Refine
Has Fulltext
- yes (24)
Is part of the Bibliography
- yes (24)
Year of publication
Document Type
Language
- English (24)
Keywords
Institute
- Theodor-Boveri-Institut für Biowissenschaften (24) (remove)
Several types of "irregular" structures in the arrangement of lateral fibrils were noted in electron microscopic preparations of transcriptionally active nucleolar chromatin from various plant and animal cells. Such forms include: I. Disproportionately long lateral fibrils which occur either as individual fibrils or in groups; 2. "Prelude complexes" and other arrangements of lateral fibrils in apparent spacer intercepts; 3. Thickening of the rDNA chromatin axis at the starting end of pre-rRNA matrix units; 4. Extremely long matrix units , the length of which exceeds that of the rDNA (double-strand) sequence complementary to the specific pre-rRN A (for abbreviations see text). In addition, the stability of high molecular weight RNAs contained in the nucleolar ribonucleoproteins during the preparation for electron microscopy was demonstrated by gel electrophoresis. The observations indicate that the morphological starting point of a pre-rRNA matrix unit is not necessarily identical with the initiation site for synthesis of pre-rRNA, but they rather suggest that the start of the transcriptional unit is located at least O.2-D.8 JLm before the matrix unit and that parts of the "apparent spacer" are transcribed. It is proposed that the pre-rRN A molecules do not represent the primary product of rDNA transcription but rather relatively stable intermediate products that have already been processed during transcription.
The ultrastructure of twO kinds of transcription ally active chromatin, the lampbrush chromosome loops and the nucleoli from amphibian oocytes and primary nuclei of the green alga Acetabularia, has been examined after manual isolation and dispersion in low salt media of slightly alkaline pH using various electron microscopic staining techniques (positive staining, metal shadowing, negative staining, preparation on positively charged films, etc.) and compared with the appearance of chromatin from various somatic cells (hen erythrocytes, rat hepatocytes, ClIltured murine sarcoma cells) prepared in parallel. While typical nucleosomes were revealed with all the techniques for chromatin from the latter three cell system, no nucleosomes were identified in either the lampbrush chromosome structures or the nucleolar chromatin. Nucleosomal arrays were absent not only in maximally fibril-covered matrix units but also in fibril-free regions between transcriptional complexes, including the apparent spacer intercepts between different transcriptional units. Moreover, comparisons of the length of the repeating units of rDNA in the transcribed state with those determined in the isolated rDNA and with the lengths of the first stable product of rDNA transcription, the pre-rRNA, demonstrated that the transcribed rDNA was not significantly shortened and/or condensed but rather extended in the transcriptional units. Distinct granules of about nucleosomal size which were sometimes found in apparent spacer regions as well as within matrix units of reduced fibril density were shown not to represent nucleosomes since their number per spacer unit was not inversely correlated with the length of the specific unit and also on the basis of their resistance to treatment with the detergent Sarkosyl NL-30. It is possible to structurally distinguish between transcriptionally active chromatin in which the DNA is extended in a non-nucleosomal form of chromatin and condensed, inactive chromatin within the typical nucleosomal package. The characteristic extended structure of transcriptionally active chromatin is found not only in the transcribed genes but also in non-transcribed regions within or between ("spacer") transcriptional units as well as in transcriptional units that are untranscribed amidst transcribed ones and/or have been inactivated for relatively short time. It is hypothesized that activation of transcription involves a transition from a nucleosomal to an extended chromatin organisation and that this structural transition is not specific for single "activated" genes but may involve larger chromatin regions, including adjacent untranscribed intercepts.
Some decades ago it was noted by cytologists that within the interphase nucleus large portions of the transcriptionally ("genetically," in their terms) inactive chromosomal material are contained in aggregates of condensed chromatin, the "chromocenters," whereas transcriptionally active regions of chromosomes appear in a more dispersed form and are less intensely stained with DNA-directed staining procedures (Heitz 1929, 1932, 1956; Bauer 1933). The hypothesis that condensed chromatin is usually characterized by very low or no transcriptional activity, and that transcription occurs in loosely packed forms of chromatin (including, in most cells, the nucleolar chromatin) has received support from studies of ultrathin sections in the electron microscope and from the numerous attempts to separate transcriptionally active from inactive chromatin biochemically (for references, see Anderson et al. 1975; Berkowitz and Doty 1975; Krieg and Wells 1976; Rickwood and Birnie 1976; Gottesfeld 1977). Electron microscopic autoradiography has revealed that sites of RNA synthesis are enriched in dispersed chromatin regions located at the margins of condensed chromatin (Fakan and Bernhard 1971, 1973; Bouteille et al. 1974; Bachellerie et al. 1975) and are characterized by the occurrence of distinct granular and fibrillar ribonucleoprotein (RNP) structures, such as perichromatin granules and fibrils. The discovery that, in most eukaryotic nuclei, major parts of the chromatin are organized in the form of nucleosomes (Olins and Olins 1974; Kornberg 1974; Baldwin et al. 1975) has raised the question whether the same nucleosomal packing of DNA is also present in transcriptionally active chromatin strands. Recent detailed examination of the morphology of active and inactive chromatin involving a diversity of electron microscopic methods, particularly the spreading technique by Miller and coworkers (Miller and Beatty 1969; Miller and Bakken 1972), has indicated that the DNA of some actively transcribed regions is not packed into nucleosomal particles but is present in a rather extended form within a relatively thin (4-7 nm) chromatin fiber.
Segregation of the nucleolar components is described in the differentiated nucleus of the generative cell in the growing Clivia and Lilium pollen tubes. This finding of a natural nucleolar segregation is discussed against the background of current views of the correlations of nucleolar morphology and transcriptional activity.
The arrangement of transcriptional units in the loops of lampbrush chromosomes from oocyte nuclei of urodele amphibia and from primary nuclei of the green alga Acetabularia have been studied in the electron microscope using spread preparations. Loops with different patterns of arrangement of matrix units (i.e. to a first approximation, transcriptional units) can be distinguished: (i) loops consisting of one active transcriptional unit; (ii) loops containing one active transcriptional unit plus additional fibril-free, i.e. apparently untranscribed, intercepts that may include 'spacer' regions; (iii) loops containing two or more transcriptional units arranged in identical or changing polarities, with or without interspersed apparent spacer regions. Morphological details of the transcriptional complexes are described. The observations are not compatible with the concept that one loop reflects one and only one transcriptional unit but, rather, lead to a classification of loop types according to the arrangement of their transcriptional units. We propose that the lampbrush chromosome loop can represent a unit for the coordinate transcription of either one gene or a set of several (different) genes.
The disintegration of the nuclear envelope has been examined in nuclei and nuclear envelopes isolated from amphibian oocytes and rat liver tissue, using different electron microscope techniques (ultrathin sections and negatively or positively stained spread preparations). Various treatments were studied, including disruption by surface tension forces, very low salt concentrations, and non ionic detergents such as Triton X-lOO and Nonidet P-40. The high local stability of the cylinders of nonmembranous pore complex material is emphasized. As progressive disintegration occurred in the membrane regions, a network of fibrils became apparent which interconnects the pore complexes and is distinguished from the pore complexassociated intranuclear fibrils. This network might correspond to an indistinct lamella, about 15 - 20 nm thick, located at the level of the inner nuclear membrane, which is recognized in thin sections to bridge the interpore distances. With all disintegration treatments a somewhat higher susceptibility of the outer nuclear membrane is notable, but a selective removal does not take place. Final stages of disintegration are generally characterized by the absence of identifiable, membrane- like structures. Analysis of detergent-treated nuclei and nuclear membrane fractions shows almost complete absence of lipid components but retention of significant amount of glycoproteins with a typical endomembrane-type carbohydrate pattern. Various alternative interpretations of these observations are discussed. From the present observations and those of Aaronson and Blobel (1,2), we favor the notion that threadlike intrinsic membrane components are stabilized by their attachment to the pore complexes, and perhaps also to peripheral nuclear structures, and constitute a detergent-resistant, interpore skeleton meshwork.