasmid DNAs in transient assays. In addition, rather than affecting individual promoters, locus control regions appear to activate large chromosome domains, presumably by inducing long-range alterations in chromatin structure.

The separation between chromosomal domains is maintained by boundary sequences or insulator elements, which prevent the chromatin structure of one domain from spreading to its neighbors. In addition, insulators act as barriers that prevent enhancers in one domain from acting on promoters located in an adjacent domain. Like locus control regions, insulators function only in the context of chromosomal DNA, suggesting that they regulate higher-order chromatin structure. Although the mechanisms of action of locus control regions and insulators remain to be elucidated, their functions clearly indicate the importance of higher-order chromatin organization in the control of eukaryotic gene expression.

Functional Domains within the Nucleus

An internal organization of the nucleus is indicated also by the localization of some nuclear processes to distinct regions of the nucleus. Rather than taking place throughout the nucleus, activities such as DNA replication and pre-mRNA processing may be localized to discrete subnuclear structures or domains. The nature and function of these nuclear substructures are not yet clear, however, and understanding the organization of functional domains within the nucleus is an incompletely explored area of cell biology.

The nuclei of mammalian cells appear to contain clustered sites of DNA replication within which the replication of multiple DNA molecules takes place. These discrete sites of DNA replication have been defined by experiments in which newly synthesized DNA was visualized within cell nuclei (Figure 8.20). This was accomplished by labeling cells with bromodeoxyuridine—an analog of thymidine that can be incorporated into DNA and then detected by staining with fluorescent antibodies. In such experiments, the newly replicated DNA was detected in approximately 200 discrete clusters distributed throughout the nucleus. Since approximately 4000 origins of replication are active in a diploid mammalian cell at any given time, each of these clustered sites of DNA replication must contain approximately 40 replication forks. Thus, DNA replication appears to take place in large structures that contain multiple replication complexes organized into distinct functional domains, which have been called replication factories.

Figure 8.20

Clustered sites of DNA replication. Newly replicated DNA was labeled by a brief exposure of cells to bromodeoxyuridine, which is incorporated into DNA in place of thymidine. This substitution allows detection of newly synthesized DNA by immunofluorescence (more...)

Actively transcribed genes appear to be distributed throughout the nucleus, but components of the splicing machinery are concentrated in discrete subnuclear structural domains. The localization of splicing components to discrete domains within the nucleus has been demonstrated by immunofluorescent staining with antibodies against snRNPs and splicing factors (Figure 8.21). Rather than being distributed uniformly throughout the nucleus, these components of the splicing apparatus are concentrated in 20 to 50 discrete structures termed nuclear speckles. It is thought that speckles are storage sites of splicing components, which are then recruited from the speckles to actively transcribed genes where pre-mRNA processing occurs.

Figure 8.21

Localization of splicing components. Staining with immunofluorescent antibodies indicates that splicing factors are concentrated in discrete domains within the nucleus. (Courtesy of David L. Spector, Cold Spring Harbor Laboratory.)

In addition to speckles, nuclei contain several other types of morphologically distinct structures, collectively called nuclear bodies. The two major types of these nuclear structures are coiled bodies and PML bodies. Coiled bodies are enriched in snRNPs and are believed to function as sites of snRNP assembly. The function of PML bodies is unknown; they are not enriched in snRNPs and do not appear to be major sites of transcription or DNA replication. Thus, although nuclear bodies indicate the presence of substructural domains within the nucleus, their functions remain to be fully elucidated.

Some scientists believe that a central component of the internal organization of the nucleus is the nuclear matrix, which is defined as the structural skeleton of the nucleus. The concept of the nuclear matrix was proposed in 1975 on the basis of experiments in which nuclei were treated with DNase to digest most of the DNA and extracted with high salt buffers to remove histones and other nuclear proteins. Such treatments left a residual network of fibers (the nuclear matrix), which was suggested to provide an internal structural framework for the nucleus, analogous to the role of the cytoskeleton as the structural framework of the cell. In addition, it has been proposed that the nuclear matrix serves to organize and anchor functional domains of the nucleus, including chromatin loops, DNA replication factories, splicing domains, and structures involved in mRNA transport.

Many scientists, however, are not convinced that the nuclear matrix exists in the living cell. They believe that the matrix observed after extraction of nuclei may be the result of artificial aggregation of proteins and nucleic acids during preparation. Although the nuclear matrix is visualized after extraction of nuclei by several different methods, its molecular composition has not been clearly defined. In the absence of definitive characterization of its structural components, the nuclear matrix remains an area of dispute among cell biologists.

Box

Molecular Medicine: Systemic Lupus Erythematosus.