Interaction of various nuclear matrix proteins with DNA

An attempt has been made to localize the previously detected strong and weak bonds of nuclear matrix proteins with DNA in some groups of proteins, using fractionation of the matrix into lamina and intranuclear fibrils, isolation of the "elementary globules", fractionation of matrix nucleoproteins on hydroxyapatite. It was shown that both weak and strong bonds are localized on the nuclear lamina and in the intranuclear fibrils. The single-stranded DNA enriched fraction of the matrix nucleoproteins contained mostly strong bonds. The strong bond is less resistant to pronase treatment. A method for isolating nuclear matrix nucleoprotein fractions carrying only strong or only weak bonds is proposed.     

A novel class of nuclear RNA (nRNA) with a long life span as a gene repressor candidate.

Different systemic organs of fetal mice were continuously labelled with 5-3H-uridine during the organogenesis periods, and chased for various lengths of time after birth. In the autoradiographs made on paraffin-embedded sections of the organs taken after the chase for longer periods than 1 week, including a 12-months chase, specific labels were present exclusively in all the nuclei. The specific nuclear labels were resistant to RNase A or H digestion and to acid hydrolysis with 1 N HCl at 60 degrees C for 5 min, but were completely abolished by DNase digestion or prolonged acid hydrolysis for 10 min, the optimum condition for the Feulgen reaction to stain DNA. Electrophoretic analysis of the total nucleic acids extracted from the different organs chased for 3 or 12 months showed all the tritium radioactivity to be present in the DNA fraction before digestion with DNase or RNase A, and to be completely absent from the DNA fraction and shifted to the RNA fraction, or to be largely destroyed by degradation, after each digestion, respectively. By HPLC analysis of the total nucleic acid extract after further successive digestions with nuclease P1 and alkaline phosphatase into the constituent nucleotides, it was shown that all tritium activity was incorporated in uridine, without any detectable label in other nucleotides. By the simultaneous labeling of human peripheral lymphocytes at the late S-phase with 5-3H-uridine and BrdU, it was demonstrated that the autoradiographic labels, which, this time, were labile to RNase A digestion, were present in the G-bands of the spread chromosomes as identified by BrdU immunohistochemistry. The findings strongly indicate the presence of a novel class of nuclear RNA (nRNA). This type of RNA (a) may be localized in the nucleus in close association or hybridization with nuclear DNA, (b) have a life span as long as that of the cell, and (c) be duplicated in the late phase of DNA replication. The nRNA may play a fundamental role as gene repressor existing in the G-bands of metaphase chromosomes in the process of ontogeny and cytodifferentiation.     

Low ionic strength extraction of nuclease-treated nuclei destroys the attachment of transcriptionally active DNA to the nuclear skeleton

We have studied how the conditions in which the nuclear matrix is isolated influence the association of transcribing DNA with the nuclear matrix. Extraction of nuclease-treated nuclei with a low ionic strength solution before a high salt nuclei with a low ionic strength solution before a high salt extraction completely abolishes this association. However, RNA removal by RNAase treatment does not affect the binding of transcribing DNA to the nuclear matrix. The nature of the association of active genes with the nuclear matrix is discussed.     

Solid phase gene extraction isolates mRNA at high spatial and temporal resolution

Rapid, localized changes in gene expression require mRNA extraction at high temporal and spatial resolution. Current small-scale mRNA extractions depend on the removal of the cells/tissue from an organism or preserved specimens. What these methods have in common is that they are destructive and do not distinguish between genomic DNA and RNA. Therefore, extracted (m)RNA is typically contaminated by extracted cytoplasm, nuclear DNA, or other compounds, and the required purification leads to loss of especially low-abundant mRNA. The need to repeatedly remove mRNA from living material has led to the development of solid phase gene extraction (SPGE). SPGE sampling can be achieved using gene-specific or generic sequences and is not species-specific. Here we demonstrate the versatility and validity of this novel RNA extraction by simultaneously profiling nanos and bicoid mRNA in individual Drosophila eggs. The SPGE technique detects previously described distribution profiles of nanos and bicoid. Its low impact is underscored by the normal development of repeatedly sampled eggs. In our study, quantification of actin mRNA in germinating flax seeds linked gene expression to distinct developmental processes. These data demonstrate the universality of SPGE as a simple generic, analytical, and diagnostic procedure.     

Spatial distribution of DNA loop attachment and replicational sites in the nuclear matrix

Biochemical fractionation was combined with high resolution electron microscopic autoradiography to study the localization in rat liver nuclear matrix of attached DNA fragments, in vivo replicated DNA, and in vitro synthesized DNA. In particular, we determined the distribution of these DNA components with the peripheral nuclear lamina versus more internally localized structural elements of isolated nuclear matrix. Autoradiography demonstrated that the bulk of in vivo newly replicated DNA associated with the nuclear matrix (71%) was found within internal matrix regions. A similar interior localization was observed in isolated nuclei and in situ in whole liver tissue. Likewise, isolated nuclear lamina contained only a small amount (12%) of the total matrix-bound, newly replicated DNA. The structural localization of matrix-bound DNA fragments was examined following long-term in vivo labeling of the DNA. The radioactive DNA fragments were found predominantly within interior regions of the matrix structure (77%), and isolated nuclear lamina contained less than 15% of the total nuclear matrix-associated DNA. Most of the endogenous DNA template sites for the replicative enzyme DNA polymerase alpha (approximately 70%) were also sequestered within interior regions of the matrix. In contrast, a majority of the endogenous DNA template sites for DNA polymerase beta (a presumptive repair enzyme) were closely associated with the peripheral nuclear lamina. A similar spatial distribution for both polymerase activities was measured in isolated nuclei before matrix fractionation. Furthermore, isolated nuclear lamina contained only a small proportion of total matrix-bound DNA polymerase alpha endogenous and exogenous template activities (3-12%), but a considerable amount of the corresponding beta polymerase activities (47-52%). Our results support the hypothesis that DNA loops are both anchored and replicated at nuclear matrix-bound sites that are predominantly but not exclusively associated with interior components of the matrix structure. Our results also suggest that the sites of nuclear DNA polymerase beta-driven DNA synthesis are uniquely sequestered within the characteristic peripheral heterochromatin shell and associated nuclear envelope structure, where they may potentially participate in DNA repair and/or replicative functions.     

Affinity purification of influenza virus ribonucleoprotein complexes from the chromatin of infected cells

Like all negative-strand RNA viruses, the genome of influenza viruses is packaged in the form of viral ribonucleoprotein complexes (vRNP), in which the single-stranded genome is encapsidated by the nucleoprotein (NP), and associated with the trimeric polymerase complex consisting of the PA, PB1, and PB2 subunits. However, in contrast to most RNA viruses, influenza viruses perform viral RNA synthesis in the nuclei of infected cells. Interestingly, viral mRNA synthesis uses cellular pre-mRNAs as primers, and it has been proposed that this process takes place on chromatin. Interactions between the viral polymerase and the host RNA polymerase II, as well as between NP and host nucleosomes have also been characterized. Recently, the generation of recombinant influenza viruses encoding a One-Strep-Tag genetically fused to the C-terminus of the PB2 subunit of the viral polymerase (rWSN-PB2-Strep) has been described. These recombinant viruses allow the purification of PB2-containing complexes, including vRNPs, from infected cells. To obtain purified vRNPs, cell cultures are infected, and vRNPs are affinity purified from lysates derived from these cells. However, the lysis procedures used to date have been based on one-step detergent lysis, which, despite the presence of a general nuclease, often extract chromatin-bound material only inefficiently. Our preliminary work suggested that a large portion of nuclear vRNPs were not extracted during traditional cell lysis, and therefore could not be affinity purified. To increase this extraction efficiency, and to separate chromatin-bound from non-chromatin-bound nuclear vRNPs, we adapted a step-wise subcellular extraction protocol to influenza virus-infected cells. Briefly, this procedure first separates the nuclei from the cell and then extracts soluble nuclear proteins (here termed the "nucleoplasmic" fraction). The remaining insoluble nuclear material is then digested with Benzonase, an unspecific DNA/RNA nuclease, followed by two salt extraction steps: first using 150 mM NaCl (termed "ch150"), then 500 mM NaCl ("ch500") (Fig. 1). These salt extraction steps were chosen based on our observation that 500 mM NaCl was sufficient to solubilize over 85% of nuclear vRNPs yet still allow binding of tagged vRNPs to the affinity matrix. After subcellular fractionation of infected cells, it is possible to affinity purify PB2-tagged vRNPs from each individual fraction and analyze their protein and RNA components using Western Blot and primer extension, respectively. Recently, we utilized this method to discover that vRNP export complexes form during late points after infection on the chromatin fraction extracted with 500 mM NaCl (ch500).     

DNA binding properties of the nuclear matrix and individual nuclear matrix proteins. Evidence for salt-resistant DNA binding sites

The DNA binding characteristics of the rat nuclear matrix were investigated. A saturable and temperature-dependent, salt-resistant DNA binding to the nuclear matrix was discovered, with 70-80% of total bound DNA resistant to extraction with high concentrations of salt at 37 degrees C, compared to less than 5% at 0 degrees C. The initial binding of DNA to nuclear matrix is sensitive to salt concentration, indicating a transition to a salt-resistant binding state. The nuclear matrix shows a preference for single-stranded DNA, both in saturation and competition assays, with little binding of RNA or double-stranded DNA. Further competition studies show a preference for matrix-attached DNA probably involving predominantly AT-rich sequences, while a specific sequence defined previously as a matrix-attached region (MAR; Cockerill, P. N., and Garrard, W. T. (1986) Cell 46, 273-282) only showed preference for a limited number of the total matrix binding sites. These results and estimates from saturation data of approximately 150,000 single-stranded DNA binding sites per matrix lead us to propose that the nuclear matrix contains different classes of DNA binding sites, each with a separate sequence specificity. Binding of DNA to individual matrix polypeptides separated on sodium dodecyl sulfate-polyacrylamide gels and transferred to nitrocellulose blots was also temperature-dependent, salt-resistant, and showed a preference for binding DNA over RNA and nuclear matrix DNA over total genomic DNA. Subnuclear fractionation experiments further demonstrated that the nuclear matrix is enriched in the subset of higher molecular weight (greater than 50,000) DNA binding proteins of isolated nuclei and correspondingly depleted of the lower molecular weight ones. Of the approximately 12 major proteins separated on nonequilibrium two-dimensional gels, 7 were identified as specific DNA binding proteins including lamins A and C (but not B), and the internal nuclear matrix proteins, matrins D, E, F, G, and 4.