The nucleus is the center for genetic inheritance and gene expression within the cells of eukaryotes. It contains a myriad of reactions that occur within several macromolecular compartments and particles. This article provides an overview of nuclear structure and function. Many of the more specific topics are treated in more detail in their entries.
1. Genomic Inheritance
The nucleus is the most conspicuous organelle within eukaryotic cells. Although earlier microscopists made note of what must have been nuclei(1), historians attribute the discovery of the nucleus to Robert Brown in 1833 because of his careful descriptions of orchid cell nuclei(2). Once the nucleus was firmly established by Brown as a constant entity within cells, Hertwig and Strasburger in 1884 independently described the nucleus as the organelle of cellular inheritance. These and other late nineteenth and early twentieth century researchers theorized that inheritance is based on segregation of the nuclear chromosomes (3, 4).
Most eukaryotic cells contain one nucleus, but there are many examples of multinucleated cells or syncytia (eg, differentiated muscle cells, the early Drosophila embryo). Conversely, mature mammalian red blood cells lack nuclei. Nuclei differ in size depending on the cell type: the yeast (Saccharomyces cerevisiae) nucleus is approximately 2 pm in diameter, a typical mammalian nucleus is 4 to 6 pm, and nuclei (germinal vesicles) in mature amphibian oocytes are about 600 pm in diameter. Most nuclei are spherical, but multilobed nuclei are common, such as those found in polymorphonuclear leukocytes or mammalian epididymal cells. The nucleus is bordered by a double-membrane nuclear envelope that contains nuclear pore complexes for macromolecular import and export from and to the cytoplasm.
The nucleus is the repository for the complex eukaryotic genome. It normally consists of unique, middle-repetitive, and highly repetitive sequences, defined by renaturation kinetics(5) (see t curve). Generally, unique DNA includes single-copy genes and unique but nontranscribed sequences. Middle-repetitive DNA contains genes with multiple copies, such as those found in the tandem arrays of repeated genes for histones or the 47S RNA precursor for ribosomes and the spacer sequences that separate the individual genes within these clusters. Highly repetitive sequences include relatively short DNA elements that constitute, for example, the centromeres of chromosomes. Highly repetitive, noncentromeric sequences have also been detected. Examples include the 5S ribosomal genes (over 20,000 copies per haploid Xenopus genome), the L1 transposable elements that constitute about 4% of primate genomes, and the Alu pseudogene/transposable elements that constitute about 5% of the human genome .
The eukaryotic cell must efficiently package all of its genomic DNA (which in human cells has a combined linear length of more than one meter) within a nucleus with a diameter of only 5 pm. Yet the manner in which the DNA is packaged must also accommodate the complex processes of DNA replication, recombination, DNA repair, and transcription . Chromatin is the term used to described the packaged genome. It consists of genomic DNA, a nearly equal mass of histone proteins, a relatively small amount of non-histone chromosomal protein, and some nascent RNA. The most basic structural unit of chromatin is the nucleosome, which consists of approximately 146 bp of genomic DNA wrapped around a complex of eight histone molecules. In the presence of histone H nucleosomes condense to form a solenoid with a diameter of 30 nm and six nucleosomes per turn. The solenoid is cast into looped domains that are attached to the interphase nuclear matrix by adenine/thymine-rich DNA elements called matrix attachment regions (MARS). Such compacted chromatin is then organized into one or more chromosomes. The complexity of chromosomal structure within the nucleus is well illustrated by the lampbrush chromosomes of the nucleus of the amphibian oocyte (Fig. 1).
Figure 1. A phase contrast image of portions of lampbrush chromosomes isolated from an oocyte nucleus of the North American newt, Notophthalmus viridescens. The chromosomes were centrifuged onto a microscope slide and fixed with formaldehyde. Lateral loops of transcriptionally active DNA display nascent ribonucleoprotein matrices (arrow heads). Multiple nucleoli (arrows) within the oocyte nucleus produce enormous numbers of ribosomes. These particular nucleoli have partially fragmented during preparation. Many other small ribonucleoprotein particles are clearly evident (see text).
Ultrastructural examination of glutaraldehyde-fixed interphase cells reveals patches of condensed chromatin within their nuclei, usually just beneath the nuclear envelope or near the nucleolus. The term ‘heterochromatin’ was defined by E. Heitz in 1928 to describe chromatin that remains condensed when the chromosomes normally decondense in the late telophase and early interphase of the cell cycle. Because of its condensed state during the interphase, this heterochromatin is generally believed to be transcriptionally silent. Different cell types contain different amounts of heterochromatin. Metabolically active cancer cells, for example, display very little condensed chromatin, whereas orthochromatic erythroblasts (red blood cells that will soon lose their nuclei) contain large amounts of condensed chromatin(7). There are actually two types of heterochromatin. Constitutive heterochromatin remains condensed throughout the cell cycle and development. This chromatin contains highly repetitive sequences that play a structural role in chromosomal structure and mechanics (ie, centromeric DNA functioning in chromosomal movement during mitosis) rather than actual gene expression. Facultative heterochromatin has the potential for gene expression at some point in development, and it can be either condensed or decondensed, depending on the cell type. The mammalian Barr body is a excellent example of facultative heterochromatin. The Ban-body is one of the two X-chromosomes in mammalian female cells. It remains condensed during the interphase and is almost completely silent in transcription(8), whereas the other X-chromosome in the cell is decondensed and transcriptionally very active. Euchromatin is the term used to describe decondensed interphase chromatin that is either transcriptionally active or has the potential for transcription.
Replication of eukaryotic chromosomes requires elaborate DNA polymerase complexes that replicate the entire genome within the physical confines of the nucleus (see DNA replication). To replicate the enormous lengths of eukaryotic chromosomes, several autonomously replicating sequences (ARS), situated along the length of each chromosome, initiate replication in both directions along the DNA, thus leading to rapid replication of the entire chromosome. Eukaryotic genomic replication occurs during the interphase of the cell cycle, specifically during the S-phase between growth phases 1 and 2 (Gj and G2). The precisely timed onset of DNA replication is controlled by elaborate signal transduction cascades that work in conjunction with cyclin-dependent kinases (see Cell Cycle). Enhanced production of new histone proteins is concomitant with genomic DNA replication to ensure the timely packaging of the nascent DNA. Telomeres are unique structures at the very ends of chromosomes that provide for completion of their replication(9) and also interact with the nuclear envelope during meiosis (10, 11), perhaps to help align the chromosomes for recombination.
Fully replicated chromosomes condense during mitosis in preparation for orderly segregation into progeny cells. Each mitotic chromosome consists of two sister chromatids held together by their centromeres. Centromeres play a structural/mechanical function in chromatid segregation during mitosis and meiosis. They are sites for peripheral kinetochoric assembly during the prophase, one kinetochore for each chromatid. A set of microtubules from the mitotic spindle attaches to the kinetochores to align the chromosomes on the metaphase plate. At anaphase, the centromeric connection between the chromatids splits, and the kinetochoric microtubules depolymerize to "pull" the separated chromatids (now called chromosomes) toward opposite poles of the spindle.
2. Genomic Expression
Transcription is the initial step in gene expression in the nucleus. Transcription units (TU) are defined as thin-to-thick ribonucleoprotein (RNP) fibrils that remain attached to active RNA polymerase complexes on genes as viewed by electron microscopy of spread chromatin preparations (13) or of the transcriptionally active loops of lampbrush chromosomes (14; Fig. 1). The nascent RNP fibrils are shortest near transcription start sites, and they are longest near transcription stop sites (the so-called "Christmas tree"). The pre-messenger RNA resulting from transcription of genes that encode proteins is also known as heterogeneous nuclear RNA (hnRNA). Almost immediately after its synthesis by RNA polymerase II, the 5′ ends of the nascent hnRNA are capped with a 7-methyl guanine nucleotide (see 5′-Cap ). The cap is important for transporting the mature mRNA from the nucleus, for protecting of the 5′ end from exonucleases, and for recognizing the mRNA by translational initiation factors within the cytoplasm. Polyadenylation of pre-mRNA also occurs in the nucleus before mRNA export, and is generally believed to stabilize the transcripts in the cytoplasm.
The ribosomal RNA (rRNA) genes are transcribed within the nucleolus. The precursor rRNA is processed to yield the mature 18S, 5.8S, and 28S rRNA found in cytoplasmic ribosomes.
2.1. Spliceosomes and Heterogeneous Nuclear RNA Processing
One of the most important aspects of eukaryotic gene expression, other than the transcription of the gene itself, is the precise splicing of premessenger RNA (pre-mRNA) transcripts, first to remove nonprotein coding introns and then to join (ligate) the protein-coding exons (see RNA Splicing). Both events must occur precisely to maintain the proper protein translational reading frame. Sequences within the nascent transcript direct the assembly of the nuclear splicing machinery. Nuclear ribonucleoprotein (RNP) particles that mediate the transesterification and ligation reactions of splicing are called spliceosomes. They consist of five different small nuclear RNP particles (snRNP). Each snRNP consists of a unique small nuclear RNA (snRNA), either U1, U2, U4, U5, or U6. The snRNP particle is referred to by its constituent snRNA (eg, the U1 snRNP). In addition to the snRNA, each snRNP consists of proteins that are either unique to the specific snRNP particle (eg, the 70-kDa U1-specific protein) or common to the different snRNP (ie, the snRNP core proteins). The common proteins initiate snRNP complex formation upon recognizing a single-stranded region of the snRNA at the Sm site.
Spliceosome assembly is a complex process involving the various snRNPs and the pre-mRNA. Auxiliary proteins are also involved in either spliceosome assembly or pre-mRNA splice site selection where alternative donor and acceptor splice site choices are possible. Upon excising the intron and ligating of the two exons, the spliceosome disassembles, and the ligated, processed mRNA is released. Individual snRNPs are conserved and reutilized for other splicing reactions.
The genes for transfer RNA in yeast, wheat germ, and vertebrates also contain small introns (6 to 80 nucleotides) just downstream of the central anticodon region. These introns are removed by endonuclease activities, and the two halves of the tRNA are ligated(15).
3. Subnuclear Organelles, Compartments, Particles, and their Functions
3.1. The Nuclear Envelope
The nucleus is bordered by two membrane bilayers—the nuclear envelope. Metabolically active cells (eg, the Xenopus oocyte) may display folds or blebs in the nuclear envelope as a mechanism to increase its surface area. The outer membrane is continuous with that of the endoplasmic reticulum. In fact, some electron micrographs show ribosomes on the surface of the outer nuclear membrane. The space between the two nuclear membranes, the perinuclear space, is continuous with the lumen of the endoplasmic reticulum. The nuclear side of the inner membrane is lined with a laminar matrix (the nuclear lamina) that consists of nuclear lamins A, B, and C, each an intermediate filament-type protein. The lamina lends support and shape to the nucleus and is considered part of the nuclear matrix.
The inner and outer membranes fuse at the nuclear pore complexes that traverse both bilayers. Nuclear pore complexes mediate bidirectional transport of macromolecules between the nucleus and the cytoplasm (see Nuclear Import, Export). An individual pore complex displays eightfold symmetry and is a huge macromolecular assembly with a combined mass of approximately 124 megadaltons.
3.2. Nuclear Matrix
The nuclear matrix is a three-dimensional fibrogranular latticework that permeates the nucleus but has been elusive to define completely. Components include the nuclear envelope with its pores, interchromatin RNP granules, perichromatin RNP fibrils, the nucleolus, tightly bound chromatin (matrix attachment regions, MARS), and associated pre-mRNA. Chromatin does not simply occupy random placement within the interphase nucleus (16, 17). Instead, specific regions of euchromatin and heterochromatin associate with the nuclear envelope(18). Observations such as these suggest that specific three-dimensional positioning of genetic loci occurs within the nucleus. This is an important aspect of the gene gating hypothesis of Blobel(19), which originally proposed that nuclear pore complexes, the nuclear lamina, and components of the nuclear core (the matrix) topologically position (eg, gate) transcribable chromatin near the envelope pores for efficient export of the mRNA.
3.3. The Nucleolus and Ribosomal Subunit Production
The nucleolus is the most conspicuous organelle within the interphase nucleus. Our traditional understanding of the nucleolus is that it is responsible for (1) synthesizing a large nascent precursor ribosomal RNA (pre-rRNA) that is 47S in mammals and approximately 13,000 nucleotides long in humans; (2) the processing (cleavage and base modification) of this RNA to yield mature ribosomal RNA of 18S (2,000 nucleotides), 5.8S (160 nucleotides), and 28S (5,000 nucleotides); and (3) the concomitant assembly of these RNAs with incoming ribosomal proteins to generate small and large ribosomal subunits that then pass into the cytoplasm (see Nucleolus).
Three distinct ultrastructural regions constitute the prototypical interphase nucleolus: the fibrillar center (FC), the surrounding dense fibrillar region (DFR), and the peripheral granular region (GR). Particles of preribosomal subunits are visible within the GR and are 15 to 20 nm in size. Vacuoles are sometimes found within nucleoli, but their significance is unknown. The relative sizes and distributions of the three principal nucleolar compartments vary among different cell types and among different metabolic states and cell cycle phases within the same cell type (21, 22).
The number of nucleoli per nucleus also differ. The yeast cell contains one relatively large nucleolus with respect to its nuclear volume, and most metazoan cells display one or a few nucleoli. At the other extreme, Xenopus oocytes contain over 1000 nucleoli per nucleus (germinal vesicle) because of selective ribosomal DNA amplification in the very early stages of meiosis I (pachytene). These multiple nucleoli are independent of the lampbrush chromosomes within these oocytes (see Fig. 1), a unique feature that allows exclusive study of nucleoli without interference from nonnucleolar chromatin. These nucleoli produce an incredible 300,000 ribosomes per second during mid-oogenesis in Xenopus .The number of nucleoli per nucleus also varies during the cell cycle. For example, cells in the early interphase may show several small nucleoli that reflect the number of NOR unique to the species. As the interphase progresses, the small nucleoli may fuse to reduce the number of nucleoli.
Nucleoli originate from and are intimately associated with the nucleolar organizer regions (NOR) that appear as secondary constrictions within certain ("nucleolar") mitotic chromosomes(20). The NOR are genetic loci that contain repeated 47S ribosomal RNA genes in tandem array. Each gene within the array is separated by intergenic spacer regions. The repeated ribosomal RNA genes each encode (in order) an external transcribed spacer, the 18S region, the first internal transcribed spacer, the 5.8S region, the second internal transcribed spacer, and finally the 28S region. Processing of the nascent transcript yields mature 18S, 5.8S, and 28S rRNA found in cytoplasmic ribosomes. The precise site of transcription has been controversial, although fibrillar centers (FC) of interphase nucleoli clearly contain ribosomal DNA. Good arguments place transcription within the FC, within the DFR, or on the borders between the two compartments . Nascent transcript processing (cleavage, folding, and base modification) and early ribosomal subunit assembly clearly take place within the DFR. Processing events include cleavage of the nascent transcript into smaller intermediates and then into the individual mature rRNA molecules, site-specific methylation of the pre-rRNA, and conversion of specific uridine residues to pseudouridine (see Ribosomes). Processing of the 47S transcript into the 18S rRNA (for the small ribosomal subunit) and the 5.8S and 28S rRNA (for the large subunit) occurs concomitantly with the association of ribosomal proteins that enter the nucleolus from their site of synthesis in the cytoplasm (there are approximately 85 different ribosomal proteins).
The 5S rRNA is transcribed by RNA polymerase III from multiple tandem genes that exist in genomic clusters outside the nucleolus. How the 5S transcripts make their way to the nucleolus and associate with the large ribosomal subunit remains uncertain. Assembly of large and small ribosomal subunits is a vectoral process beginning at the site of 47S RNA synthesis. By the time the particles reach the GR, the ribosomal subunits are nearly complete. Their assembly is completed once the subunits translocate to the cytoplasm, where a few more ribosomal proteins join the complexes. The large and small subunits finally associate as a complete ribosome in the recognition and translation of mRNA in the cytoplasm (see Translation).
3.4. Nuclear Distributions of Splicing Components
Antibodies against spliceosome components have been used in immunofluorescence microscopy to localize the sites of splicing and the nascent pre-mRNA within the nuclei of various cell types. The amphibian oocyte that has large lampbrush chromosomes has been particularly informative (25, 26). Antibodies directed against snRNPs or the splicing factor SC35 stain the lateral loops of giant lampbrush chromosomes within Notophthalmus viridescens (newt) oocytes (Fig. 1). The chromosomal loops are sites of active transcription, and their staining with antibodies directed against splicing components strongly argues that initial splicing events occur while the transcripts are still attached to RNA polymerase II. Beautiful electron micrographs of active Drosophila chorion genes clearly show the processing of nascent transcripts while still associated with RNA polymerase II(27).
Antibodies directed against splicing components brightly stain a limited number of discrete foci above a lightly and uniformly stained background in interphase nuclei. For example, anti-Sm antibodies stain 20 to 50 foci in the interphase nuclei of CHOC 400 cells (Fig. 2) (24, 28). These speckles, as they are called, contain interchromatin granule clusters (IGC) interconnected by perichromatin fibrils (PF) to form a nuclear latticework. Functionally, the ICG may be centers for snRNP assembly and distribution, whereas the PF are sites of pre-mRNA synthesis and splicing as the transcripts make their way to the nuclear envelope. This latticework of clusters and interconnecting fibrils is a dynamic structure that changes its morphology in response to changes in cell and nuclear physiology. In addition, the speckles themselves disassemble during mitosis and reassemble during the interphase. The existence and distinct distribution of the speckles within the interphase nucleoplasm clearly emphasizes the structural and functional compartmentalization of the interphase nucleus (Fig. 2).
Figure 2. Factors involved in pre-mRNA splicing are organized in a speckled distribution pattern in the interphase nuclei of mammalian cells. The speckles are composed of both interchromatin granule clusters (IGC) and perichromatin fibrils (PF). The PF represent sites of active transcription, and the IGC are thought to be storage and/or reassembly sites for splicing factors.
Using antibodies directed against the Sm antigens and the unique trimethylguanosine cap of snRNA in conjunction with in situ hybridizations to detect snRNA, Wu et al. (26)described three types of snRNP-containing granules (A, B, and C types), which are called "snurposomes"(29). Although the antibodies also stain the lampbrush chromosomes, for the most part these granules are independent of the chromosomes. The A granules contain exclusively U1 snRNP, whereas the B granules (Fig. 3) contain all five snRNPs(26). The C granules are most interesting. They are the originally described sphere organelles(30) found near or at the histone gene clusters on the large lampbrush chromosomes of at least two amphibian species (31-33). The C granules contain the U7 snRNP(34) reserved for histone transcript processing(35). Interestingly, B particles often reside on the surface of C granules or within C granules as inclusion bodies, but the functional significance of this relationship remains unknown.
Figure 3. Contents from a Xenopus laevis oocyte nucleus spread for light microscopic examination. (a) DIC image. The multiple nucleoli were stained with propidium iodide. Arrows show micronucleoli. (b) Snurposomes were stained with monoclonal antibody Y12 which specifically binds the Sm epitope of several snRNP proteins. See the text for descriptions of the nucleoli, B snurposomes, snRNP proteins, and Sm epitopes.
3.5. Coiled Bodies
Coiled bodies were first discovered by Ramon y Cajal in 1903(36) as nucleolar accessory bodies in neuronal cell nuclei, but their function remains uncertain. They were described in detail as internal coiled fibers at the ultrastructural level by Monneron and Bernhard(37), who first used the name coiled bodies. The bodies are spherical, have diameters of 0.5 ^m to 1 mm, and are found in both plant and animal cell nuclei. One to five coiled bodies exist per nucleus. The molecular constitution of coiled bodies has been examined only recently (24, 38-40). The protein p80-coilin is highly enriched in the coiled body and is considered a marker for the coiled body, although it has also been detected throughout the nucleus in relatively low concentrations (38, 41, 42). Its function is unknown.
Evidence suggests that the coiled body is related to the C snurposome (see previous): (1) Sph-1, a resident protein of the sphere organelle, is structurally homologous to p80-coilin(43); (2) antibodies directed against human p80-coilin stain the C granules intensely; (3) when mRNA-encoding, epitope-tagged, human p80-coilin was injected into Xenopus oocytes, the protein quickly localized to the C granules. Therefore coiled bodies and amphibian sphere organelles are likely to be similar in function. As in the case of amphibian C granule/sphere organelles, antibodies against Sm antigens and trimethylguanosine cap structures stain coiled bodies and, like the sphere organelles, coiled bodies in human and mouse culture cells contain high concentrations of U7 snRNA(44). Significant differences exist between coil bodies and the amphibian sphere organelles, however. First, unlike the sphere organelles, where only U7 snRNP has been detected, all five mRNA-splicing snRNAs (U1, U2, U4, U5, U6) have been detected within coiled bodies by in situ hybridization. Parenthetically, the coiled bodies are not the speckles described before because the splicing-dependent SC35 protein has not been observed within coiled bodies. Secondly, whereas no nucleolar components have been found in amphibian sphere organelles, fibrillarin and snRNA U3 (components intimately associated with nucleoli and preribosomal RNA processing, see Nucleolus) have been detected in coiled bodies of cultured cells (41, 45). As originally reported by Ramon y Cajal(36), coiled bodies closely associate with nucleoli(46) , but the physiological significance of this association remains uncertain.
Novel nuclear particles, referred to as Gemini of coiled bodies (Gems), were recently identified in HeLa cell nuclei using monoclonal antibodies against the spinal muscular atrophy (SMN) gene product(47). The SMN protein of 32 kDa binds several hnRNP proteins and the nucleolar protein, fibrillarin, to form nuclear particles that closely resemble and associate with coiled bodies. Although closely related to coiled bodies in size and composition, Gems may be more closely related to the amphibian sphere organelles (C snurposomes). Their function remains unknown.
4. Mitotic Regulation of Nuclear Disassembly and Reassembly
For cells to divide, the nucleus must disassemble, and the genome must be divided exactly in two. Impressive intracellular reorganizations occur with the onset of mitosis. Cytoplasmic microtubules disassemble into tubulin dimers that then reassemble to form the mitotic spindle. The endoplasmic reticulum and the Golgi apparatus vesiculate. The nuclear envelope and nucleoli disassemble, and chromatin supercondenses. The onset of mitosis is driven by the activity of maturation promoting factor (MPF), a complex between cyclin B and p34cdc2 kinase(48). MPF in turn is regulated by several factors and signal transduction events, one of which is the completion of genomic DNA synthesis and any requisite DNA repair. The driving mechanism behind these morphological changes is phosphorylation of cellular constituent proteins directly or indirectly by MPF (see Mitosis).
